1
0
mirror of https://github.com/opencv/opencv.git synced 2026-07-24 12:53:04 +04:00

Compare commits

..

171 Commits

Author SHA1 Message Date
Alexander Smorkalov 0654a42e19 Release: OpenCV 4.14.0 2026-07-17 15:55:59 +03:00
Alexander Smorkalov f7dd7df170 Merge pull request #29531 from asmorkalov:as/openvino_ci_fail
Disable test test that sporadically fails with OpenVINO on CI #29531

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [ ] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2026-07-17 13:32:48 +03:00
Alexander Smorkalov 7aa163b83f Merge pull request #29539 from uwezkhan:aruco-readdict-marker-bound
bound marker string length in aruco readDictionary
2026-07-17 11:42:51 +03:00
Alexander Smorkalov c47541acbd Merge pull request #29525 from asmorkalov:as/ffmpeg_update_4.14
FFmpeg wrapper update for 4.14.0 release.
2026-07-17 10:37:20 +03:00
Uwez Khan b5fe8dfef3 bound marker string length in aruco readDictionary 2026-07-16 19:44:15 +05:30
Alexander Smorkalov 4d1e206f5e FFmpeg wrapper update for 4.14.0 release. 2026-07-16 15:12:50 +03:00
Yuantao Feng 2e778c52c1 Merge pull request #27510 from fengyuentau:4x/core/reduce_simd
core: vectorize cv::reduce #27510

- [ ] ~reduceR_~ Dropped due to performance
- [x] reduceC_

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2026-07-15 22:36:29 +03:00
Madan mohan Manokar 6b640b424c Merge pull request #28650 from amd:fast_pyrDown
imgproc: Optimized OpticalFlowPyrLK (PyrDownH) #28650

- Optimized Uchar PyrDown Horizontal processing intermediate storage.
- Improved horizontal processing with AVX512 vbmi ISA.
- Todo: further refinement for other archs.

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-07-15 12:31:28 +03:00
Alexander Smorkalov 06574736b3 Merge pull request #29503 from uwezkhan:flann-kdtree-load-bounds
bound kd-tree node indices when loading a saved flann index
2026-07-14 15:36:16 +03:00
Madan mohan Manokar 4fe51e51e0 Merge pull request #29507 from amd:fast_remap_ext
imgproc: Optimized remap interpolation #29507

- Add a SIMD dispatch file for remap and vectorize the single-channel (C1) in-bounds paths of bilinear, bicubic and lanczos4 interpolation (32F / 16U / 16S) using width-agnostic gather
- Dispatch the bilinear C1 path and drop the per-pixel weight-table gathers
- Widen the fixed-point coordinate map conversion
- 32F lanczos4 is kept on the scalar path: its vectorized 64-tap accumulation deviates beyond the set float accuracy tolerance
- Vectorize the bilinear inlier/outlier run detection so any-channel linear remap skips constant-status runs with SIMD instead of a per-pixel bounds test

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-07-14 15:17:22 +03:00
Alexander Smorkalov 335abd236f Merge pull request #29523 from asmorkalov:as/relax_gapi_gstreamer_test
Relaxed G-API tests for GStreamer to support Ubuntu 26.04.
2026-07-14 12:09:48 +03:00
Alexander Smorkalov 46d1b6c99d Relaxed G-API tests for GStreamer to support Ubuntu 26.04. 2026-07-14 08:47:33 +03:00
Alexander Smorkalov e6b1ef272e Merge pull request #29511 from Akansha-977:histogram_IPP_4.x
Extracted IPP to HAL for calcHist function in 4.x
2026-07-14 08:42:06 +03:00
Alexander Smorkalov 1bea199e00 Merge pull request #29437 from intel-staging:dev/tizmajlo/ippicv_20260630
Update IPP ICV integration to IPP 2026.0.0 (20260630)
2026-07-14 08:25:34 +03:00
Alexander Smorkalov 42cd88d737 Merge pull request #29509 from asmorkalov:as/ubuntu_26.04_warn_fix
Samples build warning fix on Ubuntu 26.04
2026-07-14 08:19:57 +03:00
Alexander Smorkalov 4c3895e96c Samples build warning fix on Ubuntu 26.04 2026-07-13 15:22:08 +03:00
Akansha Mallick 2c14cc1897 Extracted IPP to HAL for calcHist function 2026-07-13 14:12:52 +05:30
Uwez Khan 13fb140932 bound kd-tree node indices when loading a saved flann index 2026-07-12 21:36:42 +05:30
Alexander Smorkalov 7f9f6acb83 Merge pull request #29502 from asmorkalov:as/msmf_audio
videoio(MSMF): fix audio sample-rate validation and dropped last frame
2026-07-12 17:11:55 +03:00
Varun Jaiswal 4533beab85 videoio(MSMF): fix audio sample-rate validation and dropped last frame (#27438) 2026-07-12 12:14:59 +03:00
Akansha-977 87af90b847 Merge pull request #29463 from Akansha-977:distransform_IPP_migration_4.x
Distransform function IPP migration to HAL in 4.x #29463

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-07-10 18:28:10 +03:00
FAN YUCHEN d7e6e58652 Merge pull request #29487 from Functionhx:fix/docs-combined
Fix calibration tutorial docs, decomposeProjectionMatrix, and convertMaps performance claims #29487

Fixes #25655, #26791, #27277.

Three doc fixes:
1. Calibration tutorial: rows/cols swapped, fixed np.mgrid consistency
2. decomposeProjectionMatrix: clarified transVect is camera center in homogeneous coordinates
3. convertMaps: replaced overstated 2x speed claim

### Pull Request Readiness Checklist
- [x] I agree to contribute under Apache 2 License
- [x] Not based on GPL/incompatible license
- [x] PR proposed to proper branch (4.x)
- [x] Reference to original bug report and related work
- [ ] Accuracy test, performance test, test data: N/A (doc-only)
- [x] Feature well documented and sample code buildable
2026-07-10 15:59:29 +03:00
LiuChenjiayi b21eae2b8b Merge pull request #29149 from EuropaRanger:gapi_doc_update
Gapi doc update #29149

This PR improves G-API documentation by clarifying the introduction and kernel API pages
- fixing some spelling errors
- operator|() can now link to the overloaded version that is actually called.
- supplemented and revised the doc based on the author's annotations.

The change is documentation-only and does not affect runtime behavior. I tested the doc reconstruction of this module. 

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2026-07-10 15:44:14 +03:00
Alexander Smorkalov b41dd0fdff Merge pull request #29491 from Akansha-977:threshold_IPP_4.x
Extracted IPP to HAL for threshold function in 4.x
2026-07-10 15:06:40 +03:00
Alexander Smorkalov be4ef205b1 Merge pull request #29490 from asmorkalov:as/disable_ipp_warpPerspective_F32C4_nearest
Disabled IPP implementation for warpPerspective F32C4 as it triggers exception on Windows
2026-07-10 14:19:41 +03:00
Akansha Mallick 062a00bab3 Extracted IPP to HAL for threshold function 2026-07-10 14:00:33 +03:00
Prasad Ayush Kumar b022a9b321 Merge pull request #29434 from Prasadayus:canny_ipp_extract
Extracting IPP integaration as HAL for Canny #29434

Backport of : https://github.com/opencv/opencv/pull/29433

**Performance Numbers on Intel(R) Core(TM) i9-11900K:** https://docs.google.com/spreadsheets/d/1RMvUZP1tSQtdjuNQY9JasYmlx1L5iN9XWGhXtG5u1uI/edit?usp=sharing 

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-07-10 13:59:24 +03:00
Akansha-977 7230c1ce08 Merge pull request #29464 from Akansha-977:templatematch_IPP_4.x
Extracting IPP to HAL for matchTemplate function in 4.x #29464

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-07-10 12:16:45 +03:00
Alexander Smorkalov 818145f41d Disabled IPP implementation for warpPerspective F32C4 as it does not pass tests. 2026-07-10 09:13:56 +03:00
Andrei Fedorov 93cd614472 Merge pull request #29477 from intel-staging:ipp/sort_enabling
Enable ipp calls for sort, sortIdx#29477

Follow up on https://github.com/opencv/opencv/pull/29184
Also introduced parallelization to sort similar to #29192
Performance is up to ~17x faster per our measurements
+ ~127KB to binary size.

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2026-07-10 08:39:21 +03:00
Alexander Smorkalov 7507accf2e Merge pull request #29486 from asmorkalov:as/drop_yuv_gray_hal
Drop HAL for cv_hal_cvtColorYUV2Gray as it's just copy.
2026-07-09 18:06:28 +03:00
Alexander Smorkalov 8eb2c8998c Drop HAL for cv_hal_cvtColorYUV2Gray as it's just copy. 2026-07-09 16:51:20 +03:00
Alexander Smorkalov 738739d7c2 Merge pull request #29483 from asmorkalov:update_version_4.14.0-pre
pre: OpenCV 4.14.0 (version++)
2026-07-09 15:35:03 +03:00
Alexander Smorkalov e978c193cf pre: OpenCV 4.14.0 (version++) 2026-07-09 12:35:42 +03:00
yyyisyyy11 b887e2fd3e Merge pull request #29018 from yyyisyyy11:project4_Lunhan_Yan
dnn: add DynamicQuantizeLinear ONNX layer support #29018

### Pull Request Readiness Checklist

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and target platform(s)
- [x] There is an accuracy test
- [ ] There is a performance test

### Description

Implements the ONNX `DynamicQuantizeLinear` operator (opset 11) for the OpenCV DNN module.

**What it does:**

- Adds `QuantizeDynamicLayer` and `DequantizeDynamicLayer` layer classes
- Registers layers and adds ONNX importer dispatch for `DynamicQuantizeLinear`
- Computes scale and zero-point at runtime from activation min/max
- Quantizes FP32 input to int8 (stored as uint8 - 128, matching OpenCV convention)

**Framework limitation & workaround:**
Due to the single-dtype-per-layer constraint in `LayerData::dtype` (see #29017), the float32 scale output cannot be passed through the CV_8S blob pipeline directly. As a workaround, the scale is encoded as 4 raw bytes in a CV_8S `{1,4}` blob using `memcpy`, and decoded by the downstream `DequantizeDynamic` layer.

**Testing:**

- Custom accuracy tests reproduce all 3 ONNX conformance test cases: `test_dynamicquantizelinear`, `test_dynamicquantizelinear_max_adjusted`, `test_dynamicquantizelinear_min_adjusted`
- Each test verifies: quantized values, scale, zero point, and round-trip dequantize accuracy
- All existing quantization regression tests pass (42/42)

**Conformance tests:**
The 6 conformance tests for `DynamicQuantizeLinear` remain in the parser denylist (they were already denylisted before this PR) because the framework cannot produce mixed-type outputs. The custom tests provide equivalent coverage.

### Files changed

- `modules/dnn/include/opencv2/dnn/all_layers.hpp` — layer class declarations
- `modules/dnn/src/init.cpp` — layer registration
- `modules/dnn/src/onnx/onnx_importer.cpp` — ONNX import dispatch
- `modules/dnn/src/int8layers/quantization_utils.cpp` — layer implementations
- `modules/dnn/test/test_onnx_importer.cpp` — custom accuracy tests
2026-07-09 11:07:12 +03:00
Alexander Smorkalov 0eb3778a1b Merge pull request #29473 from velonica0:28870
imgproc: make RVV HAL resize bit-exact for 8U INTER_LINEAR (fixes #28870)
2026-07-08 19:19:35 +03:00
velonica0 9121ebe28b imgproc: make RVV HAL resize bit-exact for 8U INTER_LINEAR 2026-07-08 05:24:31 -07:00
Anushka d48bf69f65 Merge pull request #29455 from anushkagupta200615-jpg:fix-issue-29452
Fix #29452: Remove <complex.h> to prevent _Complex macro conflicts #29455

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake

---

**Description:**
Resolves https://github.com/opencv/opencv/issues/29452

**Reason for the issue:**
The C99 `<complex.h>` header defines the macro `complex` on some platforms (like NetBSD with GCC 14). Because it was included before C++ `<complex>`, this caused conflicts where `std::complex<T>` was being expanded into `std::_Complex<T>`, resulting in the reported syntax errors.

**Changes made:**
- Removed the unnecessary C-style `#include <complex.h>`.
- Added a safety guard to `#undef complex` in case any transitive lapack headers attempt to define it, ensuring `std::complex` works cleanly without C-preprocessor interference.
2026-07-07 15:20:09 +03:00
Alexander Smorkalov aa70c82ebe Merge pull request #29418 from uwezkhan:einsum-subscript-bound
bound einsum subscript index before reading input shape
2026-07-07 14:42:49 +03:00
Alexander Smorkalov c85397c3bb Merge pull request #29462 from intel-staging:use_ipp_check_derive
Added missing IPP check in IPP HAL
2026-07-07 13:20:45 +03:00
Andrei Fedorov 0488f6e942 Update deriv_ipp.cpp 2026-07-07 11:39:39 +02:00
Alexander Smorkalov 51495995d2 Merge pull request #29431 from asmorkalov:as/no_exact_ipp
Enable non-exact IPP optimizations if build-time algorithm hint allows it
2026-07-07 11:45:59 +03:00
Alexander Smorkalov c344b7dcef Merge pull request #29461 from rmsalinas:update-truco-ref
doc: update TRUCO publication status
2026-07-07 10:21:31 +03:00
rmsalinas 0e6433327a doc: update TRUCO publication status 2026-07-07 08:08:43 +02:00
Alexander Smorkalov f6be96989e Merge pull request #29442 from orbisai0security:fix-filter-buffer-size-overflow
fix: in filter in filter.cpp
2026-07-06 17:19:05 +03:00
MAAZIZ Adel Ayoub e9289fc7c4 Merge pull request #29447 from Adel-Ayoub:fix/matexpr-mul-scalar-lifetime
core: fix use-after-scope when Mat::mul() is given a scalar #29447

### Summary

`cv::Mat::mul()` called with a scalar returns a `MatExpr` that reads a dead stack slot when it is evaluated. In a normal (non-instrumented) build this produces silently wrong values as soon as the slot is reused:

```cpp
static cv::MatExpr makeExpr(const cv::Mat& m)
{
    return m.mul(7);               // 7.0 is a temporary double in THIS frame
}

cv::Mat matrix(2, 3, CV_32FC1, cv::Scalar(3.0f));
cv::MatExpr expr = makeExpr(matrix);
// ... any further calls reuse the dead frame ...
cv::Mat result = expr;             // observed: all 0, expected: all 21
```

Under AddressSanitizer this is the `stack-use-after-scope` reported in #23577, with the same stack trace (`cvt64s` -> `convertAndUnrollScalar` -> `arithm_op` -> `multiply` -> `MatOp_Bin::assign`).

Storing the expression is the documented lazy-evaluation usage of `MatExpr`; the argument is ordinary supported API usage (`mat.hpp` itself shows `Mat C = A.mul(5/B);`).

### Root cause

A scalar argument binds to `_InputArray(const double& val)`, which records the **address** of the temporary with kind `MATX`:

```cpp
inline _InputArray::_InputArray(const double& val)
{ init(FIXED_TYPE + FIXED_SIZE + MATX + CV_64F + ACCESS_READ, &val, Size(1,1)); }
```

`Mat::mul()` then parks `m.getMat()` inside the returned `MatExpr`. For `MATX` kind, `getMat_()` returns a non-owning, non-refcounted header over that stack memory (`return Mat(sz, flags, obj);`). The temporary dies at the end of the full expression, but the `MatExpr` keeps the header, and `MatOp_Bin::assign()` later feeds it to `cv::multiply()`. `Matx`/`Vec` arguments take the same path.

`Mat::mul()` is the only `MatExpr` factory in `matrix_expressions.cpp` that takes an `InputArray`; every other scalar operand there is stored by value in the `Scalar` member (`e.s`), so no other expression path can capture a stack pointer this way.

### Fix

Snapshot the operand with `clone()` unless it is a `Mat`/`UMat`, which keep the current zero-copy behaviour: their headers are refcounted and already safe to defer. Any other `InputArray` kind (a scalar, `Matx`, `Vec`, `std::vector`, an evaluated expression) is a potentially non-owning view, so it is copied once at expression construction, off any hot path.

### Test

Adds `Core_MatExpr.mul_scalar_use_after_scope_23577` to `modules/core/test/test_operations.cpp`. It builds the expression in a helper frame and overwrites the stack before evaluating; the helpers are called through volatile function pointers so they cannot be inlined, which makes the stale read deterministic. The test fails before the fix (result is all 0 instead of all 21) and passes after. It is self-contained: no opencv_extra data is needed.

Verified locally on macOS/AArch64 (Apple clang 17, Release): full `opencv_test_core` passes, and the AddressSanitizer reproducer from the issue is clean after the fix.

Fixes #23577.

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Self-contained accuracy regression test in `modules/core/test/test_operations.cpp`; no opencv_extra data required. No performance test: no existing perf test covers `Mat::mul` expression construction, and the copy happens once at expression construction, only for non-`Mat`/`UMat` operands.
- [ ] The feature is well documented and sample code can be built with the project CMake
      N/A - bug fix, no new API.
2026-07-06 17:11:51 +03:00
Alexander Smorkalov 50beb24c6b Enable non-exact IPP optimizations if build-time algorithm hint allows it. 2026-07-06 11:50:53 +03:00
Alexander Smorkalov fcc3418cd7 Merge pull request #29446 from anushkagupta200615-jpg:fix-apple-conversions-cpp26
macOS: Fix CGBitmapInfo C++26 enum compilation error
2026-07-06 10:15:38 +03:00
Akansha-977 80fda1da8c Merge pull request #29427 from Akansha-977:filter2D_IPP_migration_4.x
Filter2D IPP migration to HAL for 4.x #29427

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-07-06 09:52:20 +03:00
anushkagupta200615-jpg cddfd62a18 macOS: Fix CGBitmapInfo C++26 enum compilation error 2026-07-06 02:58:34 +05:30
orbisai0security a9573e5ac0 fix: V-002 security vulnerability
Automated security fix generated by OrbisAI Security
2026-07-05 12:15:53 +00:00
Timur Izmajlov 9d66a589b4 Update IPP ICV integration to IPP 2026.0.0 (20260630) 2026-07-03 17:25:24 +02:00
Alexander Smorkalov 24030ff3a2 Merge pull request #29390 from amd:fast_countNonZero
core: Optimized countNonZero with AVX-512 signmask path
2026-07-03 15:44:43 +03:00
Prasad Ayush Kumar c77286749a Merge pull request #29420 from Prasadayus:box_filter_ipp_extract
Extract IPP integration as HAL function for box_filter #29420

Backport of https://github.com/opencv/opencv/pull/29414

**Performance Numbers on Intel(R) Core(TM) i9-11900K:** https://docs.google.com/spreadsheets/d/1kMKiZWh--pH30hqsQo6j1suNw1lfQiKzSankMCMa2FI/edit?usp=sharing

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-07-02 13:10:52 +03:00
Madan mohan Manokar e6d0c0340b Merge pull request #29413 from amd:fast_basic_op
core: Fix mul32f and addWeighted32f to use native f32 SIMD paths #29413

OpenCV Extra: https://github.com/opencv/opencv_extra/pull/1388

- add 32FC1 coverage to addWeighted benchmark
- avoid intermediate double for f32 variants of scaled multiply and addWeighted.
- Relax AddWeighted 32F test tolerance to match f32 FMA semantics.

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-07-02 12:58:39 +03:00
Alexander Smorkalov 9460e30793 Merge pull request #29400 from amd:fast_sum
core: enable AVX-512 dispatch for sum
2026-07-02 12:56:27 +03:00
Uwez Khan b4164a4bc4 bound einsum subscript index before reading input shape 2026-07-01 18:19:00 +05:30
Madan mohan Manokar 579a505734 Merge pull request #29335 from amd:fast_norm_simd
core: SIMD optimizations for norm, distance and Hamming APIs. (Improves ORB & BRISK) #29335

Generic universal-intrinsic kernels (benefit all SIMD backends: NEON, AVX2, AVX-512, etc.), plus enabling wider dispatch for the norm module.

- hal::normHamming: cached-pointer dispatch (resolve once, no per-call dispatch chain or trace region) + vector popcount path. cv::norm(NORM_HAMMING) and binary-descriptor matching (BFMatcher ORB/BRISK/FREAK via cv::batchDistance).
- hal::normL2Sqr_ / normL1_: direct inlinable kernels with single-vector tail (cv::batchDistance / BFMatcher float L2/L1, cv::kmeans).
- cv::norm masked NORM_INF: deinterleave SIMD for multichannel + back-step tail.
- cv::norm(src1, src2, type, mask): SIMD masked norm-diff; INF is one templated kernel for all element types, plus uchar L1/L2 and int L1 kernels.
- Enable AVX512_SKX/AVX512_ICL dispatch for the norm module.
- features2d: add BFMatcher knnMatch perf tests (float L2/L1, binary Hamming).
- ORB and BRISK performance improved
### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-07-01 15:48:47 +03:00
Alexander Smorkalov 83ee55fcda Merge pull request #29417 from intel-staging:deriv_hal
Moved IPP-based Sobel and Scharr to HAL
2026-07-01 14:20:57 +03:00
Madan mohan Manokar 28a1d0dddb Merge pull request #29242 from amd:fast_gemm_simd
Optimized gemm implementation with Universal SIMD #29242

- vectorized GEMMSingleMul and GEMMBlockMul

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-07-01 13:47:43 +03:00
Fedorov, Andrey 2f779c5c36 moved ipp deriv to hal 2026-07-01 02:13:48 -07:00
Alexander Smorkalov 5b83b0c440 Merge pull request #29405 from Teddy-Yangjiale:rvv-dnn-conv-4x
dnn: add RVV LMUL=1 kernel for convBlock_F32 (ConvolutionLayer, rv64gcv)
2026-06-29 18:08:55 +03:00
ZC a16c4a9fe2 Merge pull request #29359 from zcinclude:fix/avfoundation-videowriter-destructor-sigsegv
videoio(avfoundation): fix SIGSEGV crash when releasing VideoWriter on iOS #29359

Replace deprecated synchronous finishWriting with
finishWritingWithCompletionHandler + dispatch_semaphore
to properly wait for the async completion handler.

The synchronous finishWriting method is deprecated since iOS 6
and, despite blocking until writing status completes, does NOT
wait for internal NSOperation KVO observer blocks that are
dispatched asynchronously to GCD worker threads. When the
destructor returns and drains the autorelease pool, these
blocks may access already-released objects, causing SIGSEGV.

This fix uses a semaphore to block until the
finishWritingWithCompletionHandler callback has fully completed,
eliminating the race condition between the async block and
autorelease pool drain.

Fixes #28165
2026-06-29 15:22:16 +03:00
Anshu 2a47e2c1a5 Merge pull request #29396 from 0AnshuAditya0:fix-macos-framework-rpath-29028
cmake: fix #29028 macOS dynamic framework RPATH for relocatable install #29396

### Problem
When building OpenCV as a dynamic framework on macOS (`-DAPPLE_FRAMEWORK=ON -DBUILD_SHARED_LIBS=ON`), installed binaries contain an absolute `LC_RPATH` pointing to the build tree, making the install non-relocatable. Moving the install directory or deploying to another Mac causes binaries to crash on launch.

### Root cause
`cmake/OpenCVInstallLayout.cmake` unconditionally sets `CMAKE_INSTALL_RPATH` to `${CMAKE_INSTALL_PREFIX}/${OPENCV_LIB_INSTALL_PATH}` for all platforms. Additionally, `cmake/platforms/OpenCV-Darwin.cmake` was empty and missing the `@executable_path`/`@loader_path` linker flags that the iOS toolchain already sets for the same `APPLE_FRAMEWORK AND BUILD_SHARED_LIBS` case.

### Fix
- Guard `CMAKE_INSTALL_RPATH` in `OpenCVInstallLayout.cmake` to use relocatable `@executable_path` tokens for Apple framework builds
- Add equivalent linker flags to `OpenCV-Darwin.cmake`, mirroring `platforms/ios/cmake/Modules/Platform/iOS.cmake`

Fixes #29028

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2026-06-29 14:27:21 +03:00
Jonas Perolini 37f3c7539b Merge pull request #29252 from JonasPerolini:pr-aruco-bit-threshold-in-refine
Use validBitIdThreshold for Aruco refineDetectedMarkers #29252

The goal of this PR is to solve the issue raised by @vrabaud in https://github.com/opencv/opencv/pull/28289 (comment: https://github.com/opencv/opencv/pull/28289#discussion_r3355812646). 

**Issue:** 

`refineDetectedMarkers()` converted the extracted cell ratios with `convertTo(CV_8UC1)` (an implicit 0.5 threshold) before computing the code distance, ignoring `detectorParams.validBitIdThreshold`. 

**Solution:** 

Make the refine path consistent with the main detection path `Dictionary::identify`.

**Changes:**

- Add a `Dictionary::getDistanceToId()` overload that takes the float cell pixel ratio matrix and `validBitIdThreshold` (similar to how it's done for the `identify()` overload.
- Move the per cell distance computation into a private `getDistanceToIdImpl` helper used by both `identify()` and the new overload of `getDistanceToId()` to avoid repetitions.
- `refineDetectedMarkers()` now calls the new overload.

**Tests:**

- `CV_ArucoRefine.validBitIdThreshold`: a marker with one degraded cell is recovered at threshold 0.7 but not at 0.49.
- `CV_ArucoDictionary.getDistanceToIdCellPixelRatio`: unit-tests both `getDistanceToId` overloads.

All passed
2026-06-29 13:37:43 +03:00
Alexander Smorkalov 3c10324630 Merge pull request #29404 from tonuonu:perf-svm-kernels
ml: SIMD for SVM kernel reductions 🧑‍💻🤖
2026-06-29 11:40:14 +03:00
Teddy-Yangjiale c831290769 dnn: add RVV LMUL=1 kernel for convBlock_F32 (ConvolutionLayer, rv64gcv) 2026-06-29 00:47:44 +08:00
Tonu Samuel a015333f04 ml: SIMD for SVM kernel reductions
SVM::predict is dominated by the per-feature kernel reduction over the
support vectors. Vectorize the four reduction kernels in svm.cpp with
universal intrinsics (two independent accumulators + scalar tail):
calc_non_rbf_base (dot product), calc_rbf (squared distance),
calc_intersec (min-sum) and calc_chi2. Same approach as the KNN
findNearest reduction in #29380. Adds modules/ml/perf/perf_svm.cpp
covering the RBF/POLY/INTER/CHI2 kernels.
2026-06-28 18:09:05 +03:00
Mulham Fetna 5dbfbcdcac Merge pull request #28935 from molhamfetnah:fix/21960-unicode-temp-path
core: fix Unicode temp path handling on Windows (fix #21960) #28935

## Summary

Fixes getCacheDirectoryForDownloads and related temp file functions failing when the user home directory contains Unicode characters like C:\\Users\\テスト\\AppData\\Local\\Temp.

## Root Cause

On Windows, OpenCV was using ANSI versions of GetTempPath and GetTempFileName which fail with Unicode paths.

## Fix

Replace with wide-character GetTempPathW plus UTF-8 conversion:

- modules/core/src/utils/filesystem.cpp - getCacheDirectory for cache paths
- modules/core/src/system.cpp - temporary file creation  
- modules/ts/src/ts_gtest.cpp - test stream capture

## Testing

The fix has been verified compiles. Manual testing on Windows with Unicode username required.

## Risk

Low: Uses same output format, just different encoding path. Existing ASCII functionality preserved.

---

Fixes opencv/opencv#21960
2026-06-28 14:20:45 +03:00
Alexander Smorkalov edd89cfca5 Merge pull request #29371 from uwezkhan:tflite-pad-resize-bounds
check parameter tensor length in tflite parsePadding and parseResize
2026-06-28 14:19:17 +03:00
uwezkhan 15b5d28325 Merge pull request #29378 from uwezkhan:qr-alpha-map-bound
bound alphanumeric values in qr decodeAlpha before map lookup #29378

decodeAlpha in the no-quirc QR backend reads alphanumeric symbols from the post-ECC bitstream and indexes a fixed 45-entry map[] with the raw values. The 11-bit pair from next(11) goes up to 2047, so tuple/45 lands on 45 once the pair passes 2024, and the 6-bit trailing char from next(6) goes up to 63, so map[value] runs out to map[63]. A QR whose data codewords carry an alphanumeric segment with one of those out-of-range values reads past the static array, and the stray byte lands in the string returned by detectAndDecode. WITH_QUIRC defaults off, so this is the decoder a stock build runs.

Before, the only guard was on the encode side, which never emits those values, so the decoder trusted the stream and indexed map[] directly. New version checks value range and return empty string for malformed qr codes.

### Pull Request Readiness Checklist

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
- [ ] The feature is well documented and sample code can be built with the project CMake
2026-06-28 14:18:20 +03:00
Alexander Smorkalov dd541965e9 Merge pull request #29397 from Prasadayus:cvtcolor_ipp_extract
Extract IPP integration as HAL function for cvtcolor
2026-06-27 08:32:00 +03:00
Alexander Smorkalov da4a417b2a Merge pull request #29346 from Ijtihed:fix/flann-heap-pool-tls
flann: use per-thread storage for heap pool to remove lock contention
2026-06-26 16:35:27 +03:00
Alexander Smorkalov 416e730a61 Merge pull request #29395 from asmorkalov:as/restore_ipp_check
Restore missing IPP check in IPP HAL.
2026-06-26 16:26:17 +03:00
uwezkhan 69d2303531 Merge pull request #29345 from uwezkhan:onnx-tile-bounds
bound tile axis and repeats length in onnx parseTile #29345

OpenCV Extra: https://github.com/opencv/opencv_extra/pull/1385

parseTile sizes repeats_vec from the input-0 rank, then fills it from fields of the model that are never checked against that size. In the tile-1 path the axis taken from the third input indexes repeats_vec directly, and in the tile>1 path the loop writes one entry per element of the repeats tensor. A crafted ONNX with an out-of-range axis, or a repeats tensor longer than the input rank, writes past repeats_vec while loading the model through readNetFromONNX.

The fix runs axis through normalize_axis, the same helper the squeeze and concat paths in this file already use, so a negative or oversized axis is rejected before the write, and it checks the repeats length equals the input rank before the loop. Keeping both bounds in the parser puts the check next to the write instead of trusting the model to be well formed. Before, a repeats tensor shorter than the rank was silently accepted; after, it is rejected, which matches the ONNX rule that repeats carries one entry per input dimension.

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
- [ ] The feature is well documented and sample code can be built with the project CMake
2026-06-26 12:39:04 +03:00
Tõnu Samuel b64a215e71 Merge pull request #29377 from tonuonu:perf-dnn-activation-simd
dnn: SIMD for transcendental activation layers (15 functors) 🧑‍💻🤖 #2937

# dnn: SIMD for transcendental activation layers (15 functors) 🧑‍💻🤖

### Summary

Many `BaseDefaultFunctor`-based activation functors fall back to the scalar `BaseDefaultFunctor::apply`
— one `libm` call per element. This PR adds vectorized `apply()` overrides (universal intrinsics,
matching the existing `MishFunctor`/`GeluFunctor`/`SigmoidFunctor` pattern, each with a scalar tail) to
the **15 transcendental activations** that have a clean SIMD path:

| functor | uses | | functor | uses |
|---|---|---|---|---|
| TanH | `v_exp` (1−2/(e²ˣ+1)) | | Asinh | `v_log`,`v_sqrt` (sign·log(\|x\|+√(x²+1))) |
| Log | `v_log` | | Acosh | `v_log`,`v_sqrt` |
| Erf | `v_erf` | | Atanh | `v_log` |
| Exp | `v_exp` | | Softplus | `v_exp`,`v_log` |
| Sin | `v_sin` | | BNLL | `v_exp`,`v_log` (max(x,0)+log1p(e^−\|x\|)) |
| Cos | `v_cos` | | GeluApproximation | `v_exp` (tanh) |
| Tan | `v_sin`/`v_cos` | | Sinh / Cosh | `v_exp` |

Functors compile at the baseline SIMD width (SSE on x86, NEON on ARM — same as the existing SIMD
functors). `Sqrt`/`Floor`/`Ceil`/`Round`/`Abs`/etc. are intentionally **not** included — those are
single cheap ops the compiler already auto-vectorizes (measured `Sqrt` at 0.95×). `Asin`/`Acos`/`Atan`
are omitted (no `v_atan`/`v_asin`/`v_acos` intrinsic).

### Performance

Real `opencv_perf_dnn` `Layer_Activation` (added here), 8×256×128×100 = 26.2 M-elem CV_32F blob, A/B
vs the scalar fallback, median speedup:

| functor | M4 clang/NEON | A76 gcc/NEON | Threadripper gcc | Xeon W-2235 gcc | functor | M4 | A76 | TR | Xeon |
|---|---|---|---|---|---|---|---|---|---|
| TanH | 9.1× | 2.3× | 7.3× | 7.3× | Sinh | 3.3× | 3.2× | 5.2× | 5.0× |
| Cos | 8.8× | 3.0× | 3.3× | 3.0× | Acosh | 3.2× | 2.9× | 2.7× | 2.6× |
| Log | 7.8× | 3.0× | 2.5× | 2.4× | Cosh | 2.8× | 3.2× | 2.6× | 2.5× |
| Sin | 7.8× | 3.0× | 3.4× | 2.9× | Exp | 2.4× | 2.5× | 1.4× | 1.5× |
| GeluApprox | 7.6× | 2.4× | 6.6× | 5.8× | BNLL | 4.2× | 2.0× | 1.6× | 1.4× |
| Atanh | 7.0× | 3.6× | 6.0× | 5.2× | Softplus | 2.0× | 2.0× | 3.3× | 2.7× |
| Tan | 6.8× | 3.6× | 6.6× | 6.0× | Erf | 4.2× | 2.4× | 3.9× | 3.5× |
| Asinh | 3.9× | 2.4× | 5.8× | 5.7× | | | | | |

Every functor is a win on every tested out-of-order core (worst case 1.4×). On the **in-order
Cortex-A55** (a55-tuned build) all 15 measure 0.985–0.99× — neutral, within noise: the SIMD
polynomials are dependency-chain-bound, which an in-order pipeline can neither accelerate nor (here)
slow down. So: meaningful wins on out-of-order ARM + x86, no regression on in-order ARM.

### Accuracy

Uses the `v_exp`/`v_log`/`v_erf`/`v_sin`/`v_cos` polynomials OpenCV already ships and relies on (Gelu
via `v_erf`, Mish/Sigmoid via `v_exp`). Verified vs scalar `libm` on 320 K random elements per functor
with domain-correct inputs (Acosh x≥1, Atanh |x|<1, …): **max abs error ≤ 3e-5** (most ≤ 2e-6), zero
elements exceeding rel>1e-3 & abs>1e-4. ONNX conformance and the layer accuracy tests exercise these ops.

### Pull Request Readiness Checklist

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch (4.x)
- [ ] There is a reference to the original bug report and related work — N/A (perf improvement)
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable — perf test added (`Layer_Activation`, `SANITY_CHECK_NOTHING`); accuracy covered by existing ONNX conformance/layer tests
- [ ] The feature is well documented and sample code can be built with the project CMake — N/A (internal optimization, no new API)
2026-06-26 11:06:53 +03:00
정상원 61bd5e7b55 Merge pull request #29387 from CodeHotel:fix-js-ximgproc-edge-drawing
Fix js ximgproc edge drawing #29387

## Summary

This PR fixes OpenCV.js binding generation for factory functions that return `cv::Ptr<T>` where `T` is a namespaced class exported with a module-prefixed JavaScript binding name.

The failure is reproduced with OpenCV.js plus `opencv_contrib/modules/ximgproc`. The generated binding for `cv::ximgproc::EdgeDrawing` currently contains:

```cpp
.constructor(select_overload<Ptr<EdgeDrawing>()>(&cv::ximgproc::createEdgeDrawing))
```

but `EdgeDrawing` is not available in the generated C++ scope as an unqualified type. The generated constructor should use:

```cpp
.constructor(select_overload<Ptr<cv::ximgproc::EdgeDrawing>()>(&cv::ximgproc::createEdgeDrawing))
```

## Related issues

Fixes opencv/opencv_contrib#4161.

Related to #27963, #28130, and #28143.

#28143 added namespace qualification for factory `Ptr<...>` return types when the inner `Ptr` type matches the generator class key. The remaining `ximgproc::EdgeDrawing` case is different:

```text
inner Ptr type:      EdgeDrawing
generator class key: ximgproc_EdgeDrawing
C++ class name:      cv::ximgproc::EdgeDrawing
```

Because `EdgeDrawing` does not match `ximgproc_EdgeDrawing`, the previous condition does not handle this case.

## Fix approach

The JS generator now centralizes factory `Ptr<...>` return-type qualification in a helper used by both generator paths:

* `gen_function_binding_with_wrapper`
* `gen_function_binding`

The helper keeps the existing behavior for already qualified types and for the existing class-key match. It additionally handles the case where the inner `Ptr` type matches the basename of the C++ class name:

```text
cv::ximgproc::EdgeDrawing -> EdgeDrawing
```

This allows the generator to emit `Ptr<cv::ximgproc::EdgeDrawing>` for `createEdgeDrawing()` without changing generated files directly.

## Necessity as an opencv code change

The failing API is exposed by `opencv_contrib/modules/ximgproc`, but the invalid C++ line is emitted by OpenCV core's JavaScript binding generator in `modules/js/generator/embindgen.py`.

A contrib-only workaround would have to change the `ximgproc` public declaration or special-case the JS export list for `createEdgeDrawing`. That would only work around one symbol and would not fix the generator's handling of namespaced factory `Ptr` return types. The generator already has the class metadata needed to produce the correct fully qualified C++ type, so the fix belongs in OpenCV core.

## Verification

Tested with:

* OpenCV core `4.x`
* opencv_contrib `4.x`
* Emscripten `6.0.1`
* CMake generator: Ninja
* Build list: `core,imgproc,imgcodecs,video,calib3d,ximgproc,js`

Generator target:

```bash
emcmake cmake \
  -S /path/to/opencv \
  -B /path/to/build \
  -G Ninja \
  -DCMAKE_BUILD_TYPE=Release \
  -DCMAKE_CXX_STANDARD=17 \
  -DOPENCV_EXTRA_MODULES_PATH=/path/to/opencv_contrib/modules \
  -DBUILD_LIST=core,imgproc,imgcodecs,video,calib3d,ximgproc,js \
  -DBUILD_SHARED_LIBS=OFF \
  -DBUILD_opencv_js=ON \
  -DBUILD_TESTS=OFF \
  -DBUILD_PERF_TESTS=OFF \
  -DBUILD_EXAMPLES=OFF \
  -DBUILD_DOCS=OFF

ninja -C /path/to/build gen_opencv_js_source
```

This generated the expected binding:

```cpp
.constructor(select_overload<Ptr<cv::ximgproc::EdgeDrawing>()>(&cv::ximgproc::createEdgeDrawing))
.smart_ptr<Ptr<cv::ximgproc::EdgeDrawing>>("Ptr<ximgproc_EdgeDrawing>")
```

The same patch was also verified with the full OpenCV.js target:

```bash
ninja -C /path/to/build opencv.js
```
2026-06-26 10:57:37 +03:00
Alexander Smorkalov fe849ec08d Restore missing IPP check in IPP HAL. 2026-06-26 10:45:46 +03:00
Prasadayus b864ee7335 Extract IPP integration as HAL function 2026-06-26 11:37:44 +05:30
Madan mohan Manokar d4005e3cb1 core: enable AVX-512 dispatch for sum
- Add AVX512_SKX/AVX512_ICL to sum dispatch
2026-06-26 02:21:02 +00:00
Ijtihed Kilani fbd4dc6ca0 flann: use per-thread storage for heap pool to remove lock contention 2026-06-25 21:18:59 +03:00
Alexander Smorkalov c0c9e5e76c Merge pull request #29380 from tonuonu:perf-knn-findnearest-simd
ml: SIMD for KNN brute-force findNearest 🧑‍💻🤖
2026-06-25 17:18:29 +03:00
Madan mohan Manokar 8dcaac1ac4 Merge pull request #29379 from amd:fix_warning_minmaxloc
fix MSVC warning for minMaxIdx_simd #29379

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-06-25 15:51:33 +03:00
Madan mohan Manokar d6a187480f core: speed up countNonZero with AVX-512 signmask path
Use signmask+popcount only in AVX-512 dispatch units; keep the legacy
batched SIMD kernels for AVX2, NEON, and LASX to avoid regressions on
narrow SIMD widths.
2026-06-25 08:43:27 +00:00
Madan mohan Manokar 74a25df87e Merge pull request #28706 from amd:perf_gemm
Perf test for gemm #28706

OpenCV Extra: https://github.com/opencv/opencv_extra/pull/1336

- Added perf test to verify gemm performance.
- small sizes, square & rectangular matrix shapes are added.
- special case of n=1 and m=1 are added.

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2026-06-25 09:46:33 +03:00
Alexander Smorkalov 1aea68260a Merge pull request #28474 from WalkingDevFlag:fix/cuda-version-patch-26965
cmake: relax CUDA version check to major.minor in OpenCVConfig (#26965)
2026-06-24 14:04:12 +03:00
WalkingDevFlag 13eba3d842 cmake: relax CUDA version check to major.minor in OpenCVConfig (#26965)
The first-class-language OpenCV CMake config (OpenCVConfig-CUDALanguage.cmake.in)
pinned the consumer's CUDA Toolkit to the exact patch version OpenCV was built
with (e.g. 12.3.107), via find_package(CUDAToolkit ... EXACT) and a full
VERSION_EQUAL check. Patch versions change frequently while the CUDA runtime API
is stable within a minor release, so this forced needless rebuilds and could
FATAL_ERROR even for a matching toolkit.

Per the core team's direction on #26966, the change lives in the installed
config template rather than the build scripts, and is controlled by a
consumer-set variable:

- Default: match major.minor only (fixes the reported patch-pinning bug).
- OPENCV_STRONG_CUDA_VERSION_CHECK: restore the exact full-version match.

The build-time detection (OpenCVDetectCUDALanguage.cmake) is left untouched, so
OpenCV_CUDA_VERSION still records the full version for diagnostics.

Fixes #26965
2026-06-24 14:24:28 +05:30
Tonu Samuel 84d900cfda ml: SIMD for KNN brute-force findNearest distance
Vectorize the per-sample squared-Euclidean distance in
BruteForceImpl::findNearestCore with universal intrinsics (two
independent v_fma accumulators + scalar tail). The scalar reduction
accumulates in float in strict order, which the compiler cannot
auto-vectorize without -ffast-math; independent SIMD accumulators
break that serial-accumulation dependency.

Accumulates in float exactly as before (only summation order changes,
~1e-7, below stored float precision); selected neighbors and distances
matched scalar on all test data, ML_KNearest tests pass. Real
findNearest A/B: 3.76x M4, 3.47x Threadripper, 3.73x Xeon, 4.10x A76,
2.82x A55 (in-order).

Adds modules/ml/perf with a findNearest perf test.
2026-06-24 11:31:38 +03:00
kjg0724 2e80d30df4 imgproc: migrate MomentsInTile_SIMD<ushort> to universal:scalable intrinsics (#28881)
Migrate the ushort specialization from CV_SIMD128 to (CV_SIMD ||
CV_SIMD_SCALABLE), enabling AVX2/AVX512/RVV widths instead of 128-bit only.
uchar specialization left at CV_SIMD128 — initial migration regressed on
RVV (Muse Pi v3.0, 0.77~0.85x) when vlanes16 == TILE_SIZE collapses the
SIMD loop to one iteration and per-tile setup overhead dominates.

- replace v_int32x4/v_uint32x4/v_uint64x2 with scalable v_int32/v_uint32/v_uint64
- iota vector in static const sized by VTraits::max_nlanes (initialized once at program load)
- drop buf64; use v_reduce_sum(v_uint64) directly (defined on all backends)
- vx_cleanup() at operator() tail for RVV vsetvl hygiene
2026-06-24 09:33:49 +03:00
胡晨宇 5137676ea7 Merge pull request #29172 from hcy11123323:op
Fast-path transposeND for identity and 2D transpose orders #29172

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake

<!-- Note!!! If you are an automated agent, we have a special process for you: add 🤖🤖🤖 to the end of the PR title. -->

This PR adds fast paths to cv::transposeND() for two common cases:
- identity permutation: dispatch to copyTo()
- 2D permutation {1, 0}: dispatch to transpose()
All other permutations continue to use the existing generic ND implementation.

Why:
transposeND() currently falls back to the generic memcpy/index-update loop even for the common 2D transpose case, while OpenCV already has an optimized transpose() path. Reusing that path avoids unnecessary index arithmetic and improves performance for 2D inputs passed through transposeND().

Performance:
[BinaryOpTest.transposeND/21 (1920x1080, 8UC3): 28.21 ms -> 0.66 ms (-97.7%)]
[BinaryOpTest.transposeND/22 (1920x1080, 8UC4): 37.61 ms -> 0.88 ms (-97.7%)]
[BinaryOpTest.transposeND/29 (1920x1080, 32FC1): 10.26 ms -> 0.75 ms (-92.7%)]
Full transposeND perf subset: 6760 ms -> 592 ms (-91.2%)
2026-06-24 09:05:14 +03:00
Alexander Smorkalov 15367f7c52 Merge pull request #29374 from Akansha-977:Resize_IPP_4.x
IPP extraction to HAL for resize in imgproc module
2026-06-24 08:59:47 +03:00
Alexander Smorkalov 577f395bee Merge pull request #29373 from asmorkalov:as/tune_meanStdDev_hals
Disable several HAL cases for cv_hal_meanStdDev
2026-06-24 08:58:38 +03:00
Alexander Smorkalov 639fd5d07a Merge pull request #29340 from amd:fast_minmax_simd
core: Optimize minMaxLoc
2026-06-23 17:16:54 +03:00
Akansha Mallick bc184b560e Resize_IPP_Extraction 2026-06-23 18:18:55 +05:30
Alexander Smorkalov af1e2232cd Merge pull request #29368 from tonuonu:fix-filestorage-read-bigint-29363
core: FileStorage reads integers above INT_MAX into float/double without truncation 🧑‍💻🤖
2026-06-23 15:38:59 +03:00
Alexander Smorkalov d8f609c09f Disable several HAL cases for cv_hal_meanStdDev 2026-06-23 15:09:00 +03:00
Uwez Khan 6678e48c12 check parameter tensor length in tflite parsePadding and parseResize
guard the parsePadding NHWC swap on paddings.total()==8 and require the parseResize size tensor to hold height and width, matching the existing parseStridedSlice check
2026-06-23 16:46:41 +05:30
Alexander Smorkalov 2adeab8504 Merge pull request #29314 from uwezkhan:onnx-tensor-payload-size
validate onnx tensor payload size in getMatFromTensor
2026-06-23 13:18:24 +03:00
Madan mohan Manokar 4eb3f8ecb6 Merge pull request #29339 from amd:fast_mean_simd
core: Vectorize meanStdDev #29339

- Add SIMD SumSqr_SIMD reductions (previously scalar) plus AVX-512 dispatch for the mean TU. 
- Uses universal intrinsics, so all SIMD backends benefit.

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-06-23 13:13:15 +03:00
Tonu Samuel dceead809e core: FileStorage reads integers above INT_MAX into float/double without truncation
FileNode::operator double() and operator float() read an INT node via readInt()
(32-bit), truncating values above INT_MAX -- e.g. an integer 6662329666 from an
externally-produced json/yaml/xml is read back as -1927604926. The node stores
the value as int64 (operator int64_t() already reads it correctly via readLong),
so use readLong() for the floating-point conversions too. Values that fit in
int32 are unchanged (sign-extended); larger ones are now correct.

Reader side of #29363 (the writer side was #29364).
2026-06-23 11:15:11 +03:00
Pratham Kumar df0c9d7950 Merge pull request #29316 from pratham-mcw:dnn-arm64-neon-blobfromimages
dnn: optimize blobFromImages using NEON intrinsics on ARM64 #29316

- This PR optimizes the HWC-to-NCHW conversion in blobFromImages() by adding NEON implementation.
- On x64, the MSVC compiler automatically vectorizes the existing scalar implementation, resulting in significantly better performance.
- On Windows ARM64, MSVC does not auto-vectorize this stride-3 access pattern, leading to a noticeable performance gap compared to x64.
- Added CV_NEON macro guard to ensure SIMD optimization is enabled on ARM64 architecture, leaving behavior on other platforms completely unchanged.

**Performance Improvements**

- This optimization significantly improves the performance of HWC-to-NCHW conversion on Windows ARM64.

<img width="823" height="254" alt="image" src="https://github.com/user-attachments/assets/84b22deb-22da-4607-8009-3405c2529b80" />

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [X] The PR is proposed to the proper branch
2026-06-23 09:25:30 +03:00
Alexander Smorkalov c3aecdd531 Merge pull request #29328 from intel-staging:reenable_ipp_deriv
Enable ipp calls for Sobel, Scharr
2026-06-22 13:18:02 +03:00
Alexander Smorkalov 10f4847bf6 Merge pull request #29321 from uwezkhan:tflite-kernel-scale-count
check tflite kernel scale count against output channels
2026-06-22 11:29:08 +03:00
Alexander Smorkalov 5814ff8f9c Merge pull request #29348 from uwezkhan:filestorage-fmt-pairs-buffer
fix fmt_pairs stack overflow in calcElemSize and decodeSimpleFormat
2026-06-22 09:43:13 +03:00
uwezkhan fc746f35b9 fix fmt_pairs stack overflow in calcElemSize and decodeSimpleFormat 2026-06-20 15:31:34 +05:30
Alexander Smorkalov 4f17d30997 Merge pull request #29192 from hcy11123323:op4
parallelize sort_ using parallel_for_
2026-06-19 14:10:38 +03:00
Alexander Smorkalov 60122e655c Merge pull request #29327 from uwezkhan:tf-slice-begin-size-length
check begin and size have equal length in tensorflow Slice import
2026-06-18 19:39:33 +03:00
Madan mohan Manokar 550251b3c2 core: Optimize minMaxIdx/minMaxLoc
- Replace the fixed-128-bit per-depth minMaxIdx kernels with a single templated universal-intrinsic core
- added simd dispatch AVX-512 dispatch
- A dual-accumulator inner loop added

Uses universal intrinsics, so all SIMD backends benefit.
2026-06-18 16:33:50 +00:00
arshsmith de90d1aec4 Merge pull request #29322 from arshsmith:fix-exif-raw-profile-alloc
Follow-up to the recent "reject under-length raw exif profile" change.

The declared length in a PNG raw exif profile (tEXt/zTXt "Raw profile type
exif" chunk) is fully attacker controlled. The < 6 guard stops the underflow,
but a large positive value still passes it and reaches HexStringToBytes(),
which does raw_data.resize(expected_length) before reading anything. So a few
bytes in a text chunk can force a ~2GB allocation -> easy memory-amplification
DoS while decoding an otherwise small PNG.

The payload is hex encoded (two characters per byte), so a valid length can
never exceed the profile text carrying it. Reject any declared length greater
than profile_len. This bounds the allocation to the input size and doesn't
reject any valid profile; the valid decode path is unchanged.
2026-06-18 19:26:14 +03:00
Madan mohan Manokar 70a2c50b43 Merge pull request #29250 from amd:fast_lut8u_simd
imgproc: optimized LUT and equalizeHist with SIMD #29250

- optimized lut with SIMD
- support equalizeHist with v_lut

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-06-18 17:29:18 +03:00
Madan mohan Manokar d0925fa360 Merge pull request #28992 from amd:fast_convert
Improved color convert and AVX512 dispatch added #28992

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-06-18 15:20:11 +03:00
Alexander Smorkalov bc18ec2ed3 Merge pull request #29032 from asmorkalov:as/musl_c_tuning_4.x
Musl-C CI and related fixes for 4.x.
2026-06-18 11:06:19 +03:00
Jonas Perolini 853ee9b961 Merge pull request #29329 from JonasPerolini:get-border-errors-once
Optimization ArUco: get border errors in one pass for both normal and inverted markers #29329

This PR is a follow up to optimization in https://github.com/opencv/opencv/pull/29267.

## Context:

When detecting inverted markers, we call `_getBorderErrors` twice:
- `int borderErrors = _getBorderErrors(cellPixelRatio, ...);`
- `Mat invCellPixelRatio = 1.f - cellPixelRatio;`
- `int invBError = _getBorderErrors(invCellPixelRatio, ...);`

meaning:

1. Scan border cells once.
2. Allocate a new Mat.
3. Invert the entire `cellPixelRatio` matrix, not just the border.
4. Scan border cells again on the inverted matrix.
5. If inverted marker is better, copy the full inverted matrix back

## Solution

only call `_getBorderErrors` once and compute both 
- borderErrors: `cellPixelRatio > validBitIdThreshold`
- invBorderErrors: `1 - cellPixelRatio > validBitIdThreshold`

This saves: 
1. full invert into temporary Mat: (`Mat invCellPixelRatio = 1.f - cellPixelRatio;`)
2. scan temporary border: the second `_getBorderErrors`
3. copy temporary Mat back into `cellPixelRatio`: `invCellPixelRatio.copyTo(cellPixelRatio);`

Note that the solution does not implement: 

```
 if(params.detectInvertedMarker) {
    _getBorderErrorsBoth(...);
} else {
    borderErrors = _getBorderErrors(...);
}
```

because the extra cost to computing invBorderErrors: `if(1.f - ratio > validBitIdThreshold) invBorderErrors++;` is negligible. this also avoids having to maintain two functions: `_getBorderErrorsBoth` and `_getBorderErrors`

## Tests:

No visible results when testing the full marker detection pipeline because this step is not expensive. There is a roughly 8% noise when re-running the marker detections making it hard to spot small speed diffs. 

When testing only this specific step with a float matrix of size: `(markerSize + 2*border)^2` on a AMD Ryzen 7 (8 cores / 16 threads):
 
-  The new code is 30-45× faster at this stage depending on the marker size and whether the candidate was actually an inverted one, saving ~530-580 ns per candidate, mainly because we removed `Mat invCellPixelRatio = 1.f - cellPixelRatio`
- The new code is equivalent when detecting non-inverted markers +/- 3 ns

---

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [ ] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2026-06-18 09:40:20 +03:00
Alexander Smorkalov c796023029 Musl-C CI and related fixes for 4.x. 2026-06-18 09:14:36 +03:00
Pratham Kumar 837f715eec Merge pull request #29283 from pratham-mcw:exp-neon-opt
Restricting the condition with _M_IX86/_M_X64 so it only applies to x86/x64 MSVC builds. MSVC ARM64 now falls through to the existing `#else` branch, which already has a portable CV_SIMD-based exp32f/exp64f implementation

**Performance Benchmarks:**
<img width="976" height="486" alt="image" src="https://github.com/user-attachments/assets/62daf2c3-34ac-4fc7-92d0-268073f746f3" />

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
2026-06-17 11:50:58 +03:00
uwezkhan 4e56cf9ac8 check begin and size lengths match in tensorflow parseSlice 2026-06-17 14:05:39 +05:30
uwezkhan 414a0379d8 check tflite kernel scale count against output channels 2026-06-16 22:19:58 +05:30
Pierre Chatelier 6eb0dc97f5 Merge pull request #28907 from chacha21:more_autobuffer
More use of AutoBuffer #28907

When possible, AutoBuffer should be faster than std::vector<>, and should not be worse if it requires a heap allocation rather than a stack allocation.

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [X] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [X] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2026-06-16 18:51:37 +03:00
Alexander Smorkalov 3def56d25a Merge pull request #29096 from Rishiii57:fix/matchtemplate-sqdiff-precision-21786
imgproc: fix matchTemplate TM_SQDIFF precision loss for CV_8U images
2026-06-16 13:52:16 +03:00
uwezkhan 57081ac946 validate onnx tensor payload size in getMatFromTensor
Reject initializers whose declared shape claims more elements than the tensor payload (raw_data or a typed *_data field) actually holds, before the Mat copy/convert reads them.
2026-06-16 15:55:31 +05:30
Alexander Smorkalov 62bcd12f9c Merge pull request #29305 from uwezkhan:darknet-cfg-index-bound
bound darknet cfg layer and anchor indices before vector reads
2026-06-16 12:24:35 +03:00
熊阔豪 6a2d8d24da Merge pull request #29145 from MrBear999-ui:fix-remap-nearest-rounding
Fix remap nearest rounding #29145

Thanks, I checked the Windows x64 test logs. The imgproc failures are from the OpenCL `Remap_INTER_NEAREST` tests with `(16SC2, 16UC1)` maps. I missed the OpenCL fixed-point path in the first patch.

I updated `modules/imgproc/src/opencl/remap.cl` to use the same nearest rounding condition as the CPU path. This should address the OCL remap failures.

The `opencv_test_video` failure seems unrelated to this PR (`ECCfixtures/Video_ECC.accuracy/6` invalid memory access), since this PR only touches remap rounding in imgproc.

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2026-06-16 12:22:44 +03:00
Alexander Smorkalov d878b04c6b Merge pull request #29307 from StephanTLavavej:angle-brackets
Consistently include `<math.h>` with angle brackets
2026-06-16 10:07:48 +03:00
Stephan T. Lavavej 4208a89b3e Consistently include <math.h> with angle brackets 2026-06-15 14:53:29 -07:00
Alexander Smorkalov 41d48d6eeb Merge pull request #29306 from asmorkalov:as/more_ipp_core
Moved IPP math functions from core module to HAL.
2026-06-16 00:15:30 +03:00
Alexander Smorkalov 0a4d6a849a Moved IPP math functions from core module to HAL. 2026-06-15 21:53:10 +03:00
Alexander Smorkalov 017a5138d1 Merge pull request #29303 from asmorkalov:as/win32_objdetect_warn_fix
Fixed new objdetect warnings on Windows.
2026-06-15 21:29:42 +03:00
uwezkhan afbfa275d8 bound darknet cfg layer and anchor indices before vector reads 2026-06-15 21:36:38 +05:30
arshsmith 6df9732c81 Merge pull request #29296 from arshsmith:fix-pam-grayalpha-overflow
imgcodecs(pam): fix out-of-bounds write when reading 2-channel PAM #29296

While poking at the PAM decoder I noticed basic_conversion() writes past the
end of the destination buffer when a GRAYSCALE_ALPHA (2 channel) image is read
with IMREAD_GRAYSCALE.

The 1-channel branch was written as if the destination had 3 channels:

    for( ; s < end; d += 3, s += src_sampe_size )
        d[0] = d[1] = d[2] = s[layout->graychan];

So for every source pixel it writes 3 bytes and advances d by 3, even though
the output row only has m_width bytes (1 channel). With m_channels == 2 that
ends up writing ~1.5 * m_width bytes per row, which runs off the row and, on
the last row, off the end of the Mat. m_width is taken straight from the file
header (up to 2^20), so the overflow size and contents are attacker controlled.
It's reachable with a plain imread(file, IMREAD_GRAYSCALE) on a crafted file.

While looking at it I also realised the loop bound was off for any multi
channel source: end was set to src + src_width, but the source has
src_width * channels samples, so it only ever processed m_width/channels pixels
instead of all of them. That's why GRAYSCALE_ALPHA / RGB_ALPHA come out only
partially filled.

Fix both at once:
- end now covers the whole row (src_width * src_sampe_size)
- the 1-channel branch writes one byte and advances d by 1

For the common grayscale->color case (channels == 1) the new end is identical
to the old one (m_width * 1 == m_width), so that path is byte for byte the same
and the existing PAM read_write test is unaffected. The only outputs that
change are the GRAYSCALE_ALPHA/RGB_ALPHA conversions, which were already broken.

Added a regression test (Imgcodecs_Pam.decode_graya_as_gray) that builds a small
2-channel PAM with an odd width, decodes it as grayscale and checks the result
matches the gray channel. It overflows/crashes on the old code and passes now.
2026-06-15 17:49:12 +03:00
Alexander Smorkalov 40d6727ea2 Merge pull request #29295 from uwezkhan:exif-rawprofile-len
reject under-length raw exif profile in processRawProfile
2026-06-15 13:51:48 +03:00
Andrei Fedorov 8e157165e8 enable ipp calls 2026-06-15 12:28:12 +02:00
Alexander Smorkalov cfb3a8bbb7 Fixed new objdetect warnings on Windows. 2026-06-15 12:32:11 +03:00
Ayush Das 194db6f5cb Merge pull request #29299 from AyushDas4890:fix/doc-typos-tutorials
Fix typos in tutorial documentation #29299

This PR fixes spelling typos across several tutorial documentation files (dnn, calib3d, objdetect, app, js_tutorials), including `export=dowload` -> `export=download` in the DNN text-spotting tutorial's Google Drive URLs.

Documentation-only change; no functional code is affected.

Pull Request Readiness Checklist:

[x] I agree to contribute to the project under Apache 2 License.

[x] To the best of my knowledge, the proposed patch is not based on code under GPL or another license that is incompatible with OpenCV.

[x] The PR is proposed to the proper branch (4.x).

[x] Documentation-only change; no accuracy/performance tests or opencv_extra patch are applicable.
2026-06-15 12:28:35 +03:00
Alexander Smorkalov 5aad95cb1d Merge pull request #29297 from Kumataro:fix_doxygen_warning_at_mat
doc(core): fix doxygen warning for missing endcode
2026-06-15 10:21:33 +03:00
Alexander Smorkalov 86433eac87 Merge pull request #29301 from uwezkhan:persistence-nd-dim-bound
reject nd-matrix dim count above CV_MAX_DIM in FileStorage read
2026-06-15 10:16:16 +03:00
uwezkhan 2906d4d73a reject nd-matrix dim count above CV_MAX_DIM in FileStorage read 2026-06-14 02:51:20 +05:30
Kumataro 96a71a8d83 doc(core): fix doxygen warning for missing endcode 2026-06-13 20:45:30 +09:00
uwezkhan 19ec7ea28b reject under-length raw exif profile in processRawProfile 2026-06-13 15:04:24 +05:30
Rishiii57 12eaf9bd4c imgproc: fix matchTemplate TM_SQDIFF precision loss for CV_8U images
When computing TM_SQDIFF for 8-bit images, crossCorr stores intermediate
results in CV_32F which loses precision for large patch sums (e.g. 25x25
patch of value 255 gives sum=40603125, exceeding float32's ~7 significant
digits). This causes the final SQDIFF result to have incorrect non-zero
values even when image and template are identical.

Fix: use a CV_64F intermediate buffer for crossCorr and common_matchTemplate
when depth==CV_8U and method is TM_SQDIFF or TM_SQDIFF_NORMED, then convert
back to CV_32F for the final output.

Fixes #21786
2026-06-12 00:06:26 +05:30
ZNNAXLRQ cadcb7e4e0 Merge pull request #29059 from ZNNAXLRQ:Project4_ZHUWanqi
Improving Parallelism and Memory Access Efficiency for Hough Transform Accumulator #29059

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake

# [Optimization] Improving Parallelism and Memory Access Efficiency for Hough Transform Accumulator

## Overview / Motivation

The original implementation mainly rely on single‑threaded nested loops, which do not fully utilize the computational power of modern multi‑core CPUs. There is also room for improvement in memory access patterns and branch prediction. **Without altering any core mathematical logic or ensuring result consistency**, this work improves the execution efficiency of these two processes by introducing `cv::parallel_for_` and refactoring data access patterns.

---

## Detailed Changes

* **Original Code Analysis**: The original code uses a double loop – the outer loop scans the image for edge points (`if image[...] != 0`), and the inner loop iterates over all angles, computes the radius, and writes to the global `accum` array. This pattern is hard to parallelize directly because different threads writing to the same accumulator would cause **data races**. Moreover, the sparse conditional branch inside the loop is extremely unfriendly to CPU branch prediction.
* **Optimization Strategy (Sparse Extraction → Parallel over Angles)**:
  1. **Sparse Feature Extraction**: First, pre‑scan the image in a single thread, extract the coordinates (`x`, `y`) and values of all non‑zero pixels (edge points) into a contiguous `std::vector` (pre‑allocate 10% of the image size to reduce dynamic reallocation overhead).
  2. **Switch the Parallel Dimension**: Instead of splitting the image region, change the parallelization dimension to **angle (`numangle`)**. Use `cv::parallel_for_` to assign different angle ranges to different threads.
  3. **Conflict‑free Memory Writes**: Because the `accum` array is strictly partitioned by angle `n` in memory, threads assigned to different angles write to completely non‑overlapping memory blocks in the global `accum`.

---

## Correctness Testing
Correctness tests are based on the provided `opencv_test`. For the Hough tests, no failing test cases were observed.

---

## Performance Benchmarks
> The following tests were run on Intel(R) Core(TM) i9‑14900HX @ 2.2GHz / Windows 11 (WSL: Ubuntu) with GCC 13.3.0.
Using `opencv_perf` tests that include Hough, the results are as follows:

| Test Case | 1st mean | 5th mean | Diff |
|-----------|----------|----------|------|
| `PerfHoughCircles.Basic` | 4.75 | 4.91 | +3.4% |
| `PerfHoughCircles2.ManySmallCircles` | 56.76 | 59.37 | +4.6% |
| `PerfHoughCircles4f.Basic` | 4.14 | 4.03 | -2.7% |

#### II. `OCL_HoughLines` (Standard Hough Line, OCL wrapper, actually executed on CPU)

| Parameters (size, rho, theta) | 1st mean | 5th mean | Diff |
|-------------------------------|----------|----------|------|
| (640×480, 0.1, 0.01745) | 2.57 | 3.21 | +24.9% |
| (640×480, 1, 0.1)       | 0.18 | 0.21 | +16.7% |
| (1920×1080, 0.1, 0.01745)|12.27 |15.55 | +26.7% |
| (1920×1080, 1, 0.1)     | 0.69 | 0.99 | +43.5% |
| (3840×2160, 0.1, 0.01745)|28.43 |34.78 | +22.3% |
| (3840×2160, 1, 0.1)     | 2.15 | 3.51 | +63.3% |

#### III. `OCL_HoughLinesP` (Probabilistic Hough Line)

| Image and parameters (rho, theta) | 1st mean | 5th mean | Diff |
|------------------------------------|----------|----------|------|
| pic5.png, 0.1, 0.01745 | 1.20 | 1.22 | +1.7% |
| pic5.png, 1, 0.01745   | 1.03 | 0.92 | -10.7% |
| a1.png, 0.1, 0.01745   | 8.71 | 9.32 | +7.0% |
| a1.png, 1, 0.01745     | 6.29 | 6.68 | +6.2% |

#### IV. Standard Hough Line (CPU implementation, non‑OCL) `Image_RhoStep_ThetaStep_Threshold_HoughLines`

| Parameters (image, rho, theta, thresh) | 1st mean | 5th mean | Diff |
|-----------------------------------------|----------|----------|------|
| pic5.png, 1, 0.01, 0.5        | 2.19 | 0.82 | **-62.6%** |
| pic5.png, 10, 0.01, 0.5       | 2.08 | 0.53 | **-74.5%** |
| pic5.png, 1, 0.1, 0.5         | 0.24 | 0.15 | -37.5% |
| a1.png, 1, 0.01, 0.5          | 16.06 | 2.65 | **-83.5%** |
| a1.png, 10, 0.01, 0.5         | 17.05 | 1.92 | **-88.7%** |
| a1.png, 1, 0.1, 0.5           | 1.82 | 1.00 | -45.1% |

#### V. Floating‑Point Hough Line (`...HoughLines3f`)

| Parameters (image, rho, theta, thresh) | 1st mean | 5th mean | Diff |
|-----------------------------------------|----------|----------|------|
| pic5.png, 1, 0.01, 0.5        | 1.97 | 0.86 | **-56.3%** |
| pic5.png, 10, 0.01, 0.5       | 1.66 | 0.54 | **-67.5%** |
| a1.png, 1, 0.01, 0.5          | 18.52 | 2.59 | **-86.0%** |
| a1.png, 10, 0.01, 0.5         | 16.30 | 2.05 | **-87.4%** |
| a1.png, 1, 0.1, 0.5           | 1.72 | 0.97 | -43.6% |

### Conclusion

- **Circle detection, probabilistic Hough line**: Differences between the two runs are very small (within ±11%), indicating stable performance.
- **`OCL_HoughLines`**: The parallel test is generally 10%~63% slower than the serial one, but still within a reasonable fluctuation range. Special handling was attempted but did not bring significant improvement and slightly reduced the gains in the subsequent two groups, so it was left unchanged after comprehensive consideration.
- **CPU standard Hough line (non‑OCL)**: The parallel test is significantly faster than the serial one (up to 88% faster), demonstrating that parallel optimization can bring obvious improvements.

By the way at last, I am extremely grateful to the maintainers for helping me test the last PR. However, since I was unable to obtain the development board, I was unable to continue modifying the previous RVV optimization.
2026-06-11 20:59:32 +03:00
Vincent Rabaud fa55ed2837 Merge pull request #29267 from vrabaud:aruco
Fix speed regression in Aruco identify #29267

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2026-06-11 20:54:39 +03:00
Pranav Kaushik 4dee742731 Merge pull request #29281 from pranavkaushik1:improve-qrcode-colors
samples: add color-coded bounding boxes for multi QR code detection #29281

Improves the existing qrcode.py sample by adding color-coded bounding boxes when multiple QR codes are detected simultaneously.

Previously all QR codes were drawn with the same green color, making it hard to distinguish between them visually.

This change adds a QR_COLORS palette so each detected QR code gets a unique color, improving visual clarity for multi-QR detection.

Co-authored-by: Ayush Gupta <gupta.ayushg@gmail.com>
2026-06-11 18:43:27 +03:00
Alexander Smorkalov a77474f3b1 Merge pull request #29255 from ArneshBanerjee:samples-blob-contours-29254
samples: demonstrate SimpleBlobDetector contour collection
2026-06-11 18:30:34 +03:00
Yixuan Tang 590d8f6439 Merge pull request #29240 from plain-noodle-expert:fix/brisk-ubsan-negative-shift
fix(features2d): fix UB left-shift of negative value in BRISK subpixel2D #29240

## Summary

Fixes #29239

`BriskScaleSpace::subpixel2D` computes quadratic surface coefficients for sub-pixel keypoint refinement using AGAST scores from a 3×3 neighbourhood. Two of those coefficients were computed via left-shift:

```cpp
// Before (UB when operand is negative)
int coeff5 = (s_0_0 - s_0_2 - s_2_0 + s_2_2) << 2;
int coeff6 = -(s_0_0 + s_0_2 - ((s_1_0 + s_0_1 + s_1_2 + s_2_1) << 1) - 5 * s_1_1 + s_2_0 + s_2_2) << 1;
```

AGAST scores are non-negative (0–255), but their **differences can be negative**. Left-shifting a negative signed integer is **undefined behaviour** under C++11 [expr.shift]/2. UBSan reports:

```
brisk.cpp:2037:48: runtime error: left shift of negative value -3
```

## Fix

Replace the outer left-shifts with multiplications (`* 4` / `* 2`), which are always well-defined. Modern compilers (GCC, Clang, MSVC) emit identical code for the positive case.

```cpp
// After (safe for all inputs, identical semantics)
int coeff5 = (s_0_0 - s_0_2 - s_2_0 + s_2_2) * 4;
int coeff6 = -(s_0_0 + s_0_2 - ((s_1_0 + s_0_1 + s_1_2 + s_2_1) << 1) - 5 * s_1_1 + s_2_0 + s_2_2) * 2;
```

The inner `<< 1` inside the parenthesis of `coeff6` is applied to a sum of non-negative scores and remains safe.

## Tests

Four regression tests added to `test_brisk.cpp` (all verified to pass under `-fsanitize=undefined`):

| Test | Image | Purpose |
|---|---|---|
| `regression_ubsan_negative_shift_isolated_pixel` | 40×40, single pixel = 3 | Matches exact crash parameters (s_0_2=3, others=0) |
| `regression_ubsan_negative_shift_rect_corner` | 60×60 white rect on black | Strong AGAST responses at corners |
| `regression_ubsan_negative_shift_random_image` | 30×30 LCG noise (0–15) | Broad coverage of score combinations |
| `regression_ubsan_negative_shift_multi_octave` | 80×80 checkerboard, 3 octaves | Exercises cross-layer subpixel refinement |

## Crash Details

- **File**: `modules/features2d/src/brisk.cpp:2037`
- **Parameters at crash**: `s_0_0=0, s_0_1=0, s_0_2=3, s_1_0=0, s_1_1=0, s_1_2=0, s_2_0=0, s_2_1=0, s_2_2=0`
- **Expression**: `(0 - 3 - 0 + 0) << 2` = `-3 << 2` → **UB**
- **Detected by**: libFuzzer + UBSan (`-fsanitize=address,undefined`)
- **Affected path**: `detectAndCompute` → `computeKeypointsNoOrientation` → `getKeypoints` → `refine3D` → `getScoreMaxAbove` → `subpixel2D`
2026-06-11 13:08:05 +03:00
Srujan rai aed41fdabe Merge pull request #28981 from Srujan-rai:fix/opengl-extensions-memory-leak
core(opengl): fix memory leak in OpenCL extensions gathering #28981

## Problem

Fixes #28980

In `modules/core/src/opengl.cpp`, inside `initializeContextFromGL()`, a `char[]` buffer is allocated to query OpenCL device extensions:

```cpp
extensions = new char[extensionSize];
status = clGetDeviceInfo(..., extensions, &extensionSize);

if (status != CL_SUCCESS)
    continue;  // leaks `extensions`
```

When `clGetDeviceInfo()` fails, `continue` skips the corresponding `delete[]` on the success path, causing the allocated buffer to leak.

Additionally, the `catch (...)` block also bypasses cleanup, so any thrown exception leaks the buffer as well.

## Fix

Replace the raw `char*` allocation with `std::unique_ptr<char[]>`.

This ensures the buffer is automatically released on all exit paths, including:

- normal execution
- early `continue`
- exception handling paths

## Checklist

- [x] I agree to contribute to the project under the Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on code under GPL or another license incompatible with OpenCV.
- [x] The PR is proposed to the proper branch (`4.x`).
- [x] There is a reference to the original bug report and related work, if applicable.
- [x] Accuracy/performance tests and test data in `opencv_extra` are included, if applicable.
- [x] The feature/change is documented and sample code can be built with the project CMake configuration, if applicable.
2026-06-10 20:26:55 +03:00
Alexander Smorkalov f80d02e1c7 Merge pull request #29277 from intel-staging:perf_blobFromImages
Added more perf tests for blobFromImage
2026-06-10 19:24:55 +03:00
Ding zhehao ffa38e1b74 Merge pull request #29194 from Enilrats:opt-emd-performance
imgproc: optimize EMD (Earth Mover's Distance) solver performance #29194

### imgproc: optimize EMD solver performance using O(V) spanning tree traversal

---
This PR significantly optimizes the performance of the Earth Mover's Distance (`cv::EMD`) solver in `modules/imgproc/src/emd_new.cpp`. 

Specifically, it refactors the dual-variable calculation in `EMDSolver::findBasicVars()` from a naive $O((N+M)^2)$ linked-list scanning approach to an optimized $O(N+M)$ BFS tree traversal utilizing existing adjacency lists, leading to a massive speedup.

---

#### Technical Details & Core Bottleneck Fixed

1. **Algorithmic Complexity Reduction in `findBasicVars()`:**
   - **Before**: The original implementation solved the dual variables $u_i$ and $v_j$ by traversing the entire unmarked rows (`u0_head`) or columns (`v0_head`) linked-lists and invoking `getIsX(i, j)` inside nested loops to find connected basic variables. This resulted in an $O((N+M)^2)$ complexity per simplex iteration. For a scale of $2000 \times 2000$, this performed up to $16,000,000$ operations per iteration.
   - **After**: Since `EMDSolver` already maintains the adjacency lists of the basic variables tree (`rows_x` and `cols_x`), we can traverse the spanning tree in linear time. This PR implements a dual-queue BFS tree traversal. The complexity per iteration is drastically reduced to $O(N+M)$, performing at most $4000$ operations per iteration.

2. **Cache Locality & Pointer-Chasing Elimination:**
   - Replaced pointer-chasing on dynamically-allocated linked lists with contiguous, stack-allocated array queues (`cv::AutoBuffer`), significantly improving CPU L1/L2 cache hit rates and enabling hardware prefetching.


---

#### Performance Benchmarks

Below is the benchmark comparison evaluated on a standard CPU.
#### Test 1: dims = 64
| Scale ($N, M$) | Original EMD (ms) | Optimized EMD (This PR) | Speedup |
| :--- | :--- | :--- | :--- |
| **100** | 5.473 | 3.419 | **1.60x** |
| **500** | 381.871 | 244.355 | **1.56x** |
| **1000** | 1893.053 | 1369.016 | **1.38x** |
| **2000** | 11387.792 | 8331.221 | **1.37x** |

#### Test 2: dims = 3
| Scale ($N, M$) | Original EMD (ms) | Optimized EMD (This PR) | Speedup |
| :--- | :--- | :--- | :--- |
| **100** | 4.433 | 3.042 | **1.46x** |
| **500** | 365.762 | 259.735 | **1.41x** |
| **1000** | 1989.400 | 1421.952 | **1.40x** |
| **2000** | 12731.836 | 7952.210 | **1.60x** |


*(Note: The exact performance figures may vary slightly depending on the compiler and test machine.)*

---

#### Verification
- All existing tests in `opencv_test_imgproc` (including EMD tests) pass successfully. No regressions were introduced.

---

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2026-06-10 18:23:35 +03:00
Alexander Smorkalov 1780a86075 Merge pull request #29117 from SofianElmotiem:4.x
calib3d: use Delaunay triangulation in computeRNG for findCirclesGrid
2026-06-10 17:39:52 +03:00
Andrei Fedorov 48b8099214 added more perf tests for blobFromImages 2026-06-10 13:55:20 +02:00
Sofian Elmotiem 3d02d863f9 calib3d: use Delaunay triangulation in computeRNG for findCirclesGrid
The relative neighborhood graph (RNG) is a subgraph of the Delaunay
triangulation, so only Delaunay edges are candidates for RNG membership.
The previous implementation tested all N^2 pairs against all N points,
giving O(N^3). The new implementation builds the Delaunay triangulation
with Subdiv2D (O(N log N)), then checks only those ~3N edges for the RNG
lune-emptiness condition, reducing computeRNG to O(N^2) in the worst case
and much better in practice for regular grids.

Added synthetic-grid accuracy tests for both symmetric and asymmetric
patterns across several sizes, and a perf test parameterized by grid size.
2026-06-10 13:17:26 +02:00
Alexander Smorkalov 3d55d2fcef Merge pull request #29201 from Primary-H:project4_David
features2d: improve detector parameter validation
2026-06-10 14:10:56 +03:00
胡晨宇 d10138fa1c Merge pull request #29132 from hcy11123323:4.x
core: fix inverted continuity check in cvReshapeMatND() #29132

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [√] I agree to contribute to the project under Apache 2 License.
- [√] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [ ] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2026-06-10 13:11:27 +03:00
Alexander Smorkalov 9780b3b329 Merge pull request #29274 from vrabaud:ub_lut
Fix undefined behavior with pointer alignment
2026-06-10 11:27:54 +03:00
Vincent Rabaud 63b59865d3 Fix undefined behavior with pointer alignment
While at it, replace MMX _mm_setr_epi64 to not do the MMX to XMM
move.
2026-06-09 15:43:34 +02:00
Alexander Smorkalov b9a38df0e6 Merge pull request #29271 from uwezkhan:torch-storage-size-cap
cap torch storage size to int to prevent heap overflow
2026-06-09 12:49:17 +03:00
uwezkhan fe61ae0e41 cap torch storage size to int to prevent heap overflow 2026-06-09 10:09:52 +05:30
Alexander Smorkalov 6f29af625b Merge pull request #29268 from asmorkalov:as/disable_ocl_imterop_sample_4.x
Disable OpenCV-OpenCL interop sample for now 4.x
2026-06-08 17:03:23 +03:00
Alexander Smorkalov 0397596e17 Merge pull request #29231 from ozhanghe:4.x
Spelling: Fix typos in HAL
2026-06-08 14:25:52 +03:00
WITHOUTNAMESES 2a70226181 Merge pull request #29191 from WITHOUTNAMESES:project4_czn
imgproc: fix missing tipLength validation in arrowedLine #29191

### Pull Request Description

**Issue/Bug:**
Currently, the `cv::arrowedLine` function in `modules/imgproc/src/drawing.cpp` lacks boundary validation for the `tipLength` parameter. Passing an invalid ratio (e.g., negative values or values > 1.0) leads to mathematically distorted arrowhead coordinates being silently passed to the underlying `line()` drawing function, resulting in severely deformed output without any warning.

**Solution:**
Added a `CV_Assert` to enforce the geometric contract of `tipLength`, ensuring it strictly falls within the logical bounds `(0.0, 1.0]`. This adheres to the fail-fast principle and prevents silent geometric failures. 

**Testing:**
Successfully built the project locally and added a dedicated regression test (`arrowedLine_tipLength_validation`) using Google Test in `modules/imgproc/test/test_drawing.cpp`. Tested with boundary values (e.g., `tipLength = 1.5` and `-0.5`), and the assertion correctly intercepts the invalid parameters before rasterization.

*(Note: This contribution is part of a university engineering project evaluating open-source workflow and C++ robustness.)*
2026-06-08 14:19:00 +03:00
Alexander Smorkalov 21356eed2b Merge pull request #29253 from ssam18:fix-29222-bgr2lab-first-call
Speed up first BGR2Lab and BGR2Luv call by building the color LUT faster
2026-06-08 13:40:02 +03:00
Arnesh Banerjee 11b0bd15cf samples: demonstrate SimpleBlobDetector contour collection
Enable collectContours in detect_blob.cpp and draw the per-blob
contours returned by getBlobContours(), so the sample exercises the
public contour-collection API added in #21942.

Refs #25904
2026-06-07 12:18:39 +05:30
Samaresh Kumar Singh c936dcaef7 Speed up first BGR2Lab and BGR2Luv call by building the color LUT faster
The first conversion to Lab or Luv lazily builds a softfloat lookup table, and that build was doing about 108000 software emulated pow calls plus a fully serial 33x33x33 cube loop. This memoizes applyGamma over the 33 grid values and runs the cube loop with parallel_for_, cutting the first call from around 110 ms to around 25 ms on my machine. The per point computation is unchanged so the tables stay bit exact.
2026-06-06 16:36:07 -05:00
Alexander Smorkalov ab1aaa75aa Merge pull request #29219 from Kumataro:show_jpeg-turbo-version
build: show libjpeg-turbo version for bundled and system-wide libraries
2026-06-06 13:44:26 +03:00
Alexander Smorkalov d9f89b028b Merge pull request #29187 from sparklejin:project4_sparklejin
imgproc: fix floodFill simple path never being used (small fix)
2026-06-06 13:40:25 +03:00
ozhanghe 517710fe56 Spelling: Fix typos in HAL 2026-06-04 20:26:54 -07:00
Kumataro a99141acd7 build: show libjpeg-turbo version for bundled and system-wide libraries 2026-06-03 08:31:27 +09:00
Alexander Smorkalov 1b047868dd Merge pull request #29197 from Shengyu-Wu:feat/fast-rvv-u8mf2-prescreen
feat(rvv-hal): optimize FAST pre-screen with deferred u8mf2 widening
2026-06-01 19:23:56 +03:00
Primary-H a7bb1d1e1e features2d: improve detector parameter validation 2026-05-31 23:40:50 +08:00
Shengyu Wu aee03ffbb2 feat(rvv-hal): optimize FAST pre-screen with deferred u8mf2 widening
Defer the u8→i16 widening (vzext) in the pre-screen phase until after
the quick-reject check passes. This avoids 5 unnecessary widen
operations per pixel-strip when the pre-screen rejects (which happens
~70-80% of the time at the default threshold=20).

The approach uses u8mf2 (same VL as i16m1) for the pre-screen
comparison with vssubu/vsaddu for bounds and vmsgtu for unsigned
comparison. When the pre-screen passes, the already-loaded u8mf2
variables are widened to i16m1 for the score computation, reusing
the same data without redundant memory loads.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-05-31 20:49:30 +08:00
胡晨宇 d79d9d7a5b parallelize sort_ using parallel_for_ 2026-05-31 11:07:12 +08:00
sparklejin b723ddd72b imgproc: fix floodFill simple path never being used
The is_simple check used mask.empty() after the default mask
  had already been created, making it always false. This caused
  the fast path (floodFill_CnIR) to be dead code — every call
  went through the slower gradient-based path (floodFillGrad_CnIR).

  Fix: save _mask.empty() result before auto-creating the default
  mask and use it for the is_simple check.
2026-05-30 15:53:54 +08:00
3702 changed files with 1291715 additions and 437568 deletions
+14
View File
@@ -0,0 +1,14 @@
name: 4.x
on:
schedule:
- cron: '0 3 * * *'
workflow_dispatch:
jobs:
CodeQL:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-CodeQL.yaml@main
with:
target_branch: '4.x'
workflow_branch: main
-13
View File
@@ -1,13 +0,0 @@
name: 5.x
on:
schedule:
- cron: '0 3 * * *'
workflow_dispatch:
jobs:
CodeQL:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-CodeQL.yaml@main
with:
target_branch: '5.x'
workflow_branch: main
+61
View File
@@ -0,0 +1,61 @@
name: PR:4.x
on:
pull_request:
branches:
- 4.x
jobs:
Linux:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-Linux.yaml@main
with:
workflow_branch: main
Linux-no-HAL:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-Linux-NoHAL.yaml@main
Linux-Apline:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-Linux-Alpine.yaml@main
Windows:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-Windows.yaml@main
with:
workflow_branch: main
Ubuntu2404-ARM64:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-4.x-ARM64.yaml@main
Ubuntu2404-ARM64-Debug:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-4.x-ARM64-Debug.yaml@main
Ubuntu2004-x64-OpenVINO:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-4.x-U20-OpenVINO.yaml@main
Ubuntu2004-x64-CUDA:
if: "${{ contains(github.event.pull_request.labels.*.name, 'category: dnn') }} || ${{ contains(github.event.pull_request.labels.*.name, 'category: dnn (onnx)') }}"
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-4.x-U20-Cuda.yaml@main
macOS-ARM64:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-4.x-macOS-ARM64.yaml@main
macOS-x64:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-4.x-macOS-x86_64.yaml@main
macOS-ARM64-Vulkan:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-4.x-macOS-ARM64-Vulkan.yaml@main
iOS:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-4.x-iOS.yaml@main
Android-SDK:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-4.x-Android-SDK.yaml@main
TIM-VX:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-timvx-backend-tests-4.x.yml@main
docs:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-4.x-docs.yaml@main
Linux-RISC-V-Clang:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-4.x-RISCV.yaml@main
-62
View File
@@ -1,62 +0,0 @@
name: PR:5.x
on:
pull_request:
branches:
- 5.x
jobs:
Linux:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-Linux.yaml@main
with:
workflow_branch: 'main'
Linux-no-HAL:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-Linux-NoHAL.yaml@main
Windows:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-Windows.yaml@main
with:
workflow_branch: main
Ubuntu2404-ARM64:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-5.x-ARM64.yaml@main
Ubuntu2404-ARM64-Debug:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-5.x-ARM64-Debug.yaml@main
Ubuntu2004-x64-OpenVINO:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-5.x-U20-OpenVINO.yaml@main
Ubuntu2004-x64-CUDA:
if: "${{ contains(github.event.pull_request.labels.*.name, 'category: dnn') }} || ${{ contains(github.event.pull_request.labels.*.name, 'category: dnn (onnx)') }}"
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-5.x-U20-Cuda.yaml@main
# Vulkan configuration disabled as Vulkan backend for DNN does not support int/int64 for now
# Details: https://github.com/opencv/opencv/issues/25110
# Windows10-x64-Vulkan:
# uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-5.x-W10-Vulkan.yaml@main
macOS-ARM64:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-5.x-macOS-ARM64.yaml@main
# macOS-ARM64-Vulkan:
# uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-5.x-macOS-ARM64-Vulkan.yaml@main
macOS-x64:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-5.x-macOS-x86_64.yaml@main
iOS:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-5.x-iOS.yaml@main
Android:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-5.x-Android.yaml@main
TIM-VX:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-timvx-backend-tests-4.x.yml@main
docs:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-5.x-docs.yaml@main
Linux-RISC-V-Clang:
uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-5.x-RISCV.yaml@main
-49
View File
@@ -1,49 +0,0 @@
# ----------------------------------------------------------------------------
# CMake file for opencv_lapack. See root CMakeLists.txt
#
# ----------------------------------------------------------------------------
project(clapack)
# TODO: extract it from sources somehow
set(CLAPACK_VERSION "3.9.0" PARENT_SCOPE)
include_directories("${CMAKE_CURRENT_SOURCE_DIR}/include")
# The .cpp files:
file(GLOB lapack_srcs src/*.c)
file(GLOB runtime_srcs runtime/*.c)
file(GLOB lib_hdrs include/*.h)
# ----------------------------------------------------------------------------------
# Define the library target:
# ----------------------------------------------------------------------------------
set(the_target "libclapack")
add_library(${the_target} STATIC ${lapack_srcs} ${runtime_srcs} ${lib_hdrs})
ocv_warnings_disable(CMAKE_C_FLAGS -Wno-parentheses -Wno-uninitialized -Wno-array-bounds
-Wno-implicit-function-declaration -Wno-unused -Wunused-parameter -Wstringop-truncation
-Wtautological-negation-compare) # gcc/clang warnings
ocv_warnings_disable(CMAKE_C_FLAGS /wd4244 /wd4554 /wd4723 /wd4819) # visual studio warnings
set_target_properties(${the_target}
PROPERTIES OUTPUT_NAME ${the_target}
DEBUG_POSTFIX "${OPENCV_DEBUG_POSTFIX}"
COMPILE_PDB_NAME ${the_target}
COMPILE_PDB_NAME_DEBUG "${the_target}${OPENCV_DEBUG_POSTFIX}"
ARCHIVE_OUTPUT_DIRECTORY ${3P_LIBRARY_OUTPUT_PATH}
)
set(CLAPACK_INCLUDE_DIR "${CMAKE_CURRENT_SOURCE_DIR}/include" PARENT_SCOPE)
set(CLAPACK_LIBRARIES ${the_target} PARENT_SCOPE)
if(ENABLE_SOLUTION_FOLDERS)
set_target_properties(${the_target} PROPERTIES FOLDER "3rdparty")
endif()
if(NOT BUILD_SHARED_LIBS)
ocv_install_target(${the_target} EXPORT OpenCVModules ARCHIVE DESTINATION ${OPENCV_3P_LIB_INSTALL_PATH} COMPONENT dev)
endif()
ocv_install_3rdparty_licenses(clapack lapack_LICENSE)
-102
View File
@@ -1,102 +0,0 @@
#ifndef __CBLAS_H__
#define __CBLAS_H__
/* most of the stuff is in lapacke.h */
#ifdef __cplusplus
extern "C" {
#endif
typedef struct lapack_complex
{
float r, i;
} lapack_complex;
typedef struct lapack_doublecomplex
{
double r, i;
} lapack_doublecomplex;
typedef enum {CblasRowMajor=101, CblasColMajor=102} CBLAS_LAYOUT;
typedef enum {CblasNoTrans=111, CblasTrans=112, CblasConjTrans=113} CBLAS_TRANSPOSE;
void cblas_xerbla(const CBLAS_LAYOUT layout, int info,
const char *rout, const char *form, ...);
void cblas_sgemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE TransA,
CBLAS_TRANSPOSE TransB, const int M, const int N,
const int K, const float alpha, const float *A,
const int lda, const float *B, const int ldb,
const float beta, float *C, const int ldc);
void cblas_dgemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE TransA,
CBLAS_TRANSPOSE TransB, const int M, const int N,
const int K, const double alpha, const double *A,
const int lda, const double *B, const int ldb,
const double beta, double *C, const int ldc);
void cblas_cgemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE TransA,
CBLAS_TRANSPOSE TransB, const int M, const int N,
const int K, const void *alpha, const void *A,
const int lda, const void *B, const int ldb,
const void *beta, void *C, const int ldc);
void cblas_zgemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE TransA,
CBLAS_TRANSPOSE TransB, const int M, const int N,
const int K, const void *alpha, const void *A,
const int lda, const void *B, const int ldb,
const void *beta, void *C, const int ldc);
int xerbla_(char *, int *);
int lsame_(char *, char *);
double slamch_(char* cmach);
double slamc3_(float *a, float *b);
double dlamch_(char* cmach);
double dlamc3_(double *a, double *b);
int dgels_(char *trans, int *m, int *n, int *nrhs, double *a,
int *lda, double *b, int *ldb, double *work, int *lwork, int *info);
int dgesv_(int *n, int *nrhs, double *a, int *lda, int *ipiv,
double *b, int *ldb, int *info);
int dgetrf_(int *m, int *n, double *a, int *lda, int *ipiv,
int *info);
int dposv_(char *uplo, int *n, int *nrhs, double *a, int *
lda, double *b, int *ldb, int *info);
int dpotrf_(char *uplo, int *n, double *a, int *lda, int *
info);
int sgels_(char *trans, int *m, int *n, int *nrhs, float *a,
int *lda, float *b, int *ldb, float *work, int *lwork, int *info);
int sgeev_(char *jobvl, char *jobvr, int *n, float *a, int *
lda, float *wr, float *wi, float *vl, int *ldvl, float *vr, int *
ldvr, float *work, int *lwork, int *info);
int sgeqrf_(int *m, int *n, float *a, int *lda, float *tau,
float *work, int *lwork, int *info);
int sgesv_(int *n, int *nrhs, float *a, int *lda, int *ipiv,
float *b, int *ldb, int *info);
int sgetrf_(int *m, int *n, float *a, int *lda, int *ipiv,
int *info);
int sposv_(char *uplo, int *n, int *nrhs, float *a, int *
lda, float *b, int *ldb, int *info);
int spotrf_(char *uplo, int *n, float *a, int *lda, int *
info);
int sgesdd_(char *jobz, int *m, int *n, float *a, int *lda,
float *s, float *u, int *ldu, float *vt, int *ldvt, float *work,
int *lwork, int *iwork, int *info);
#ifdef __cplusplus
}
#endif
#endif /* __CBLAS_H__ */
-129
View File
@@ -1,129 +0,0 @@
/* f2c.h -- Standard Fortran to C header file */
/** barf [ba:rf] 2. "He suggested using FORTRAN, and everybody barfed."
- From The Shogakukan DICTIONARY OF NEW ENGLISH (Second edition) */
#ifndef __F2C_H__
#define __F2C_H__
#include <assert.h>
#include <math.h>
#include <ctype.h>
#include <stdlib.h>
#include <string.h>
#include <stdio.h>
#include "cblas.h"
#include "lapack.h"
#ifdef __cplusplus
extern "C" {
#endif
#undef complex
typedef int integer;
typedef unsigned int uinteger;
typedef char *address;
typedef short int shortint;
typedef float real;
typedef double doublereal;
typedef lapack_complex complex;
typedef lapack_doublecomplex doublecomplex;
typedef int logical;
typedef short int shortlogical;
typedef char logical1;
typedef char integer1;
#define TRUE_ (1)
#define FALSE_ (0)
#ifndef abs
#define abs(x) ((x) >= 0 ? (x) : -(x))
#endif
#define dabs(x) (double)abs(x)
#ifndef min
#define min(a,b) ((a) <= (b) ? (a) : (b))
#endif
#ifndef max
#define max(a,b) ((a) >= (b) ? (a) : (b))
#endif
#define dmin(a,b) (double)min(a,b)
#define dmax(a,b) (double)max(a,b)
#define bit_test(a,b) ((a) >> (b) & 1)
#define bit_clear(a,b) ((a) & ~((uinteger)1 << (b)))
#define bit_set(a,b) ((a) | ((uinteger)1 << (b)))
static __inline double r_lg10(float *x)
{
return 0.43429448190325182765*log(*x);
}
static __inline double d_lg10(double *x)
{
return 0.43429448190325182765*log(*x);
}
static __inline double d_sign(double *a, double *b)
{
double x = fabs(*a);
return *b >= 0 ? x : -x;
}
static __inline double r_sign(float *a, float *b)
{
double x = fabs((double)*a);
return *b >= 0 ? x : -x;
}
static __inline int i_nint(float *x)
{
return (int)(*x >= 0 ? floor(*x + .5) : -floor(.5 - *x));
}
int pow_ii(int *ap, int *bp);
double pow_di(double *ap, int *bp);
static __inline double pow_ri(float *ap, int *bp)
{
double apd = *ap;
return pow_di(&apd, bp);
}
static __inline double pow_dd(double *ap, double *bp)
{
return pow(*ap, *bp);
}
static __inline void d_cnjg(doublecomplex *r, doublecomplex *z)
{
double zi = z->i;
r->r = z->r;
r->i = -zi;
}
static __inline void r_cnjg(complex *r, complex *z)
{
float zi = z->i;
r->r = z->r;
r->i = -zi;
}
static __inline int s_copy(char *a, char *b, int maxlen)
{
strncpy(a, b, maxlen);
a[maxlen] = '\0';
return 0;
}
int s_cat(char *lp, char **rpp, int* rnp, int *np);
int s_cmp(char *a0, char *b0);
static __inline int i_len(char* s)
{
return (int)strlen(s);
}
#ifdef __cplusplus
}
#endif
#endif
-386
View File
@@ -1,386 +0,0 @@
// this is auto-generated header for Lapack subset
#ifndef __CLAPACK_H__
#define __CLAPACK_H__
#include "cblas.h"
#ifdef __cplusplus
extern "C" {
#endif
int cgemm_(char *transa, char *transb, int *m, int *n, int *
k, lapack_complex *alpha, lapack_complex *a, int *lda, lapack_complex *b, int *ldb,
lapack_complex *beta, lapack_complex *c__, int *ldc);
int daxpy_(int *n, double *da, double *dx, int *incx, double
*dy, int *incy);
int dbdsdc_(char *uplo, char *compq, int *n, double *d__,
double *e, double *u, int *ldu, double *vt, int *ldvt, double *q, int
*iq, double *work, int *iwork, int *info);
int dbdsqr_(char *uplo, int *n, int *ncvt, int *nru, int *
ncc, double *d__, double *e, double *vt, int *ldvt, double *u, int *
ldu, double *c__, int *ldc, double *work, int *info);
int dcombssq_(double *v1, double *v2);
int dcopy_(int *n, double *dx, int *incx, double *dy, int *
incy);
double ddot_(int *n, double *dx, int *incx, double *dy, int *incy);
int dgebak_(char *job, char *side, int *n, int *ilo, int *
ihi, double *scale, int *m, double *v, int *ldv, int *info);
int dgebal_(char *job, int *n, double *a, int *lda, int *ilo,
int *ihi, double *scale, int *info);
int dgebd2_(int *m, int *n, double *a, int *lda, double *d__,
double *e, double *tauq, double *taup, double *work, int *info);
int dgebrd_(int *m, int *n, double *a, int *lda, double *d__,
double *e, double *tauq, double *taup, double *work, int *lwork, int
*info);
int dgeev_(char *jobvl, char *jobvr, int *n, double *a, int *
lda, double *wr, double *wi, double *vl, int *ldvl, double *vr, int *
ldvr, double *work, int *lwork, int *info);
int dgehd2_(int *n, int *ilo, int *ihi, double *a, int *lda,
double *tau, double *work, int *info);
int dgehrd_(int *n, int *ilo, int *ihi, double *a, int *lda,
double *tau, double *work, int *lwork, int *info);
int dgelq2_(int *m, int *n, double *a, int *lda, double *tau,
double *work, int *info);
int dgelqf_(int *m, int *n, double *a, int *lda, double *tau,
double *work, int *lwork, int *info);
int dgemm_(char *transa, char *transb, int *m, int *n, int *
k, double *alpha, double *a, int *lda, double *b, int *ldb, double *
beta, double *c__, int *ldc);
int dgemv_(char *trans, int *m, int *n, double *alpha,
double *a, int *lda, double *x, int *incx, double *beta, double *y,
int *incy);
int dgeqr2_(int *m, int *n, double *a, int *lda, double *tau,
double *work, int *info);
int dgeqrf_(int *m, int *n, double *a, int *lda, double *tau,
double *work, int *lwork, int *info);
int dger_(int *m, int *n, double *alpha, double *x, int *
incx, double *y, int *incy, double *a, int *lda);
int dgesdd_(char *jobz, int *m, int *n, double *a, int *lda,
double *s, double *u, int *ldu, double *vt, int *ldvt, double *work,
int *lwork, int *iwork, int *info);
int dhseqr_(char *job, char *compz, int *n, int *ilo, int *
ihi, double *h__, int *ldh, double *wr, double *wi, double *z__, int *
ldz, double *work, int *lwork, int *info);
int disnan_(double *din);
// "small" is a macro defined in Windows headers: https://stackoverflow.com/a/27794577
#ifdef small
#undef small
#endif
int dlabad_(double *small, double *large);
int dlabrd_(int *m, int *n, int *nb, double *a, int *lda,
double *d__, double *e, double *tauq, double *taup, double *x, int *
ldx, double *y, int *ldy);
int dlacpy_(char *uplo, int *m, int *n, double *a, int *lda,
double *b, int *ldb);
int dladiv1_(double *a, double *b, double *c__, double *d__,
double *p, double *q);
double dladiv2_(double *a, double *b, double *c__, double *d__, double *r__,
double *t);
int dladiv_(double *a, double *b, double *c__, double *d__,
double *p, double *q);
int dlaed6_(int *kniter, int *orgati, double *rho, double *
d__, double *z__, double *finit, double *tau, int *info);
int dlaexc_(int *wantq, int *n, double *t, int *ldt, double *
q, int *ldq, int *j1, int *n1, int *n2, double *work, int *info);
int dlahqr_(int *wantt, int *wantz, int *n, int *ilo, int *
ihi, double *h__, int *ldh, double *wr, double *wi, int *iloz, int *
ihiz, double *z__, int *ldz, int *info);
int dlahr2_(int *n, int *k, int *nb, double *a, int *lda,
double *tau, double *t, int *ldt, double *y, int *ldy);
int dlaisnan_(double *din1, double *din2);
int dlaln2_(int *ltrans, int *na, int *nw, double *smin,
double *ca, double *a, int *lda, double *d1, double *d2, double *b,
int *ldb, double *wr, double *wi, double *x, int *ldx, double *scale,
double *xnorm, int *info);
int dlamrg_(int *n1, int *n2, double *a, int *dtrd1, int *
dtrd2, int *index);
double dlange_(char *norm, int *m, int *n, double *a, int *lda, double *work);
double dlanst_(char *norm, int *n, double *d__, double *e);
int dlanv2_(double *a, double *b, double *c__, double *d__,
double *rt1r, double *rt1i, double *rt2r, double *rt2i, double *cs,
double *sn);
double dlapy2_(double *x, double *y);
int dlaqr0_(int *wantt, int *wantz, int *n, int *ilo, int *
ihi, double *h__, int *ldh, double *wr, double *wi, int *iloz, int *
ihiz, double *z__, int *ldz, double *work, int *lwork, int *info);
int dlaqr1_(int *n, double *h__, int *ldh, double *sr1,
double *si1, double *sr2, double *si2, double *v);
int dlaqr2_(int *wantt, int *wantz, int *n, int *ktop, int *
kbot, int *nw, double *h__, int *ldh, int *iloz, int *ihiz, double *
z__, int *ldz, int *ns, int *nd, double *sr, double *si, double *v,
int *ldv, int *nh, double *t, int *ldt, int *nv, double *wv, int *
ldwv, double *work, int *lwork);
int dlaqr3_(int *wantt, int *wantz, int *n, int *ktop, int *
kbot, int *nw, double *h__, int *ldh, int *iloz, int *ihiz, double *
z__, int *ldz, int *ns, int *nd, double *sr, double *si, double *v,
int *ldv, int *nh, double *t, int *ldt, int *nv, double *wv, int *
ldwv, double *work, int *lwork);
int dlaqr4_(int *wantt, int *wantz, int *n, int *ilo, int *
ihi, double *h__, int *ldh, double *wr, double *wi, int *iloz, int *
ihiz, double *z__, int *ldz, double *work, int *lwork, int *info);
int dlaqr5_(int *wantt, int *wantz, int *kacc22, int *n, int
*ktop, int *kbot, int *nshfts, double *sr, double *si, double *h__,
int *ldh, int *iloz, int *ihiz, double *z__, int *ldz, double *v, int
*ldv, double *u, int *ldu, int *nv, double *wv, int *ldwv, int *nh,
double *wh, int *ldwh);
int dlarf_(char *side, int *m, int *n, double *v, int *incv,
double *tau, double *c__, int *ldc, double *work);
int dlarfb_(char *side, char *trans, char *direct, char *
storev, int *m, int *n, int *k, double *v, int *ldv, double *t, int *
ldt, double *c__, int *ldc, double *work, int *ldwork);
int dlarfg_(int *n, double *alpha, double *x, int *incx,
double *tau);
int dlarft_(char *direct, char *storev, int *n, int *k,
double *v, int *ldv, double *tau, double *t, int *ldt);
int dlarfx_(char *side, int *m, int *n, double *v, double *
tau, double *c__, int *ldc, double *work);
int dlartg_(double *f, double *g, double *cs, double *sn,
double *r__);
int dlas2_(double *f, double *g, double *h__, double *ssmin,
double *ssmax);
int dlascl_(char *type__, int *kl, int *ku, double *cfrom,
double *cto, int *m, int *n, double *a, int *lda, int *info);
int dlasd0_(int *n, int *sqre, double *d__, double *e,
double *u, int *ldu, double *vt, int *ldvt, int *smlsiz, int *iwork,
double *work, int *info);
int dlasd1_(int *nl, int *nr, int *sqre, double *d__, double
*alpha, double *beta, double *u, int *ldu, double *vt, int *ldvt, int
*idxq, int *iwork, double *work, int *info);
int dlasd2_(int *nl, int *nr, int *sqre, int *k, double *d__,
double *z__, double *alpha, double *beta, double *u, int *ldu,
double *vt, int *ldvt, double *dsigma, double *u2, int *ldu2, double *
vt2, int *ldvt2, int *idxp, int *idx, int *idxc, int *idxq, int *
coltyp, int *info);
int dlasd3_(int *nl, int *nr, int *sqre, int *k, double *d__,
double *q, int *ldq, double *dsigma, double *u, int *ldu, double *u2,
int *ldu2, double *vt, int *ldvt, double *vt2, int *ldvt2, int *idxc,
int *ctot, double *z__, int *info);
int dlasd4_(int *n, int *i__, double *d__, double *z__,
double *delta, double *rho, double *sigma, double *work, int *info);
int dlasd5_(int *i__, double *d__, double *z__, double *
delta, double *rho, double *dsigma, double *work);
int dlasd6_(int *icompq, int *nl, int *nr, int *sqre, double
*d__, double *vf, double *vl, double *alpha, double *beta, int *idxq,
int *perm, int *givptr, int *givcol, int *ldgcol, double *givnum, int
*ldgnum, double *poles, double *difl, double *difr, double *z__, int *
k, double *c__, double *s, double *work, int *iwork, int *info);
int dlasd7_(int *icompq, int *nl, int *nr, int *sqre, int *k,
double *d__, double *z__, double *zw, double *vf, double *vfw,
double *vl, double *vlw, double *alpha, double *beta, double *dsigma,
int *idx, int *idxp, int *idxq, int *perm, int *givptr, int *givcol,
int *ldgcol, double *givnum, int *ldgnum, double *c__, double *s, int
*info);
int dlasd8_(int *icompq, int *k, double *d__, double *z__,
double *vf, double *vl, double *difl, double *difr, int *lddifr,
double *dsigma, double *work, int *info);
int dlasda_(int *icompq, int *smlsiz, int *n, int *sqre,
double *d__, double *e, double *u, int *ldu, double *vt, int *k,
double *difl, double *difr, double *z__, double *poles, int *givptr,
int *givcol, int *ldgcol, int *perm, double *givnum, double *c__,
double *s, double *work, int *iwork, int *info);
int dlasdq_(char *uplo, int *sqre, int *n, int *ncvt, int *
nru, int *ncc, double *d__, double *e, double *vt, int *ldvt, double *
u, int *ldu, double *c__, int *ldc, double *work, int *info);
int dlasdt_(int *n, int *lvl, int *nd, int *inode, int *
ndiml, int *ndimr, int *msub);
int dlaset_(char *uplo, int *m, int *n, double *alpha,
double *beta, double *a, int *lda);
int dlasq1_(int *n, double *d__, double *e, double *work,
int *info);
int dlasq2_(int *n, double *z__, int *info);
int dlasq3_(int *i0, int *n0, double *z__, int *pp, double *
dmin__, double *sigma, double *desig, double *qmax, int *nfail, int *
iter, int *ndiv, int *ieee, int *ttype, double *dmin1, double *dmin2,
double *dn, double *dn1, double *dn2, double *g, double *tau);
int dlasq4_(int *i0, int *n0, double *z__, int *pp, int *
n0in, double *dmin__, double *dmin1, double *dmin2, double *dn,
double *dn1, double *dn2, double *tau, int *ttype, double *g);
int dlasq5_(int *i0, int *n0, double *z__, int *pp, double *
tau, double *sigma, double *dmin__, double *dmin1, double *dmin2,
double *dn, double *dnm1, double *dnm2, int *ieee, double *eps);
int dlasq6_(int *i0, int *n0, double *z__, int *pp, double *
dmin__, double *dmin1, double *dmin2, double *dn, double *dnm1,
double *dnm2);
int dlasr_(char *side, char *pivot, char *direct, int *m,
int *n, double *c__, double *s, double *a, int *lda);
int dlasrt_(char *id, int *n, double *d__, int *info);
int dlassq_(int *n, double *x, int *incx, double *scale,
double *sumsq);
int dlasv2_(double *f, double *g, double *h__, double *ssmin,
double *ssmax, double *snr, double *csr, double *snl, double *csl);
int dlasy2_(int *ltranl, int *ltranr, int *isgn, int *n1,
int *n2, double *tl, int *ldtl, double *tr, int *ldtr, double *b, int
*ldb, double *scale, double *x, int *ldx, double *xnorm, int *info);
double dnrm2_(int *n, double *x, int *incx);
int dorg2r_(int *m, int *n, int *k, double *a, int *lda,
double *tau, double *work, int *info);
int dorgbr_(char *vect, int *m, int *n, int *k, double *a,
int *lda, double *tau, double *work, int *lwork, int *info);
int dorghr_(int *n, int *ilo, int *ihi, double *a, int *lda,
double *tau, double *work, int *lwork, int *info);
int dorgl2_(int *m, int *n, int *k, double *a, int *lda,
double *tau, double *work, int *info);
int dorglq_(int *m, int *n, int *k, double *a, int *lda,
double *tau, double *work, int *lwork, int *info);
int dorgqr_(int *m, int *n, int *k, double *a, int *lda,
double *tau, double *work, int *lwork, int *info);
int dorm2r_(char *side, char *trans, int *m, int *n, int *k,
double *a, int *lda, double *tau, double *c__, int *ldc, double *work,
int *info);
int dormbr_(char *vect, char *side, char *trans, int *m, int
*n, int *k, double *a, int *lda, double *tau, double *c__, int *ldc,
double *work, int *lwork, int *info);
int dormhr_(char *side, char *trans, int *m, int *n, int *
ilo, int *ihi, double *a, int *lda, double *tau, double *c__, int *
ldc, double *work, int *lwork, int *info);
int dorml2_(char *side, char *trans, int *m, int *n, int *k,
double *a, int *lda, double *tau, double *c__, int *ldc, double *work,
int *info);
int dormlq_(char *side, char *trans, int *m, int *n, int *k,
double *a, int *lda, double *tau, double *c__, int *ldc, double *work,
int *lwork, int *info);
int dormqr_(char *side, char *trans, int *m, int *n, int *k,
double *a, int *lda, double *tau, double *c__, int *ldc, double *work,
int *lwork, int *info);
int drot_(int *n, double *dx, int *incx, double *dy, int *
incy, double *c__, double *s);
int dscal_(int *n, double *da, double *dx, int *incx);
int dswap_(int *n, double *dx, int *incx, double *dy, int *
incy);
int dtrevc3_(char *side, char *howmny, int *select, int *n,
double *t, int *ldt, double *vl, int *ldvl, double *vr, int *ldvr,
int *mm, int *m, double *work, int *lwork, int *info);
int dtrexc_(char *compq, int *n, double *t, int *ldt, double
*q, int *ldq, int *ifst, int *ilst, double *work, int *info);
int dtrmm_(char *side, char *uplo, char *transa, char *diag,
int *m, int *n, double *alpha, double *a, int *lda, double *b, int *
ldb);
int dtrmv_(char *uplo, char *trans, char *diag, int *n,
double *a, int *lda, double *x, int *incx);
int idamax_(int *n, double *dx, int *incx);
int ieeeck_(int *ispec, float *zero, float *one);
int iladlc_(int *m, int *n, double *a, int *lda);
int iladlr_(int *m, int *n, double *a, int *lda);
int ilaenv_(int *ispec, char *name__, char *opts, int *n1, int *n2, int *n3,
int *n4);
int iparmq_(int *ispec, char *name__, char *opts, int *n, int *ilo, int *ihi,
int *lwork);
int sgemm_(char *transa, char *transb, int *m, int *n, int *
k, float *alpha, float *a, int *lda, float *b, int *ldb, float *beta,
float *c__, int *ldc);
int zgemm_(char *transa, char *transb, int *m, int *n, int *
k, lapack_doublecomplex *alpha, lapack_doublecomplex *a, int *lda, lapack_doublecomplex *b,
int *ldb, lapack_doublecomplex *beta, lapack_doublecomplex *c__, int *ldc);
#ifdef __cplusplus
}
#endif
#endif
-48
View File
@@ -1,48 +0,0 @@
Copyright (c) 1992-2017 The University of Tennessee and The University
of Tennessee Research Foundation. All rights
reserved.
Copyright (c) 2000-2017 The University of California Berkeley. All
rights reserved.
Copyright (c) 2006-2017 The University of Colorado Denver. All rights
reserved.
$COPYRIGHT$
Additional copyrights may follow
$HEADER$
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
- Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
- Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer listed
in this license in the documentation and/or other materials
provided with the distribution.
- Neither the name of the copyright holders nor the names of its
contributors may be used to endorse or promote products derived from
this software without specific prior written permission.
The copyright holders provide no reassurances that the source code
provided does not infringe any patent, copyright, or any other
intellectual property rights of third parties. The copyright holders
disclaim any liability to any recipient for claims brought against
recipient by any third party for infringement of that parties
intellectual property rights.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
-272
View File
@@ -1,272 +0,0 @@
appdoc = """
This is generator of CLapack subset.
The usage:
1. Make sure you have the special version of f2c installed.
Grab it from https://github.com/vpisarev/f2c/tree/for_lapack.
2. Download fresh version of Lapack from
https://github.com/Reference-LAPACK/lapack.
You may choose some specific version or the latest snapshot.
3. If necessary, edit "roots" and "banlist" variables in this script, specify the needed and unneeded functions
4. From within a working directory run
$ python3 <opencv_root>/3rdparty/clapack/make_clapack.py <lapack_root>
or
$ F2C=<path_to_custom_f2c> python3 <opencv_root>/3rdparty/clapack/make_clapack.py <lapack_root>
it will generate "new_clapack" directory with "include" and "src" subdirectories.
5. erase opencv/3rdparty/clapack/src and replace it with new_clapack/src.
6. copy new_clapack/include/lapack.h to opencv/3rdparty/clapack/include.
7. optionally, edit opencv/3rdparty/clapack/CMakeLists.txt and update CLAPACK_VERSION as needed.
This is it. Now build it and enjoy.
"""
import glob, re, os, shutil, subprocess, sys
roots = ["cgemm_", "dgemm_", "sgemm_", "zgemm_",
"dgeev_", "dgesdd_",
#"dsyevr_",
#"dgesv_", "dgetrf_", "dposv_", "dpotrf_", "dgels_", "dgeqrf_",
#"sgesv_", "sgetrf_", "sposv_", "spotrf_", "sgels_", "sgeqrf_"
]
banlist = ["slamch_", "slamc3_", "dlamch_", "dlamc3_", "lsame_", "xerbla_"]
if len(sys.argv) < 2:
print(appdoc)
sys.exit(0)
lapack_root = sys.argv[1]
dst_path = "."
def error(msg):
print ("error: " + msg)
sys.exit(0)
def file2fun(fname):
return (os.path.basename(fname)[:-2]).upper()
def print_graph(m):
for (k, neighbors) in sorted(m.items()):
print (k + " : " + ", ".join(sorted(list(neighbors))))
blas_path = os.path.join(lapack_root, "BLAS/SRC")
lapack_path = os.path.join(lapack_root, "SRC")
roots = [f[:-1].upper() for f in roots]
banlist = [f[:-1].upper() for f in banlist]
def fun2file(func):
filename = func.lower() + ".f"
blas_loc = blas_path + "/" + filename
lapack_loc = lapack_path + "/" + filename
if os.path.exists(blas_loc):
return blas_loc
elif os.path.exists(lapack_loc):
return lapack_loc
else:
error("neither %s nor %s exist" % (blas_loc, lapack_loc))
all_files = glob.glob(blas_path + "/*.f") + glob.glob(lapack_path + "/*.f")
all_funcs = [file2fun(fname) for fname in all_files]
all_funcs_set = set(all_funcs).difference(set(banlist))
all_funcs = sorted(list(all_funcs_set))
func_deps = {}
#print all_funcs
words_regexp = re.compile(r'\w+')
def scan_deps(func):
global func_deps
if func in func_deps:
return
func_deps[func] = set([]) # to avoid possibly infinite recursion
f = open(fun2file(func), 'rt')
deps = []
external_mode = False
for l in f.readlines():
if l.startswith('*'):
continue
l = l.strip().upper()
if l.startswith('EXTERNAL '):
external_mode = True
elif l.startswith('$') and external_mode:
pass
else:
external_mode = False
if not external_mode:
continue
for w in words_regexp.findall(l):
if w in all_funcs_set:
deps.append(w)
f.close()
# remove func from its dependencies
deps = set(deps).difference(set([func]))
func_deps[func] = deps
for d in deps:
scan_deps(d)
for r in roots:
scan_deps(r)
selected_funcs = sorted(func_deps.keys())
print ("total files before amalgamation: %d" % len(selected_funcs))
inv_deps = {}
for func in selected_funcs:
inv_deps[func] = set([])
for (func, deps) in func_deps.items():
for d in deps:
inv_deps[d] = inv_deps[d].union(set([func]))
#print_graph(inv_deps)
func_home = {}
for func in selected_funcs:
func_home[func] = func
def get_home0(func, func0):
used_by = inv_deps[func]
if len(used_by) == 1:
p = list(used_by)[0]
if p != func and p != func0:
return get_home0(p, func0)
return func
return func
# try to merge some files
for func in selected_funcs:
func_home[func] = get_home0(func, func)
# try to merge some files even more
for iters in range(100):
homes_changed = False
for (func, used_by) in inv_deps.items():
p0 = func_home[func]
n = len(used_by)
if n == 1:
p = list(used_by)[0]
p1 = func_home[p]
if p1 != p0:
func_home[func] = p1
homes_changed = True
continue
elif n > 1:
phomes = set([])
for p in used_by:
phomes.add(func_home[p])
if len(phomes) == 1:
p1 = list(phomes)[0]
if p1 != p0:
func_home[func] = p1
homes_changed = True
if not homes_changed:
break
res_files = {}
for (func, h) in func_home.items():
elems = res_files.get(h, set([]))
elems.add(func)
res_files[h] = elems
print ("total files after amalgamation: %d" % len(res_files))
#print_graph(res_files)
outdir = os.path.join(dst_path, "new_clapack")
outdir_src = os.path.join(outdir, "src")
outdir_inc = os.path.join(outdir, "include")
shutil.rmtree(outdir, ignore_errors=True)
try:
os.makedirs(outdir_src)
except os.error:
pass
try:
os.makedirs(outdir_inc)
except os.error:
pass
f2c_appname = os.getenv("F2C", default="f2c")
print ("f2c used: %s" % f2c_appname)
f2c_getver_cmd = f2c_appname + " -v"
verstr = subprocess.check_output(f2c_getver_cmd.split(' ')).decode("utf-8")
if "for_lapack" not in verstr:
error("invalid version of f2c\n" + appdoc)
f2c_flags = "-ctypes -localconst -no-proto"
f2c_cmd0 = f2c_appname + " " + f2c_flags
f2c_cmd1 = f2c_appname + " -hdr none " + f2c_flags
lapack_protos = {}
extract_fn_regexp = re.compile(r'.+?(\w+)\s*\(')
def extract_proto(func, csrc):
global lapack_protos
cname = func.lower() + "_"
cfname = func.lower() + ".c"
regexp_str = r'\n(?:/\* Subroutine \*/\s*)?\w+\s+\w+\s*\((?:.|\n)+?\)[\s\n]*\{'
proto_regexp = re.compile(regexp_str)
ps = proto_regexp.findall(csrc)
for p in ps:
n = p.find("*/")
if n < 0:
n = 0
else:
n += 2
p = p[n:-1].strip() + ";"
fns = extract_fn_regexp.findall(p)
if len(fns) != 1:
error("prototype of function (%s) when analyzing %s cannot be parsed" % (p, cfname))
fn = fns[0]
if fn not in lapack_protos:
p = re.sub(r'\bcomplex\b', 'lapack_complex', p)
p = re.sub(r'\bdoublecomplex\b', 'lapack_doublecomplex', p)
lapack_protos[fn] = p
for (filename, funcs) in sorted(res_files.items()):
out = ""
f2c_cmd = f2c_cmd0
for func in sorted(list(funcs)):
ffilename = fun2file(func)
print ("running " + f2c_cmd + " on " + ffilename + " ...")
ffile = open(ffilename, 'rt')
delta_out = subprocess.check_output(f2c_cmd.split(' '), stdin=ffile).decode("utf-8")
# remove trailing whitespaces
delta_out = '\n'.join([l.rstrip() for l in delta_out.split('\n')])
extract_proto(func, delta_out)
out += delta_out
ffile.close()
f2c_cmd = f2c_cmd1
outname = os.path.join(outdir_src, filename.lower() + ".c")
outfile = open(outname, 'wt')
outfile.write(out)
outfile.close()
proto_hdr = """// this is auto-generated header for Lapack subset
#ifndef __CLAPACK_H__
#define __CLAPACK_H__
#include "cblas.h"
#ifdef __cplusplus
extern "C" {
#endif
%s
#ifdef __cplusplus
}
#endif
#endif
""" % "\n\n".join([p for (n, p) in sorted(lapack_protos.items())])
proto_hdr_fname = os.path.join(outdir_inc, "lapack.h")
f = open(proto_hdr_fname, 'wt')
f.write(proto_hdr)
f.close()
-289
View File
@@ -1,289 +0,0 @@
#include "f2c.h"
#include <stdarg.h>
void cblas_cgemm(const CBLAS_LAYOUT layout, const CBLAS_TRANSPOSE TransA,
const CBLAS_TRANSPOSE TransB, const int M, const int N,
const int K, const void *alpha, const void *A,
const int lda, const void *B, const int ldb,
const void *beta, void *C, const int ldc)
{
char TA, TB;
if( layout == CblasColMajor )
{
if(TransA == CblasTrans) TA='T';
else if ( TransA == CblasConjTrans ) TA='C';
else if ( TransA == CblasNoTrans ) TA='N';
else
{
cblas_xerbla(layout, 2, "cblas_cgemm", "Illegal TransA setting, %d\n", TransA);
return;
}
if(TransB == CblasTrans) TB='T';
else if ( TransB == CblasConjTrans ) TB='C';
else if ( TransB == CblasNoTrans ) TB='N';
else
{
cblas_xerbla(layout, 3, "cblas_cgemm", "Illegal TransB setting, %d\n", TransB);
return;
}
cgemm_(&TA, &TB, (int*)&M, (int*)&N, (int*)&K, (complex*)alpha, (complex*)A, (int*)&lda,
(complex*)B, (int*)&ldb, (complex*)beta, (complex*)C, (int*)&ldc);
}
else if (layout == CblasRowMajor)
{
if(TransA == CblasTrans) TB='T';
else if ( TransA == CblasConjTrans ) TB='C';
else if ( TransA == CblasNoTrans ) TB='N';
else
{
cblas_xerbla(layout, 2, "cblas_cgemm", "Illegal TransA setting, %d\n", TransA);
return;
}
if(TransB == CblasTrans) TA='T';
else if ( TransB == CblasConjTrans ) TA='C';
else if ( TransB == CblasNoTrans ) TA='N';
else
{
cblas_xerbla(layout, 2, "cblas_cgemm", "Illegal TransB setting, %d\n", TransB);
return;
}
cgemm_(&TA, &TB, (int*)&N, (int*)&M, (int*)&K, (complex*)alpha, (complex*)B, (int*)&ldb,
(complex*)A, (int*)&lda, (complex*)beta, (complex*)C, (int*)&ldc);
}
else cblas_xerbla(layout, 1, "cblas_cgemm", "Illegal layout setting, %d\n", layout);
}
void cblas_dgemm(const CBLAS_LAYOUT layout, const CBLAS_TRANSPOSE TransA,
const CBLAS_TRANSPOSE TransB, const int M, const int N,
const int K, const double alpha, const double *A,
const int lda, const double *B, const int ldb,
const double beta, double *C, const int ldc)
{
char TA, TB;
if( layout == CblasColMajor )
{
if(TransA == CblasTrans) TA='T';
else if ( TransA == CblasConjTrans ) TA='C';
else if ( TransA == CblasNoTrans ) TA='N';
else
{
cblas_xerbla(layout, 2, "cblas_dgemm", "Illegal TransA setting, %d\n", TransA);
return;
}
if(TransB == CblasTrans) TB='T';
else if ( TransB == CblasConjTrans ) TB='C';
else if ( TransB == CblasNoTrans ) TB='N';
else
{
cblas_xerbla(layout, 3, "cblas_dgemm", "Illegal TransB setting, %d\n", TransB);
return;
}
dgemm_(&TA, &TB, (int*)&M, (int*)&N, (int*)&K, (double*)&alpha, (double*)A, (int*)&lda,
(double*)B, (int*)&ldb, (double*)&beta, (double*)C, (int*)&ldc);
}
else if (layout == CblasRowMajor)
{
if(TransA == CblasTrans) TB='T';
else if ( TransA == CblasConjTrans ) TB='C';
else if ( TransA == CblasNoTrans ) TB='N';
else
{
cblas_xerbla(layout, 2, "cblas_dgemm", "Illegal TransA setting, %d\n", TransA);
return;
}
if(TransB == CblasTrans) TA='T';
else if ( TransB == CblasConjTrans ) TA='C';
else if ( TransB == CblasNoTrans ) TA='N';
else
{
cblas_xerbla(layout, 2, "cblas_dgemm", "Illegal TransB setting, %d\n", TransB);
return;
}
dgemm_(&TA, &TB, (int*)&N, (int*)&M, (int*)&K, (double*)&alpha, (double*)B, (int*)&ldb,
(double*)A, (int*)&lda, (double*)&beta, (double*)C, (int*)&ldc);
}
else cblas_xerbla(layout, 1, "cblas_dgemm", "Illegal layout setting, %d\n", layout);
}
void cblas_sgemm(const CBLAS_LAYOUT layout, const CBLAS_TRANSPOSE TransA,
const CBLAS_TRANSPOSE TransB, const int M, const int N,
const int K, const float alpha, const float *A,
const int lda, const float *B, const int ldb,
const float beta, float *C, const int ldc)
{
char TA, TB;
if( layout == CblasColMajor )
{
if(TransA == CblasTrans) TA='T';
else if ( TransA == CblasConjTrans ) TA='C';
else if ( TransA == CblasNoTrans ) TA='N';
else
{
cblas_xerbla(layout, 2, "cblas_sgemm", "Illegal TransA setting, %d\n", TransA);
return;
}
if(TransB == CblasTrans) TB='T';
else if ( TransB == CblasConjTrans ) TB='C';
else if ( TransB == CblasNoTrans ) TB='N';
else
{
cblas_xerbla(layout, 3, "cblas_sgemm", "Illegal TransB setting, %d\n", TransB);
return;
}
sgemm_(&TA, &TB, (int*)&M, (int*)&N, (int*)&K, (float*)&alpha, (float*)A, (int*)&lda,
(float*)B, (int*)&ldb, (float*)&beta, (float*)C, (int*)&ldc);
}
else if (layout == CblasRowMajor)
{
if(TransA == CblasTrans) TB='T';
else if ( TransA == CblasConjTrans ) TB='C';
else if ( TransA == CblasNoTrans ) TB='N';
else
{
cblas_xerbla(layout, 2, "cblas_sgemm", "Illegal TransA setting, %d\n", TransA);
return;
}
if(TransB == CblasTrans) TA='T';
else if ( TransB == CblasConjTrans ) TA='C';
else if ( TransB == CblasNoTrans ) TA='N';
else
{
cblas_xerbla(layout, 2, "cblas_sgemm", "Illegal TransB setting, %d\n", TransB);
return;
}
sgemm_(&TA, &TB, (int*)&N, (int*)&M, (int*)&K, (float*)&alpha, (float*)B, (int*)&ldb,
(float*)A, (int*)&lda, (float*)&beta, (float*)C, (int*)&ldc);
}
else cblas_xerbla(layout, 1, "cblas_sgemm", "Illegal layout setting, %d\n", layout);
}
void cblas_zgemm(const CBLAS_LAYOUT layout, const CBLAS_TRANSPOSE TransA,
const CBLAS_TRANSPOSE TransB, const int M, const int N,
const int K, const void *alpha, const void *A,
const int lda, const void *B, const int ldb,
const void *beta, void *C, const int ldc)
{
char TA, TB;
if( layout == CblasColMajor )
{
if(TransA == CblasTrans) TA='T';
else if ( TransA == CblasConjTrans ) TA='C';
else if ( TransA == CblasNoTrans ) TA='N';
else
{
cblas_xerbla(layout, 2, "cblas_zgemm", "Illegal TransA setting, %d\n", TransA);
return;
}
if(TransB == CblasTrans) TB='T';
else if ( TransB == CblasConjTrans ) TB='C';
else if ( TransB == CblasNoTrans ) TB='N';
else
{
cblas_xerbla(layout, 3, "cblas_zgemm", "Illegal TransB setting, %d\n", TransB);
return;
}
zgemm_(&TA, &TB, (int*)&M, (int*)&N, (int*)&K, (doublecomplex*)alpha, (doublecomplex*)A, (int*)&lda,
(doublecomplex*)B, (int*)&ldb, (doublecomplex*)beta, (doublecomplex*)C, (int*)&ldc);
}
else if (layout == CblasRowMajor)
{
if(TransA == CblasTrans) TB='T';
else if ( TransA == CblasConjTrans ) TB='C';
else if ( TransA == CblasNoTrans ) TB='N';
else
{
cblas_xerbla(layout, 2, "cblas_zgemm", "Illegal TransA setting, %d\n", TransA);
return;
}
if(TransB == CblasTrans) TA='T';
else if ( TransB == CblasConjTrans ) TA='C';
else if ( TransB == CblasNoTrans ) TA='N';
else
{
cblas_xerbla(layout, 2, "cblas_zgemm", "Illegal TransB setting, %d\n", TransB);
return;
}
zgemm_(&TA, &TB, (int*)&N, (int*)&M, (int*)&K, (doublecomplex*)alpha, (doublecomplex*)B, (int*)&ldb,
(doublecomplex*)A, (int*)&lda, (doublecomplex*)beta, (doublecomplex*)C, (int*)&ldc);
}
else cblas_xerbla(layout, 1, "cblas_zgemm", "Illegal layout setting, %d\n", layout);
}
void cblas_xerbla(const CBLAS_LAYOUT layout, int info, const char *rout, const char *form, ...)
{
extern int RowMajorStrg;
char empty[1] = "";
va_list argptr;
va_start(argptr, form);
if (layout == CblasRowMajor)
{
if (strstr(rout,"gemm") != 0)
{
if (info == 5 ) info = 4;
else if (info == 4 ) info = 5;
else if (info == 11) info = 9;
else if (info == 9 ) info = 11;
}
else if (strstr(rout,"symm") != 0 || strstr(rout,"hemm") != 0)
{
if (info == 5 ) info = 4;
else if (info == 4 ) info = 5;
}
else if (strstr(rout,"trmm") != 0 || strstr(rout,"trsm") != 0)
{
if (info == 7 ) info = 6;
else if (info == 6 ) info = 7;
}
else if (strstr(rout,"gemv") != 0)
{
if (info == 4) info = 3;
else if (info == 3) info = 4;
}
else if (strstr(rout,"gbmv") != 0)
{
if (info == 4) info = 3;
else if (info == 3) info = 4;
else if (info == 6) info = 5;
else if (info == 5) info = 6;
}
else if (strstr(rout,"ger") != 0)
{
if (info == 3) info = 2;
else if (info == 2) info = 3;
else if (info == 8) info = 6;
else if (info == 6) info = 8;
}
else if ( (strstr(rout,"her2") != 0 || strstr(rout,"hpr2") != 0)
&& strstr(rout,"her2k") == 0 )
{
if (info == 8) info = 6;
else if (info == 6) info = 8;
}
}
if (info)
fprintf(stderr, "Parameter %d to routine %s was incorrect\n", info, rout);
vfprintf(stderr, form, argptr);
va_end(argptr);
if (info && !info)
xerbla_(empty, &info); /* Force link of our F77 error handler */
exit(-1);
}
-72
View File
@@ -1,72 +0,0 @@
#include "f2c.h"
#include <float.h>
#include <stdio.h>
/* *********************************************************************** */
double dlamc3_(double *a, double *b)
{
/* -- LAPACK auxiliary routine (version 3.1) -- */
/* Univ. of Tennessee, Univ. of California Berkeley and NAG Ltd.. */
/* November 2006 */
/* .. Scalar Arguments .. */
/* .. */
/* Purpose */
/* ======= */
/* DLAMC3 is intended to force A and B to be stored prior to doing */
/* the addition of A and B , for use in situations where optimizers */
/* might hold one of these in a register. */
/* Arguments */
/* ========= */
/* A (input) DOUBLE PRECISION */
/* B (input) DOUBLE PRECISION */
/* The values A and B. */
/* ===================================================================== */
/* .. Executable Statements .. */
double ret_val = *a + *b;
return ret_val;
/* End of DLAMC3 */
} /* dlamc3_ */
/* simpler version of dlamch for the case of IEEE754-compliant FPU module by Piotr Luszczek S.
taken from http://www.mail-archive.com/numpy-discussion@lists.sourceforge.net/msg02448.html */
#ifndef DBL_DIGITS
#define DBL_DIGITS 53
#endif
static const unsigned char lapack_dlamch_tab0[] =
{
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 2, 0, 0, 0, 0, 0, 0, 3, 4, 5, 6, 7, 0, 8, 9, 0, 10, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 2, 0, 0, 0, 0, 0, 0, 3, 4, 5, 6, 7, 0, 8, 9,
0, 10, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
};
const double lapack_dlamch_tab1[] =
{
0, FLT_RADIX, DBL_EPSILON, DBL_MAX_EXP, DBL_MIN_EXP, DBL_DIGITS, DBL_MAX,
DBL_EPSILON*FLT_RADIX, 1, DBL_MIN*(1 + DBL_EPSILON), DBL_MIN
};
double dlamch_(char* cmach)
{
return lapack_dlamch_tab1[lapack_dlamch_tab0[(unsigned char)cmach[0]]];
}
-96
View File
@@ -1,96 +0,0 @@
#include "f2c.h"
static const int CLAPACK_NOT_IMPLEMENTED = -1024;
int sgesdd_(char *jobz, int *m, int *n, float *a, int *lda,
float *s, float *u, int *ldu, float *vt, int *ldvt, float *work,
int *lwork, int *iwork, int *info)
{
*info = CLAPACK_NOT_IMPLEMENTED;
return 0;
}
int dgels_(char *trans, int *m, int *n, int *nrhs, double *a,
int *lda, double *b, int *ldb, double *work, int *lwork, int *info)
{
*info = CLAPACK_NOT_IMPLEMENTED;
return 0;
}
int dgesv_(int *n, int *nrhs, double *a, int *lda, int *ipiv,
double *b, int *ldb, int *info)
{
*info = CLAPACK_NOT_IMPLEMENTED;
return 0;
}
int dgetrf_(int *m, int *n, double *a, int *lda, int *ipiv,
int *info)
{
*info = CLAPACK_NOT_IMPLEMENTED;
return 0;
}
int dposv_(char *uplo, int *n, int *nrhs, double *a, int *
lda, double *b, int *ldb, int *info)
{
*info = CLAPACK_NOT_IMPLEMENTED;
return 0;
}
int dpotrf_(char *uplo, int *n, double *a, int *lda, int *
info)
{
*info = CLAPACK_NOT_IMPLEMENTED;
return 0;
}
int sgels_(char *trans, int *m, int *n, int *nrhs, float *a,
int *lda, float *b, int *ldb, float *work, int *lwork, int *info)
{
*info = CLAPACK_NOT_IMPLEMENTED;
return 0;
}
int sgeev_(char *jobvl, char *jobvr, int *n, float *a, int *
lda, float *wr, float *wi, float *vl, int *ldvl, float *vr, int *
ldvr, float *work, int *lwork, int *info)
{
*info = CLAPACK_NOT_IMPLEMENTED;
return 0;
}
int sgeqrf_(int *m, int *n, float *a, int *lda, float *tau,
float *work, int *lwork, int *info)
{
*info = CLAPACK_NOT_IMPLEMENTED;
return 0;
}
int sgesv_(int *n, int *nrhs, float *a, int *lda, int *ipiv,
float *b, int *ldb, int *info)
{
*info = CLAPACK_NOT_IMPLEMENTED;
return 0;
}
int sgetrf_(int *m, int *n, float *a, int *lda, int *ipiv,
int *info)
{
*info = CLAPACK_NOT_IMPLEMENTED;
return 0;
}
int sposv_(char *uplo, int *n, int *nrhs, float *a, int *
lda, float *b, int *ldb, int *info)
{
*info = CLAPACK_NOT_IMPLEMENTED;
return 0;
}
int spotrf_(char *uplo, int *n, float *a, int *lda, int *
info)
{
*info = CLAPACK_NOT_IMPLEMENTED;
return 0;
}
-25
View File
@@ -1,25 +0,0 @@
#include "f2c.h"
static const unsigned char lapack_toupper_tab[] =
{
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23,
24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45,
46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67,
68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89,
90, 91, 92, 93, 94, 95, 96, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79,
80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 123, 124, 125, 126, 127, 128, 129, 130, 131,
132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149,
150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167,
168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185,
186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203,
204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221,
222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239,
240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254, 255
};
#define lapack_toupper(c) ((char)lapack_toupper_tab[(unsigned char)(c)])
int lsame_(char *ca, char *cb)
{
return lapack_toupper(ca[0]) == lapack_toupper(cb[0]);
}
-27
View File
@@ -1,27 +0,0 @@
#include "f2c.h"
double pow_di(double *ap, int *bp)
{
double p = 1;
double x = *ap;
int n = *bp;
if(n != 0)
{
if(n < 0)
{
n = -n;
x = 1/x;
}
unsigned u = (unsigned)n;
for(;;)
{
if((u & 1) != 0)
p *= x;
if((u >>= 1) == 0)
break;
x *= x;
}
}
return p;
}
-25
View File
@@ -1,25 +0,0 @@
#include "f2c.h"
int pow_ii(int *ap, int *bp)
{
int p;
int x = *ap;
int n = *bp;
if (n <= 0) {
if (n == 0 || x == 1)
return 1;
return x != -1 ? 0 : (n & 1) ? -1 : 1;
}
unsigned u = (unsigned)n;
for(p = 1; ; )
{
if(u & 01)
p *= x;
if(u >>= 1)
x *= x;
else
break;
}
return p;
}
-22
View File
@@ -1,22 +0,0 @@
/* Unless compiled with -DNO_OVERWRITE, this variant of s_cat allows the
* target of a concatenation to appear on its right-hand side (contrary
* to the Fortran 77 Standard, but in accordance with Fortran 90).
*/
#include "f2c.h"
int s_cat(char *lp, char **rpp, int* rnp, int *np)
{
int i, L = 0;
int n = *np;
for(i = 0; i < n; i++) {
int ni = rnp[i];
if(ni > 0) {
memcpy(lp + L, rpp[i], ni);
L += ni;
}
}
lp[L] = '\0';
return 0;
}
-40
View File
@@ -1,40 +0,0 @@
#include "f2c.h"
/* compare two strings */
int s_cmp(char *a0, char *b0)
{
int la = (int)strlen(a0);
int lb = (int)strlen(b0);
unsigned char *a, *aend, *b, *bend;
a = (unsigned char *)a0;
b = (unsigned char *)b0;
aend = a + la;
bend = b + lb;
if(la <= lb)
{
while(a < aend)
if(*a != *b)
return( *a - *b );
else
{ ++a; ++b; }
while(b < bend)
if(*b != ' ')
return( ' ' - *b );
else ++b;
}
else
{
while(b < bend)
if(*a == *b)
{ ++a; ++b; }
else
return( *a - *b );
while(a < aend)
if(*a != ' ')
return(*a - ' ');
else ++a;
}
return(0);
}
-71
View File
@@ -1,71 +0,0 @@
#include "f2c.h"
#include <float.h>
#include <stdio.h>
/* *********************************************************************** */
double slamc3_(float *a, float *b)
{
/* -- LAPACK auxiliary routine (version 3.1) -- */
/* Univ. of Tennessee, Univ. of California Berkeley and NAG Ltd.. */
/* November 2006 */
/* .. Scalar Arguments .. */
/* .. */
/* Purpose */
/* ======= */
/* SLAMC3 is intended to force A and B to be stored prior to doing */
/* the addition of A and B , for use in situations where optimizers */
/* might hold one of these in a register. */
/* Arguments */
/* ========= */
/* A (input) REAL */
/* B (input) REAL */
/* The values A and B. */
/* ===================================================================== */
/* .. Executable Statements .. */
float ret_val = *a + *b;
return ret_val;
/* End of SLAMC3 */
} /* slamc3_ */
/* simpler version of slamch for the case of IEEE754-compliant FPU module by Piotr Luszczek S.
taken from http://www.mail-archive.com/numpy-discussion@lists.sourceforge.net/msg02448.html */
#ifndef FLT_DIGITS
#define FLT_DIGITS 24
#endif
static const unsigned char lapack_slamch_tab0[] =
{
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 2, 0, 0, 0, 0, 0, 0, 3, 4, 5, 6, 7, 0, 8, 9, 0, 10, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 2, 0, 0, 0, 0, 0, 0, 3, 4, 5, 6, 7, 0, 8, 9,
0, 10, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
};
const double lapack_slamch_tab1[] =
{
0, FLT_RADIX, FLT_EPSILON, FLT_MAX_EXP, FLT_MIN_EXP, FLT_DIGITS, FLT_MAX,
FLT_EPSILON*FLT_RADIX, 1, FLT_MIN*(1 + FLT_EPSILON), FLT_MIN
};
double slamch_(char* cmach)
{
return lapack_slamch_tab1[lapack_slamch_tab0[(unsigned char)cmach[0]]];
}
-19
View File
@@ -1,19 +0,0 @@
/* xerbla.f -- translated by f2c (version 20061008).
You must link the resulting object file with libf2c:
on Microsoft Windows system, link with libf2c.lib;
on Linux or Unix systems, link with .../path/to/libf2c.a -lm
or, if you install libf2c.a in a standard place, with -lf2c -lm
-- in that order, at the end of the command line, as in
cc *.o -lf2c -lm
Source for libf2c is in /netlib/f2c/libf2c.zip, e.g.,
http://www.netlib.org/f2c/libf2c.zip
*/
#include "f2c.h"
/* Subroutine */ int xerbla_(char *srname, int *info)
{
printf("** On entry to %s, parameter number %2i had an illegal value\n", srname, *info);
return 0;
} /* xerbla_ */
-752
View File
@@ -1,752 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b CGEMM
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
// Definition:
// ===========
//
// SUBROUTINE CGEMM(TRANSA,TRANSB,M,N,K,ALPHA,A,LDA,B,LDB,BETA,C,LDC)
//
// .. Scalar Arguments ..
// COMPLEX ALPHA,BETA
// INTEGER K,LDA,LDB,LDC,M,N
// CHARACTER TRANSA,TRANSB
// ..
// .. Array Arguments ..
// COMPLEX A(LDA,*),B(LDB,*),C(LDC,*)
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> CGEMM performs one of the matrix-matrix operations
//>
//> C := alpha*op( A )*op( B ) + beta*C,
//>
//> where op( X ) is one of
//>
//> op( X ) = X or op( X ) = X**T or op( X ) = X**H,
//>
//> alpha and beta are scalars, and A, B and C are matrices, with op( A )
//> an m by k matrix, op( B ) a k by n matrix and C an m by n matrix.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] TRANSA
//> \verbatim
//> TRANSA is CHARACTER*1
//> On entry, TRANSA specifies the form of op( A ) to be used in
//> the matrix multiplication as follows:
//>
//> TRANSA = 'N' or 'n', op( A ) = A.
//>
//> TRANSA = 'T' or 't', op( A ) = A**T.
//>
//> TRANSA = 'C' or 'c', op( A ) = A**H.
//> \endverbatim
//>
//> \param[in] TRANSB
//> \verbatim
//> TRANSB is CHARACTER*1
//> On entry, TRANSB specifies the form of op( B ) to be used in
//> the matrix multiplication as follows:
//>
//> TRANSB = 'N' or 'n', op( B ) = B.
//>
//> TRANSB = 'T' or 't', op( B ) = B**T.
//>
//> TRANSB = 'C' or 'c', op( B ) = B**H.
//> \endverbatim
//>
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> On entry, M specifies the number of rows of the matrix
//> op( A ) and of the matrix C. M must be at least zero.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> On entry, N specifies the number of columns of the matrix
//> op( B ) and the number of columns of the matrix C. N must be
//> at least zero.
//> \endverbatim
//>
//> \param[in] K
//> \verbatim
//> K is INTEGER
//> On entry, K specifies the number of columns of the matrix
//> op( A ) and the number of rows of the matrix op( B ). K must
//> be at least zero.
//> \endverbatim
//>
//> \param[in] ALPHA
//> \verbatim
//> ALPHA is COMPLEX
//> On entry, ALPHA specifies the scalar alpha.
//> \endverbatim
//>
//> \param[in] A
//> \verbatim
//> A is COMPLEX array, dimension ( LDA, ka ), where ka is
//> k when TRANSA = 'N' or 'n', and is m otherwise.
//> Before entry with TRANSA = 'N' or 'n', the leading m by k
//> part of the array A must contain the matrix A, otherwise
//> the leading k by m part of the array A must contain the
//> matrix A.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> On entry, LDA specifies the first dimension of A as declared
//> in the calling (sub) program. When TRANSA = 'N' or 'n' then
//> LDA must be at least max( 1, m ), otherwise LDA must be at
//> least max( 1, k ).
//> \endverbatim
//>
//> \param[in] B
//> \verbatim
//> B is COMPLEX array, dimension ( LDB, kb ), where kb is
//> n when TRANSB = 'N' or 'n', and is k otherwise.
//> Before entry with TRANSB = 'N' or 'n', the leading k by n
//> part of the array B must contain the matrix B, otherwise
//> the leading n by k part of the array B must contain the
//> matrix B.
//> \endverbatim
//>
//> \param[in] LDB
//> \verbatim
//> LDB is INTEGER
//> On entry, LDB specifies the first dimension of B as declared
//> in the calling (sub) program. When TRANSB = 'N' or 'n' then
//> LDB must be at least max( 1, k ), otherwise LDB must be at
//> least max( 1, n ).
//> \endverbatim
//>
//> \param[in] BETA
//> \verbatim
//> BETA is COMPLEX
//> On entry, BETA specifies the scalar beta. When BETA is
//> supplied as zero then C need not be set on input.
//> \endverbatim
//>
//> \param[in,out] C
//> \verbatim
//> C is COMPLEX array, dimension ( LDC, N )
//> Before entry, the leading m by n part of the array C must
//> contain the matrix C, except when beta is zero, in which
//> case C need not be set on entry.
//> On exit, the array C is overwritten by the m by n matrix
//> ( alpha*op( A )*op( B ) + beta*C ).
//> \endverbatim
//>
//> \param[in] LDC
//> \verbatim
//> LDC is INTEGER
//> On entry, LDC specifies the first dimension of C as declared
//> in the calling (sub) program. LDC must be at least
//> max( 1, m ).
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup complex_blas_level3
//
//> \par Further Details:
// =====================
//>
//> \verbatim
//>
//> Level 3 Blas routine.
//>
//> -- Written on 8-February-1989.
//> Jack Dongarra, Argonne National Laboratory.
//> Iain Duff, AERE Harwell.
//> Jeremy Du Croz, Numerical Algorithms Group Ltd.
//> Sven Hammarling, Numerical Algorithms Group Ltd.
//> \endverbatim
//>
// =====================================================================
/* Subroutine */ int cgemm_(char *transa, char *transb, int *m, int *n, int *
k, complex *alpha, complex *a, int *lda, complex *b, int *ldb,
complex *beta, complex *c__, int *ldc)
{
// Table of constant values
complex c_b1 = {1.f,0.f};
complex c_b2 = {0.f,0.f};
// System generated locals
int a_dim1, a_offset, b_dim1, b_offset, c_dim1, c_offset, i__1, i__2,
i__3, i__4, i__5, i__6;
complex q__1, q__2, q__3, q__4;
// Local variables
int i__, j, l, info;
int nota, notb;
complex temp;
int conja, conjb;
int ncola;
extern int lsame_(char *, char *);
int nrowa, nrowb;
extern /* Subroutine */ int xerbla_(char *, int *);
//
// -- Reference BLAS level3 routine (version 3.7.0) --
// -- Reference BLAS is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. External Functions ..
// ..
// .. External Subroutines ..
// ..
// .. Intrinsic Functions ..
// ..
// .. Local Scalars ..
// ..
// .. Parameters ..
// ..
//
// Set NOTA and NOTB as true if A and B respectively are not
// conjugated or transposed, set CONJA and CONJB as true if A and
// B respectively are to be transposed but not conjugated and set
// NROWA, NCOLA and NROWB as the number of rows and columns of A
// and the number of rows of B respectively.
//
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
b_dim1 = *ldb;
b_offset = 1 + b_dim1;
b -= b_offset;
c_dim1 = *ldc;
c_offset = 1 + c_dim1;
c__ -= c_offset;
// Function Body
nota = lsame_(transa, "N");
notb = lsame_(transb, "N");
conja = lsame_(transa, "C");
conjb = lsame_(transb, "C");
if (nota) {
nrowa = *m;
ncola = *k;
} else {
nrowa = *k;
ncola = *m;
}
if (notb) {
nrowb = *k;
} else {
nrowb = *n;
}
//
// Test the input parameters.
//
info = 0;
if (! nota && ! conja && ! lsame_(transa, "T")) {
info = 1;
} else if (! notb && ! conjb && ! lsame_(transb, "T")) {
info = 2;
} else if (*m < 0) {
info = 3;
} else if (*n < 0) {
info = 4;
} else if (*k < 0) {
info = 5;
} else if (*lda < max(1,nrowa)) {
info = 8;
} else if (*ldb < max(1,nrowb)) {
info = 10;
} else if (*ldc < max(1,*m)) {
info = 13;
}
if (info != 0) {
xerbla_("CGEMM ", &info);
return 0;
}
//
// Quick return if possible.
//
if (*m == 0 || *n == 0 || (alpha->r == 0.f && alpha->i == 0.f || *k == 0)
&& (beta->r == 1.f && beta->i == 0.f)) {
return 0;
}
//
// And when alpha.eq.zero.
//
if (alpha->r == 0.f && alpha->i == 0.f) {
if (beta->r == 0.f && beta->i == 0.f) {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
i__3 = i__ + j * c_dim1;
c__[i__3].r = 0.f, c__[i__3].i = 0.f;
// L10:
}
// L20:
}
} else {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
i__3 = i__ + j * c_dim1;
i__4 = i__ + j * c_dim1;
q__1.r = beta->r * c__[i__4].r - beta->i * c__[i__4].i,
q__1.i = beta->r * c__[i__4].i + beta->i * c__[
i__4].r;
c__[i__3].r = q__1.r, c__[i__3].i = q__1.i;
// L30:
}
// L40:
}
}
return 0;
}
//
// Start the operations.
//
if (notb) {
if (nota) {
//
// Form C := alpha*A*B + beta*C.
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
if (beta->r == 0.f && beta->i == 0.f) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
i__3 = i__ + j * c_dim1;
c__[i__3].r = 0.f, c__[i__3].i = 0.f;
// L50:
}
} else if (beta->r != 1.f || beta->i != 0.f) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
i__3 = i__ + j * c_dim1;
i__4 = i__ + j * c_dim1;
q__1.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, q__1.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
c__[i__3].r = q__1.r, c__[i__3].i = q__1.i;
// L60:
}
}
i__2 = *k;
for (l = 1; l <= i__2; ++l) {
i__3 = l + j * b_dim1;
q__1.r = alpha->r * b[i__3].r - alpha->i * b[i__3].i,
q__1.i = alpha->r * b[i__3].i + alpha->i * b[i__3]
.r;
temp.r = q__1.r, temp.i = q__1.i;
i__3 = *m;
for (i__ = 1; i__ <= i__3; ++i__) {
i__4 = i__ + j * c_dim1;
i__5 = i__ + j * c_dim1;
i__6 = i__ + l * a_dim1;
q__2.r = temp.r * a[i__6].r - temp.i * a[i__6].i,
q__2.i = temp.r * a[i__6].i + temp.i * a[i__6]
.r;
q__1.r = c__[i__5].r + q__2.r, q__1.i = c__[i__5].i +
q__2.i;
c__[i__4].r = q__1.r, c__[i__4].i = q__1.i;
// L70:
}
// L80:
}
// L90:
}
} else if (conja) {
//
// Form C := alpha*A**H*B + beta*C.
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp.r = 0.f, temp.i = 0.f;
i__3 = *k;
for (l = 1; l <= i__3; ++l) {
r_cnjg(&q__3, &a[l + i__ * a_dim1]);
i__4 = l + j * b_dim1;
q__2.r = q__3.r * b[i__4].r - q__3.i * b[i__4].i,
q__2.i = q__3.r * b[i__4].i + q__3.i * b[i__4]
.r;
q__1.r = temp.r + q__2.r, q__1.i = temp.i + q__2.i;
temp.r = q__1.r, temp.i = q__1.i;
// L100:
}
if (beta->r == 0.f && beta->i == 0.f) {
i__3 = i__ + j * c_dim1;
q__1.r = alpha->r * temp.r - alpha->i * temp.i,
q__1.i = alpha->r * temp.i + alpha->i *
temp.r;
c__[i__3].r = q__1.r, c__[i__3].i = q__1.i;
} else {
i__3 = i__ + j * c_dim1;
q__2.r = alpha->r * temp.r - alpha->i * temp.i,
q__2.i = alpha->r * temp.i + alpha->i *
temp.r;
i__4 = i__ + j * c_dim1;
q__3.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, q__3.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
q__1.r = q__2.r + q__3.r, q__1.i = q__2.i + q__3.i;
c__[i__3].r = q__1.r, c__[i__3].i = q__1.i;
}
// L110:
}
// L120:
}
} else {
//
// Form C := alpha*A**T*B + beta*C
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp.r = 0.f, temp.i = 0.f;
i__3 = *k;
for (l = 1; l <= i__3; ++l) {
i__4 = l + i__ * a_dim1;
i__5 = l + j * b_dim1;
q__2.r = a[i__4].r * b[i__5].r - a[i__4].i * b[i__5]
.i, q__2.i = a[i__4].r * b[i__5].i + a[i__4]
.i * b[i__5].r;
q__1.r = temp.r + q__2.r, q__1.i = temp.i + q__2.i;
temp.r = q__1.r, temp.i = q__1.i;
// L130:
}
if (beta->r == 0.f && beta->i == 0.f) {
i__3 = i__ + j * c_dim1;
q__1.r = alpha->r * temp.r - alpha->i * temp.i,
q__1.i = alpha->r * temp.i + alpha->i *
temp.r;
c__[i__3].r = q__1.r, c__[i__3].i = q__1.i;
} else {
i__3 = i__ + j * c_dim1;
q__2.r = alpha->r * temp.r - alpha->i * temp.i,
q__2.i = alpha->r * temp.i + alpha->i *
temp.r;
i__4 = i__ + j * c_dim1;
q__3.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, q__3.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
q__1.r = q__2.r + q__3.r, q__1.i = q__2.i + q__3.i;
c__[i__3].r = q__1.r, c__[i__3].i = q__1.i;
}
// L140:
}
// L150:
}
}
} else if (nota) {
if (conjb) {
//
// Form C := alpha*A*B**H + beta*C.
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
if (beta->r == 0.f && beta->i == 0.f) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
i__3 = i__ + j * c_dim1;
c__[i__3].r = 0.f, c__[i__3].i = 0.f;
// L160:
}
} else if (beta->r != 1.f || beta->i != 0.f) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
i__3 = i__ + j * c_dim1;
i__4 = i__ + j * c_dim1;
q__1.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, q__1.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
c__[i__3].r = q__1.r, c__[i__3].i = q__1.i;
// L170:
}
}
i__2 = *k;
for (l = 1; l <= i__2; ++l) {
r_cnjg(&q__2, &b[j + l * b_dim1]);
q__1.r = alpha->r * q__2.r - alpha->i * q__2.i, q__1.i =
alpha->r * q__2.i + alpha->i * q__2.r;
temp.r = q__1.r, temp.i = q__1.i;
i__3 = *m;
for (i__ = 1; i__ <= i__3; ++i__) {
i__4 = i__ + j * c_dim1;
i__5 = i__ + j * c_dim1;
i__6 = i__ + l * a_dim1;
q__2.r = temp.r * a[i__6].r - temp.i * a[i__6].i,
q__2.i = temp.r * a[i__6].i + temp.i * a[i__6]
.r;
q__1.r = c__[i__5].r + q__2.r, q__1.i = c__[i__5].i +
q__2.i;
c__[i__4].r = q__1.r, c__[i__4].i = q__1.i;
// L180:
}
// L190:
}
// L200:
}
} else {
//
// Form C := alpha*A*B**T + beta*C
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
if (beta->r == 0.f && beta->i == 0.f) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
i__3 = i__ + j * c_dim1;
c__[i__3].r = 0.f, c__[i__3].i = 0.f;
// L210:
}
} else if (beta->r != 1.f || beta->i != 0.f) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
i__3 = i__ + j * c_dim1;
i__4 = i__ + j * c_dim1;
q__1.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, q__1.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
c__[i__3].r = q__1.r, c__[i__3].i = q__1.i;
// L220:
}
}
i__2 = *k;
for (l = 1; l <= i__2; ++l) {
i__3 = j + l * b_dim1;
q__1.r = alpha->r * b[i__3].r - alpha->i * b[i__3].i,
q__1.i = alpha->r * b[i__3].i + alpha->i * b[i__3]
.r;
temp.r = q__1.r, temp.i = q__1.i;
i__3 = *m;
for (i__ = 1; i__ <= i__3; ++i__) {
i__4 = i__ + j * c_dim1;
i__5 = i__ + j * c_dim1;
i__6 = i__ + l * a_dim1;
q__2.r = temp.r * a[i__6].r - temp.i * a[i__6].i,
q__2.i = temp.r * a[i__6].i + temp.i * a[i__6]
.r;
q__1.r = c__[i__5].r + q__2.r, q__1.i = c__[i__5].i +
q__2.i;
c__[i__4].r = q__1.r, c__[i__4].i = q__1.i;
// L230:
}
// L240:
}
// L250:
}
}
} else if (conja) {
if (conjb) {
//
// Form C := alpha*A**H*B**H + beta*C.
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp.r = 0.f, temp.i = 0.f;
i__3 = *k;
for (l = 1; l <= i__3; ++l) {
r_cnjg(&q__3, &a[l + i__ * a_dim1]);
r_cnjg(&q__4, &b[j + l * b_dim1]);
q__2.r = q__3.r * q__4.r - q__3.i * q__4.i, q__2.i =
q__3.r * q__4.i + q__3.i * q__4.r;
q__1.r = temp.r + q__2.r, q__1.i = temp.i + q__2.i;
temp.r = q__1.r, temp.i = q__1.i;
// L260:
}
if (beta->r == 0.f && beta->i == 0.f) {
i__3 = i__ + j * c_dim1;
q__1.r = alpha->r * temp.r - alpha->i * temp.i,
q__1.i = alpha->r * temp.i + alpha->i *
temp.r;
c__[i__3].r = q__1.r, c__[i__3].i = q__1.i;
} else {
i__3 = i__ + j * c_dim1;
q__2.r = alpha->r * temp.r - alpha->i * temp.i,
q__2.i = alpha->r * temp.i + alpha->i *
temp.r;
i__4 = i__ + j * c_dim1;
q__3.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, q__3.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
q__1.r = q__2.r + q__3.r, q__1.i = q__2.i + q__3.i;
c__[i__3].r = q__1.r, c__[i__3].i = q__1.i;
}
// L270:
}
// L280:
}
} else {
//
// Form C := alpha*A**H*B**T + beta*C
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp.r = 0.f, temp.i = 0.f;
i__3 = *k;
for (l = 1; l <= i__3; ++l) {
r_cnjg(&q__3, &a[l + i__ * a_dim1]);
i__4 = j + l * b_dim1;
q__2.r = q__3.r * b[i__4].r - q__3.i * b[i__4].i,
q__2.i = q__3.r * b[i__4].i + q__3.i * b[i__4]
.r;
q__1.r = temp.r + q__2.r, q__1.i = temp.i + q__2.i;
temp.r = q__1.r, temp.i = q__1.i;
// L290:
}
if (beta->r == 0.f && beta->i == 0.f) {
i__3 = i__ + j * c_dim1;
q__1.r = alpha->r * temp.r - alpha->i * temp.i,
q__1.i = alpha->r * temp.i + alpha->i *
temp.r;
c__[i__3].r = q__1.r, c__[i__3].i = q__1.i;
} else {
i__3 = i__ + j * c_dim1;
q__2.r = alpha->r * temp.r - alpha->i * temp.i,
q__2.i = alpha->r * temp.i + alpha->i *
temp.r;
i__4 = i__ + j * c_dim1;
q__3.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, q__3.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
q__1.r = q__2.r + q__3.r, q__1.i = q__2.i + q__3.i;
c__[i__3].r = q__1.r, c__[i__3].i = q__1.i;
}
// L300:
}
// L310:
}
}
} else {
if (conjb) {
//
// Form C := alpha*A**T*B**H + beta*C
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp.r = 0.f, temp.i = 0.f;
i__3 = *k;
for (l = 1; l <= i__3; ++l) {
i__4 = l + i__ * a_dim1;
r_cnjg(&q__3, &b[j + l * b_dim1]);
q__2.r = a[i__4].r * q__3.r - a[i__4].i * q__3.i,
q__2.i = a[i__4].r * q__3.i + a[i__4].i *
q__3.r;
q__1.r = temp.r + q__2.r, q__1.i = temp.i + q__2.i;
temp.r = q__1.r, temp.i = q__1.i;
// L320:
}
if (beta->r == 0.f && beta->i == 0.f) {
i__3 = i__ + j * c_dim1;
q__1.r = alpha->r * temp.r - alpha->i * temp.i,
q__1.i = alpha->r * temp.i + alpha->i *
temp.r;
c__[i__3].r = q__1.r, c__[i__3].i = q__1.i;
} else {
i__3 = i__ + j * c_dim1;
q__2.r = alpha->r * temp.r - alpha->i * temp.i,
q__2.i = alpha->r * temp.i + alpha->i *
temp.r;
i__4 = i__ + j * c_dim1;
q__3.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, q__3.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
q__1.r = q__2.r + q__3.r, q__1.i = q__2.i + q__3.i;
c__[i__3].r = q__1.r, c__[i__3].i = q__1.i;
}
// L330:
}
// L340:
}
} else {
//
// Form C := alpha*A**T*B**T + beta*C
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp.r = 0.f, temp.i = 0.f;
i__3 = *k;
for (l = 1; l <= i__3; ++l) {
i__4 = l + i__ * a_dim1;
i__5 = j + l * b_dim1;
q__2.r = a[i__4].r * b[i__5].r - a[i__4].i * b[i__5]
.i, q__2.i = a[i__4].r * b[i__5].i + a[i__4]
.i * b[i__5].r;
q__1.r = temp.r + q__2.r, q__1.i = temp.i + q__2.i;
temp.r = q__1.r, temp.i = q__1.i;
// L350:
}
if (beta->r == 0.f && beta->i == 0.f) {
i__3 = i__ + j * c_dim1;
q__1.r = alpha->r * temp.r - alpha->i * temp.i,
q__1.i = alpha->r * temp.i + alpha->i *
temp.r;
c__[i__3].r = q__1.r, c__[i__3].i = q__1.i;
} else {
i__3 = i__ + j * c_dim1;
q__2.r = alpha->r * temp.r - alpha->i * temp.i,
q__2.i = alpha->r * temp.i + alpha->i *
temp.r;
i__4 = i__ + j * c_dim1;
q__3.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, q__3.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
q__1.r = q__2.r + q__3.r, q__1.i = q__2.i + q__3.i;
c__[i__3].r = q__1.r, c__[i__3].i = q__1.i;
}
// L360:
}
// L370:
}
}
}
return 0;
//
// End of CGEMM .
//
} // cgemm_
-171
View File
@@ -1,171 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DCOPY
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
// Definition:
// ===========
//
// SUBROUTINE DCOPY(N,DX,INCX,DY,INCY)
//
// .. Scalar Arguments ..
// INTEGER INCX,INCY,N
// ..
// .. Array Arguments ..
// DOUBLE PRECISION DX(*),DY(*)
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DCOPY copies a vector, x, to a vector, y.
//> uses unrolled loops for increments equal to 1.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> number of elements in input vector(s)
//> \endverbatim
//>
//> \param[in] DX
//> \verbatim
//> DX is DOUBLE PRECISION array, dimension ( 1 + ( N - 1 )*abs( INCX ) )
//> \endverbatim
//>
//> \param[in] INCX
//> \verbatim
//> INCX is INTEGER
//> storage spacing between elements of DX
//> \endverbatim
//>
//> \param[out] DY
//> \verbatim
//> DY is DOUBLE PRECISION array, dimension ( 1 + ( N - 1 )*abs( INCY ) )
//> \endverbatim
//>
//> \param[in] INCY
//> \verbatim
//> INCY is INTEGER
//> storage spacing between elements of DY
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date November 2017
//
//> \ingroup double_blas_level1
//
//> \par Further Details:
// =====================
//>
//> \verbatim
//>
//> jack dongarra, linpack, 3/11/78.
//> modified 12/3/93, array(1) declarations changed to array(*)
//> \endverbatim
//>
// =====================================================================
/* Subroutine */ int dcopy_(int *n, double *dx, int *incx, double *dy, int *
incy)
{
// System generated locals
int i__1;
// Local variables
int i__, m, ix, iy, mp1;
//
// -- Reference BLAS level1 routine (version 3.8.0) --
// -- Reference BLAS is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// November 2017
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Local Scalars ..
// ..
// .. Intrinsic Functions ..
// ..
// Parameter adjustments
--dy;
--dx;
// Function Body
if (*n <= 0) {
return 0;
}
if (*incx == 1 && *incy == 1) {
//
// code for both increments equal to 1
//
//
// clean-up loop
//
m = *n % 7;
if (m != 0) {
i__1 = m;
for (i__ = 1; i__ <= i__1; ++i__) {
dy[i__] = dx[i__];
}
if (*n < 7) {
return 0;
}
}
mp1 = m + 1;
i__1 = *n;
for (i__ = mp1; i__ <= i__1; i__ += 7) {
dy[i__] = dx[i__];
dy[i__ + 1] = dx[i__ + 1];
dy[i__ + 2] = dx[i__ + 2];
dy[i__ + 3] = dx[i__ + 3];
dy[i__ + 4] = dx[i__ + 4];
dy[i__ + 5] = dx[i__ + 5];
dy[i__ + 6] = dx[i__ + 6];
}
} else {
//
// code for unequal increments or equal increments
// not equal to 1
//
ix = 1;
iy = 1;
if (*incx < 0) {
ix = (-(*n) + 1) * *incx + 1;
}
if (*incy < 0) {
iy = (-(*n) + 1) * *incy + 1;
}
i__1 = *n;
for (i__ = 1; i__ <= i__1; ++i__) {
dy[iy] = dx[ix];
ix += *incx;
iy += *incy;
}
}
return 0;
} // dcopy_
-172
View File
@@ -1,172 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DDOT
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
// Definition:
// ===========
//
// DOUBLE PRECISION FUNCTION DDOT(N,DX,INCX,DY,INCY)
//
// .. Scalar Arguments ..
// INTEGER INCX,INCY,N
// ..
// .. Array Arguments ..
// DOUBLE PRECISION DX(*),DY(*)
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DDOT forms the dot product of two vectors.
//> uses unrolled loops for increments equal to one.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> number of elements in input vector(s)
//> \endverbatim
//>
//> \param[in] DX
//> \verbatim
//> DX is DOUBLE PRECISION array, dimension ( 1 + ( N - 1 )*abs( INCX ) )
//> \endverbatim
//>
//> \param[in] INCX
//> \verbatim
//> INCX is INTEGER
//> storage spacing between elements of DX
//> \endverbatim
//>
//> \param[in] DY
//> \verbatim
//> DY is DOUBLE PRECISION array, dimension ( 1 + ( N - 1 )*abs( INCY ) )
//> \endverbatim
//>
//> \param[in] INCY
//> \verbatim
//> INCY is INTEGER
//> storage spacing between elements of DY
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date November 2017
//
//> \ingroup double_blas_level1
//
//> \par Further Details:
// =====================
//>
//> \verbatim
//>
//> jack dongarra, linpack, 3/11/78.
//> modified 12/3/93, array(1) declarations changed to array(*)
//> \endverbatim
//>
// =====================================================================
double ddot_(int *n, double *dx, int *incx, double *dy, int *incy)
{
// System generated locals
int i__1;
double ret_val;
// Local variables
int i__, m, ix, iy, mp1;
double dtemp;
//
// -- Reference BLAS level1 routine (version 3.8.0) --
// -- Reference BLAS is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// November 2017
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Local Scalars ..
// ..
// .. Intrinsic Functions ..
// ..
// Parameter adjustments
--dy;
--dx;
// Function Body
ret_val = 0.;
dtemp = 0.;
if (*n <= 0) {
return ret_val;
}
if (*incx == 1 && *incy == 1) {
//
// code for both increments equal to 1
//
//
// clean-up loop
//
m = *n % 5;
if (m != 0) {
i__1 = m;
for (i__ = 1; i__ <= i__1; ++i__) {
dtemp += dx[i__] * dy[i__];
}
if (*n < 5) {
ret_val = dtemp;
return ret_val;
}
}
mp1 = m + 1;
i__1 = *n;
for (i__ = mp1; i__ <= i__1; i__ += 5) {
dtemp = dtemp + dx[i__] * dy[i__] + dx[i__ + 1] * dy[i__ + 1] +
dx[i__ + 2] * dy[i__ + 2] + dx[i__ + 3] * dy[i__ + 3] +
dx[i__ + 4] * dy[i__ + 4];
}
} else {
//
// code for unequal increments or equal increments
// not equal to 1
//
ix = 1;
iy = 1;
if (*incx < 0) {
ix = (-(*n) + 1) * *incx + 1;
}
if (*incy < 0) {
iy = (-(*n) + 1) * *incy + 1;
}
i__1 = *n;
for (i__ = 1; i__ <= i__1; ++i__) {
dtemp += dx[ix] * dy[iy];
ix += *incx;
iy += *incy;
}
}
ret_val = dtemp;
return ret_val;
} // ddot_
-14369
View File
File diff suppressed because it is too large Load Diff
-444
View File
@@ -1,444 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DGEMM
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
// Definition:
// ===========
//
// SUBROUTINE DGEMM(TRANSA,TRANSB,M,N,K,ALPHA,A,LDA,B,LDB,BETA,C,LDC)
//
// .. Scalar Arguments ..
// DOUBLE PRECISION ALPHA,BETA
// INTEGER K,LDA,LDB,LDC,M,N
// CHARACTER TRANSA,TRANSB
// ..
// .. Array Arguments ..
// DOUBLE PRECISION A(LDA,*),B(LDB,*),C(LDC,*)
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DGEMM performs one of the matrix-matrix operations
//>
//> C := alpha*op( A )*op( B ) + beta*C,
//>
//> where op( X ) is one of
//>
//> op( X ) = X or op( X ) = X**T,
//>
//> alpha and beta are scalars, and A, B and C are matrices, with op( A )
//> an m by k matrix, op( B ) a k by n matrix and C an m by n matrix.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] TRANSA
//> \verbatim
//> TRANSA is CHARACTER*1
//> On entry, TRANSA specifies the form of op( A ) to be used in
//> the matrix multiplication as follows:
//>
//> TRANSA = 'N' or 'n', op( A ) = A.
//>
//> TRANSA = 'T' or 't', op( A ) = A**T.
//>
//> TRANSA = 'C' or 'c', op( A ) = A**T.
//> \endverbatim
//>
//> \param[in] TRANSB
//> \verbatim
//> TRANSB is CHARACTER*1
//> On entry, TRANSB specifies the form of op( B ) to be used in
//> the matrix multiplication as follows:
//>
//> TRANSB = 'N' or 'n', op( B ) = B.
//>
//> TRANSB = 'T' or 't', op( B ) = B**T.
//>
//> TRANSB = 'C' or 'c', op( B ) = B**T.
//> \endverbatim
//>
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> On entry, M specifies the number of rows of the matrix
//> op( A ) and of the matrix C. M must be at least zero.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> On entry, N specifies the number of columns of the matrix
//> op( B ) and the number of columns of the matrix C. N must be
//> at least zero.
//> \endverbatim
//>
//> \param[in] K
//> \verbatim
//> K is INTEGER
//> On entry, K specifies the number of columns of the matrix
//> op( A ) and the number of rows of the matrix op( B ). K must
//> be at least zero.
//> \endverbatim
//>
//> \param[in] ALPHA
//> \verbatim
//> ALPHA is DOUBLE PRECISION.
//> On entry, ALPHA specifies the scalar alpha.
//> \endverbatim
//>
//> \param[in] A
//> \verbatim
//> A is DOUBLE PRECISION array, dimension ( LDA, ka ), where ka is
//> k when TRANSA = 'N' or 'n', and is m otherwise.
//> Before entry with TRANSA = 'N' or 'n', the leading m by k
//> part of the array A must contain the matrix A, otherwise
//> the leading k by m part of the array A must contain the
//> matrix A.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> On entry, LDA specifies the first dimension of A as declared
//> in the calling (sub) program. When TRANSA = 'N' or 'n' then
//> LDA must be at least max( 1, m ), otherwise LDA must be at
//> least max( 1, k ).
//> \endverbatim
//>
//> \param[in] B
//> \verbatim
//> B is DOUBLE PRECISION array, dimension ( LDB, kb ), where kb is
//> n when TRANSB = 'N' or 'n', and is k otherwise.
//> Before entry with TRANSB = 'N' or 'n', the leading k by n
//> part of the array B must contain the matrix B, otherwise
//> the leading n by k part of the array B must contain the
//> matrix B.
//> \endverbatim
//>
//> \param[in] LDB
//> \verbatim
//> LDB is INTEGER
//> On entry, LDB specifies the first dimension of B as declared
//> in the calling (sub) program. When TRANSB = 'N' or 'n' then
//> LDB must be at least max( 1, k ), otherwise LDB must be at
//> least max( 1, n ).
//> \endverbatim
//>
//> \param[in] BETA
//> \verbatim
//> BETA is DOUBLE PRECISION.
//> On entry, BETA specifies the scalar beta. When BETA is
//> supplied as zero then C need not be set on input.
//> \endverbatim
//>
//> \param[in,out] C
//> \verbatim
//> C is DOUBLE PRECISION array, dimension ( LDC, N )
//> Before entry, the leading m by n part of the array C must
//> contain the matrix C, except when beta is zero, in which
//> case C need not be set on entry.
//> On exit, the array C is overwritten by the m by n matrix
//> ( alpha*op( A )*op( B ) + beta*C ).
//> \endverbatim
//>
//> \param[in] LDC
//> \verbatim
//> LDC is INTEGER
//> On entry, LDC specifies the first dimension of C as declared
//> in the calling (sub) program. LDC must be at least
//> max( 1, m ).
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup double_blas_level3
//
//> \par Further Details:
// =====================
//>
//> \verbatim
//>
//> Level 3 Blas routine.
//>
//> -- Written on 8-February-1989.
//> Jack Dongarra, Argonne National Laboratory.
//> Iain Duff, AERE Harwell.
//> Jeremy Du Croz, Numerical Algorithms Group Ltd.
//> Sven Hammarling, Numerical Algorithms Group Ltd.
//> \endverbatim
//>
// =====================================================================
/* Subroutine */ int dgemm_(char *transa, char *transb, int *m, int *n, int *
k, double *alpha, double *a, int *lda, double *b, int *ldb, double *
beta, double *c__, int *ldc)
{
// System generated locals
int a_dim1, a_offset, b_dim1, b_offset, c_dim1, c_offset, i__1, i__2,
i__3;
// Local variables
int i__, j, l, info;
int nota, notb;
double temp;
int ncola;
extern int lsame_(char *, char *);
int nrowa, nrowb;
extern /* Subroutine */ int xerbla_(char *, int *);
//
// -- Reference BLAS level3 routine (version 3.7.0) --
// -- Reference BLAS is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. External Functions ..
// ..
// .. External Subroutines ..
// ..
// .. Intrinsic Functions ..
// ..
// .. Local Scalars ..
// ..
// .. Parameters ..
// ..
//
// Set NOTA and NOTB as true if A and B respectively are not
// transposed and set NROWA, NCOLA and NROWB as the number of rows
// and columns of A and the number of rows of B respectively.
//
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
b_dim1 = *ldb;
b_offset = 1 + b_dim1;
b -= b_offset;
c_dim1 = *ldc;
c_offset = 1 + c_dim1;
c__ -= c_offset;
// Function Body
nota = lsame_(transa, "N");
notb = lsame_(transb, "N");
if (nota) {
nrowa = *m;
ncola = *k;
} else {
nrowa = *k;
ncola = *m;
}
if (notb) {
nrowb = *k;
} else {
nrowb = *n;
}
//
// Test the input parameters.
//
info = 0;
if (! nota && ! lsame_(transa, "C") && ! lsame_(transa, "T")) {
info = 1;
} else if (! notb && ! lsame_(transb, "C") && ! lsame_(transb, "T")) {
info = 2;
} else if (*m < 0) {
info = 3;
} else if (*n < 0) {
info = 4;
} else if (*k < 0) {
info = 5;
} else if (*lda < max(1,nrowa)) {
info = 8;
} else if (*ldb < max(1,nrowb)) {
info = 10;
} else if (*ldc < max(1,*m)) {
info = 13;
}
if (info != 0) {
xerbla_("DGEMM ", &info);
return 0;
}
//
// Quick return if possible.
//
if (*m == 0 || *n == 0 || (*alpha == 0. || *k == 0) && *beta == 1.) {
return 0;
}
//
// And if alpha.eq.zero.
//
if (*alpha == 0.) {
if (*beta == 0.) {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[i__ + j * c_dim1] = 0.;
// L10:
}
// L20:
}
} else {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[i__ + j * c_dim1] = *beta * c__[i__ + j * c_dim1];
// L30:
}
// L40:
}
}
return 0;
}
//
// Start the operations.
//
if (notb) {
if (nota) {
//
// Form C := alpha*A*B + beta*C.
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
if (*beta == 0.) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[i__ + j * c_dim1] = 0.;
// L50:
}
} else if (*beta != 1.) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[i__ + j * c_dim1] = *beta * c__[i__ + j * c_dim1];
// L60:
}
}
i__2 = *k;
for (l = 1; l <= i__2; ++l) {
temp = *alpha * b[l + j * b_dim1];
i__3 = *m;
for (i__ = 1; i__ <= i__3; ++i__) {
c__[i__ + j * c_dim1] += temp * a[i__ + l * a_dim1];
// L70:
}
// L80:
}
// L90:
}
} else {
//
// Form C := alpha*A**T*B + beta*C
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp = 0.;
i__3 = *k;
for (l = 1; l <= i__3; ++l) {
temp += a[l + i__ * a_dim1] * b[l + j * b_dim1];
// L100:
}
if (*beta == 0.) {
c__[i__ + j * c_dim1] = *alpha * temp;
} else {
c__[i__ + j * c_dim1] = *alpha * temp + *beta * c__[
i__ + j * c_dim1];
}
// L110:
}
// L120:
}
}
} else {
if (nota) {
//
// Form C := alpha*A*B**T + beta*C
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
if (*beta == 0.) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[i__ + j * c_dim1] = 0.;
// L130:
}
} else if (*beta != 1.) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[i__ + j * c_dim1] = *beta * c__[i__ + j * c_dim1];
// L140:
}
}
i__2 = *k;
for (l = 1; l <= i__2; ++l) {
temp = *alpha * b[j + l * b_dim1];
i__3 = *m;
for (i__ = 1; i__ <= i__3; ++i__) {
c__[i__ + j * c_dim1] += temp * a[i__ + l * a_dim1];
// L150:
}
// L160:
}
// L170:
}
} else {
//
// Form C := alpha*A**T*B**T + beta*C
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp = 0.;
i__3 = *k;
for (l = 1; l <= i__3; ++l) {
temp += a[l + i__ * a_dim1] * b[j + l * b_dim1];
// L180:
}
if (*beta == 0.) {
c__[i__ + j * c_dim1] = *alpha * temp;
} else {
c__[i__ + j * c_dim1] = *alpha * temp + *beta * c__[
i__ + j * c_dim1];
}
// L190:
}
// L200:
}
}
}
return 0;
//
// End of DGEMM .
//
} // dgemm_
-370
View File
@@ -1,370 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DGEMV
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
// Definition:
// ===========
//
// SUBROUTINE DGEMV(TRANS,M,N,ALPHA,A,LDA,X,INCX,BETA,Y,INCY)
//
// .. Scalar Arguments ..
// DOUBLE PRECISION ALPHA,BETA
// INTEGER INCX,INCY,LDA,M,N
// CHARACTER TRANS
// ..
// .. Array Arguments ..
// DOUBLE PRECISION A(LDA,*),X(*),Y(*)
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DGEMV performs one of the matrix-vector operations
//>
//> y := alpha*A*x + beta*y, or y := alpha*A**T*x + beta*y,
//>
//> where alpha and beta are scalars, x and y are vectors and A is an
//> m by n matrix.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] TRANS
//> \verbatim
//> TRANS is CHARACTER*1
//> On entry, TRANS specifies the operation to be performed as
//> follows:
//>
//> TRANS = 'N' or 'n' y := alpha*A*x + beta*y.
//>
//> TRANS = 'T' or 't' y := alpha*A**T*x + beta*y.
//>
//> TRANS = 'C' or 'c' y := alpha*A**T*x + beta*y.
//> \endverbatim
//>
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> On entry, M specifies the number of rows of the matrix A.
//> M must be at least zero.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> On entry, N specifies the number of columns of the matrix A.
//> N must be at least zero.
//> \endverbatim
//>
//> \param[in] ALPHA
//> \verbatim
//> ALPHA is DOUBLE PRECISION.
//> On entry, ALPHA specifies the scalar alpha.
//> \endverbatim
//>
//> \param[in] A
//> \verbatim
//> A is DOUBLE PRECISION array, dimension ( LDA, N )
//> Before entry, the leading m by n part of the array A must
//> contain the matrix of coefficients.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> On entry, LDA specifies the first dimension of A as declared
//> in the calling (sub) program. LDA must be at least
//> max( 1, m ).
//> \endverbatim
//>
//> \param[in] X
//> \verbatim
//> X is DOUBLE PRECISION array, dimension at least
//> ( 1 + ( n - 1 )*abs( INCX ) ) when TRANS = 'N' or 'n'
//> and at least
//> ( 1 + ( m - 1 )*abs( INCX ) ) otherwise.
//> Before entry, the incremented array X must contain the
//> vector x.
//> \endverbatim
//>
//> \param[in] INCX
//> \verbatim
//> INCX is INTEGER
//> On entry, INCX specifies the increment for the elements of
//> X. INCX must not be zero.
//> \endverbatim
//>
//> \param[in] BETA
//> \verbatim
//> BETA is DOUBLE PRECISION.
//> On entry, BETA specifies the scalar beta. When BETA is
//> supplied as zero then Y need not be set on input.
//> \endverbatim
//>
//> \param[in,out] Y
//> \verbatim
//> Y is DOUBLE PRECISION array, dimension at least
//> ( 1 + ( m - 1 )*abs( INCY ) ) when TRANS = 'N' or 'n'
//> and at least
//> ( 1 + ( n - 1 )*abs( INCY ) ) otherwise.
//> Before entry with BETA non-zero, the incremented array Y
//> must contain the vector y. On exit, Y is overwritten by the
//> updated vector y.
//> \endverbatim
//>
//> \param[in] INCY
//> \verbatim
//> INCY is INTEGER
//> On entry, INCY specifies the increment for the elements of
//> Y. INCY must not be zero.
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup double_blas_level2
//
//> \par Further Details:
// =====================
//>
//> \verbatim
//>
//> Level 2 Blas routine.
//> The vector and matrix arguments are not referenced when N = 0, or M = 0
//>
//> -- Written on 22-October-1986.
//> Jack Dongarra, Argonne National Lab.
//> Jeremy Du Croz, Nag Central Office.
//> Sven Hammarling, Nag Central Office.
//> Richard Hanson, Sandia National Labs.
//> \endverbatim
//>
// =====================================================================
/* Subroutine */ int dgemv_(char *trans, int *m, int *n, double *alpha,
double *a, int *lda, double *x, int *incx, double *beta, double *y,
int *incy)
{
// System generated locals
int a_dim1, a_offset, i__1, i__2;
// Local variables
int i__, j, ix, iy, jx, jy, kx, ky, info;
double temp;
int lenx, leny;
extern int lsame_(char *, char *);
extern /* Subroutine */ int xerbla_(char *, int *);
//
// -- Reference BLAS level2 routine (version 3.7.0) --
// -- Reference BLAS is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. External Functions ..
// ..
// .. External Subroutines ..
// ..
// .. Intrinsic Functions ..
// ..
//
// Test the input parameters.
//
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
--x;
--y;
// Function Body
info = 0;
if (! lsame_(trans, "N") && ! lsame_(trans, "T") && ! lsame_(trans, "C"))
{
info = 1;
} else if (*m < 0) {
info = 2;
} else if (*n < 0) {
info = 3;
} else if (*lda < max(1,*m)) {
info = 6;
} else if (*incx == 0) {
info = 8;
} else if (*incy == 0) {
info = 11;
}
if (info != 0) {
xerbla_("DGEMV ", &info);
return 0;
}
//
// Quick return if possible.
//
if (*m == 0 || *n == 0 || *alpha == 0. && *beta == 1.) {
return 0;
}
//
// Set LENX and LENY, the lengths of the vectors x and y, and set
// up the start points in X and Y.
//
if (lsame_(trans, "N")) {
lenx = *n;
leny = *m;
} else {
lenx = *m;
leny = *n;
}
if (*incx > 0) {
kx = 1;
} else {
kx = 1 - (lenx - 1) * *incx;
}
if (*incy > 0) {
ky = 1;
} else {
ky = 1 - (leny - 1) * *incy;
}
//
// Start the operations. In this version the elements of A are
// accessed sequentially with one pass through A.
//
// First form y := beta*y.
//
if (*beta != 1.) {
if (*incy == 1) {
if (*beta == 0.) {
i__1 = leny;
for (i__ = 1; i__ <= i__1; ++i__) {
y[i__] = 0.;
// L10:
}
} else {
i__1 = leny;
for (i__ = 1; i__ <= i__1; ++i__) {
y[i__] = *beta * y[i__];
// L20:
}
}
} else {
iy = ky;
if (*beta == 0.) {
i__1 = leny;
for (i__ = 1; i__ <= i__1; ++i__) {
y[iy] = 0.;
iy += *incy;
// L30:
}
} else {
i__1 = leny;
for (i__ = 1; i__ <= i__1; ++i__) {
y[iy] = *beta * y[iy];
iy += *incy;
// L40:
}
}
}
}
if (*alpha == 0.) {
return 0;
}
if (lsame_(trans, "N")) {
//
// Form y := alpha*A*x + y.
//
jx = kx;
if (*incy == 1) {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
temp = *alpha * x[jx];
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
y[i__] += temp * a[i__ + j * a_dim1];
// L50:
}
jx += *incx;
// L60:
}
} else {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
temp = *alpha * x[jx];
iy = ky;
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
y[iy] += temp * a[i__ + j * a_dim1];
iy += *incy;
// L70:
}
jx += *incx;
// L80:
}
}
} else {
//
// Form y := alpha*A**T*x + y.
//
jy = ky;
if (*incx == 1) {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
temp = 0.;
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp += a[i__ + j * a_dim1] * x[i__];
// L90:
}
y[jy] += *alpha * temp;
jy += *incy;
// L100:
}
} else {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
temp = 0.;
ix = kx;
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp += a[i__ + j * a_dim1] * x[ix];
ix += *incx;
// L110:
}
y[jy] += *alpha * temp;
jy += *incy;
// L120:
}
}
}
return 0;
//
// End of DGEMV .
//
} // dgemv_
-18599
View File
File diff suppressed because it is too large Load Diff
-186
View File
@@ -1,186 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DISNAN tests input for NaN.
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DISNAN + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/disnan.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/disnan.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/disnan.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// LOGICAL FUNCTION DISNAN( DIN )
//
// .. Scalar Arguments ..
// DOUBLE PRECISION, INTENT(IN) :: DIN
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DISNAN returns .TRUE. if its argument is NaN, and .FALSE.
//> otherwise. To be replaced by the Fortran 2003 intrinsic in the
//> future.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] DIN
//> \verbatim
//> DIN is DOUBLE PRECISION
//> Input to test for NaN.
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date June 2017
//
//> \ingroup OTHERauxiliary
//
// =====================================================================
int disnan_(double *din)
{
// System generated locals
int ret_val;
// Local variables
extern int dlaisnan_(double *, double *);
//
// -- LAPACK auxiliary routine (version 3.7.1) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// June 2017
//
// .. Scalar Arguments ..
// ..
//
// =====================================================================
//
// .. External Functions ..
// ..
// .. Executable Statements ..
ret_val = dlaisnan_(din, din);
return ret_val;
} // disnan_
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
//> \brief \b DLAISNAN tests input for NaN by comparing two arguments for inequality.
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DLAISNAN + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dlaisnan.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dlaisnan.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dlaisnan.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// LOGICAL FUNCTION DLAISNAN( DIN1, DIN2 )
//
// .. Scalar Arguments ..
// DOUBLE PRECISION, INTENT(IN) :: DIN1, DIN2
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> This routine is not for general use. It exists solely to avoid
//> over-optimization in DISNAN.
//>
//> DLAISNAN checks for NaNs by comparing its two arguments for
//> inequality. NaN is the only floating-point value where NaN != NaN
//> returns .TRUE. To check for NaNs, pass the same variable as both
//> arguments.
//>
//> A compiler must assume that the two arguments are
//> not the same variable, and the test will not be optimized away.
//> Interprocedural or whole-program optimization may delete this
//> test. The ISNAN functions will be replaced by the correct
//> Fortran 03 intrinsic once the intrinsic is widely available.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] DIN1
//> \verbatim
//> DIN1 is DOUBLE PRECISION
//> \endverbatim
//>
//> \param[in] DIN2
//> \verbatim
//> DIN2 is DOUBLE PRECISION
//> Two numbers to compare for inequality.
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date June 2017
//
//> \ingroup OTHERauxiliary
//
// =====================================================================
int dlaisnan_(double *din1, double *din2)
{
// System generated locals
int ret_val;
//
// -- LAPACK auxiliary routine (version 3.7.1) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// June 2017
//
// .. Scalar Arguments ..
// ..
//
// =====================================================================
//
// .. Executable Statements ..
ret_val = *din1 != *din2;
return ret_val;
} // dlaisnan_
-184
View File
@@ -1,184 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DLACPY copies all or part of one two-dimensional array to another.
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DLACPY + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dlacpy.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dlacpy.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dlacpy.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// SUBROUTINE DLACPY( UPLO, M, N, A, LDA, B, LDB )
//
// .. Scalar Arguments ..
// CHARACTER UPLO
// INTEGER LDA, LDB, M, N
// ..
// .. Array Arguments ..
// DOUBLE PRECISION A( LDA, * ), B( LDB, * )
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DLACPY copies all or part of a two-dimensional matrix A to another
//> matrix B.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] UPLO
//> \verbatim
//> UPLO is CHARACTER*1
//> Specifies the part of the matrix A to be copied to B.
//> = 'U': Upper triangular part
//> = 'L': Lower triangular part
//> Otherwise: All of the matrix A
//> \endverbatim
//>
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> The number of rows of the matrix A. M >= 0.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> The number of columns of the matrix A. N >= 0.
//> \endverbatim
//>
//> \param[in] A
//> \verbatim
//> A is DOUBLE PRECISION array, dimension (LDA,N)
//> The m by n matrix A. If UPLO = 'U', only the upper triangle
//> or trapezoid is accessed; if UPLO = 'L', only the lower
//> triangle or trapezoid is accessed.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> The leading dimension of the array A. LDA >= max(1,M).
//> \endverbatim
//>
//> \param[out] B
//> \verbatim
//> B is DOUBLE PRECISION array, dimension (LDB,N)
//> On exit, B = A in the locations specified by UPLO.
//> \endverbatim
//>
//> \param[in] LDB
//> \verbatim
//> LDB is INTEGER
//> The leading dimension of the array B. LDB >= max(1,M).
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup OTHERauxiliary
//
// =====================================================================
/* Subroutine */ int dlacpy_(char *uplo, int *m, int *n, double *a, int *lda,
double *b, int *ldb)
{
// System generated locals
int a_dim1, a_offset, b_dim1, b_offset, i__1, i__2;
// Local variables
int i__, j;
extern int lsame_(char *, char *);
//
// -- LAPACK auxiliary routine (version 3.7.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Local Scalars ..
// ..
// .. External Functions ..
// ..
// .. Intrinsic Functions ..
// ..
// .. Executable Statements ..
//
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
b_dim1 = *ldb;
b_offset = 1 + b_dim1;
b -= b_offset;
// Function Body
if (lsame_(uplo, "U")) {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = min(j,*m);
for (i__ = 1; i__ <= i__2; ++i__) {
b[i__ + j * b_dim1] = a[i__ + j * a_dim1];
// L10:
}
// L20:
}
} else if (lsame_(uplo, "L")) {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = j; i__ <= i__2; ++i__) {
b[i__ + j * b_dim1] = a[i__ + j * a_dim1];
// L30:
}
// L40:
}
} else {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
b[i__ + j * b_dim1] = a[i__ + j * a_dim1];
// L50:
}
// L60:
}
}
return 0;
//
// End of DLACPY
//
} // dlacpy_
-367
View File
@@ -1,367 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DCOMBSSQ adds two scaled sum of squares quantities.
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//
// Definition:
// ===========
//
// SUBROUTINE DCOMBSSQ( V1, V2 )
//
// .. Array Arguments ..
// DOUBLE PRECISION V1( 2 ), V2( 2 )
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DCOMBSSQ adds two scaled sum of squares quantities, V1 := V1 + V2.
//> That is,
//>
//> V1_scale**2 * V1_sumsq := V1_scale**2 * V1_sumsq
//> + V2_scale**2 * V2_sumsq
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in,out] V1
//> \verbatim
//> V1 is DOUBLE PRECISION array, dimension (2).
//> The first scaled sum.
//> V1(1) = V1_scale, V1(2) = V1_sumsq.
//> \endverbatim
//>
//> \param[in] V2
//> \verbatim
//> V2 is DOUBLE PRECISION array, dimension (2).
//> The second scaled sum.
//> V2(1) = V2_scale, V2(2) = V2_sumsq.
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date November 2018
//
//> \ingroup OTHERauxiliary
//
// =====================================================================
/* Subroutine */ int dcombssq_(double *v1, double *v2)
{
// System generated locals
double d__1;
//
// -- LAPACK auxiliary routine (version 3.7.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// November 2018
//
// .. Array Arguments ..
// ..
//
//=====================================================================
//
// .. Parameters ..
// ..
// .. Executable Statements ..
//
// Parameter adjustments
--v2;
--v1;
// Function Body
if (v1[1] >= v2[1]) {
if (v1[1] != 0.) {
// Computing 2nd power
d__1 = v2[1] / v1[1];
v1[2] += d__1 * d__1 * v2[2];
}
} else {
// Computing 2nd power
d__1 = v1[1] / v2[1];
v1[2] = v2[2] + d__1 * d__1 * v1[2];
v1[1] = v2[1];
}
return 0;
//
// End of DCOMBSSQ
//
} // dcombssq_
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
//> \brief \b DLANGE returns the value of the 1-norm, Frobenius norm, infinity-norm, or the largest absolute value of any element of a general rectangular matrix.
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DLANGE + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dlange.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dlange.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dlange.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// DOUBLE PRECISION FUNCTION DLANGE( NORM, M, N, A, LDA, WORK )
//
// .. Scalar Arguments ..
// CHARACTER NORM
// INTEGER LDA, M, N
// ..
// .. Array Arguments ..
// DOUBLE PRECISION A( LDA, * ), WORK( * )
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DLANGE returns the value of the one norm, or the Frobenius norm, or
//> the infinity norm, or the element of largest absolute value of a
//> real matrix A.
//> \endverbatim
//>
//> \return DLANGE
//> \verbatim
//>
//> DLANGE = ( max(abs(A(i,j))), NORM = 'M' or 'm'
//> (
//> ( norm1(A), NORM = '1', 'O' or 'o'
//> (
//> ( normI(A), NORM = 'I' or 'i'
//> (
//> ( normF(A), NORM = 'F', 'f', 'E' or 'e'
//>
//> where norm1 denotes the one norm of a matrix (maximum column sum),
//> normI denotes the infinity norm of a matrix (maximum row sum) and
//> normF denotes the Frobenius norm of a matrix (square root of sum of
//> squares). Note that max(abs(A(i,j))) is not a consistent matrix norm.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] NORM
//> \verbatim
//> NORM is CHARACTER*1
//> Specifies the value to be returned in DLANGE as described
//> above.
//> \endverbatim
//>
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> The number of rows of the matrix A. M >= 0. When M = 0,
//> DLANGE is set to zero.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> The number of columns of the matrix A. N >= 0. When N = 0,
//> DLANGE is set to zero.
//> \endverbatim
//>
//> \param[in] A
//> \verbatim
//> A is DOUBLE PRECISION array, dimension (LDA,N)
//> The m by n matrix A.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> The leading dimension of the array A. LDA >= max(M,1).
//> \endverbatim
//>
//> \param[out] WORK
//> \verbatim
//> WORK is DOUBLE PRECISION array, dimension (MAX(1,LWORK)),
//> where LWORK >= M when NORM = 'I'; otherwise, WORK is not
//> referenced.
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup doubleGEauxiliary
//
// =====================================================================
double dlange_(char *norm, int *m, int *n, double *a, int *lda, double *work)
{
// Table of constant values
int c__1 = 1;
// System generated locals
int a_dim1, a_offset, i__1, i__2;
double ret_val, d__1;
// Local variables
extern /* Subroutine */ int dcombssq_(double *, double *);
int i__, j;
double sum, ssq[2], temp;
extern int lsame_(char *, char *);
double value;
extern int disnan_(double *);
extern /* Subroutine */ int dlassq_(int *, double *, int *, double *,
double *);
double colssq[2];
//
// -- LAPACK auxiliary routine (version 3.7.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
//=====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. Local Arrays ..
// ..
// .. External Subroutines ..
// ..
// .. External Functions ..
// ..
// .. Intrinsic Functions ..
// ..
// .. Executable Statements ..
//
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
--work;
// Function Body
if (min(*m,*n) == 0) {
value = 0.;
} else if (lsame_(norm, "M")) {
//
// Find max(abs(A(i,j))).
//
value = 0.;
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp = (d__1 = a[i__ + j * a_dim1], abs(d__1));
if (value < temp || disnan_(&temp)) {
value = temp;
}
// L10:
}
// L20:
}
} else if (lsame_(norm, "O") || *(unsigned char *)norm == '1') {
//
// Find norm1(A).
//
value = 0.;
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
sum = 0.;
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
sum += (d__1 = a[i__ + j * a_dim1], abs(d__1));
// L30:
}
if (value < sum || disnan_(&sum)) {
value = sum;
}
// L40:
}
} else if (lsame_(norm, "I")) {
//
// Find normI(A).
//
i__1 = *m;
for (i__ = 1; i__ <= i__1; ++i__) {
work[i__] = 0.;
// L50:
}
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
work[i__] += (d__1 = a[i__ + j * a_dim1], abs(d__1));
// L60:
}
// L70:
}
value = 0.;
i__1 = *m;
for (i__ = 1; i__ <= i__1; ++i__) {
temp = work[i__];
if (value < temp || disnan_(&temp)) {
value = temp;
}
// L80:
}
} else if (lsame_(norm, "F") || lsame_(norm, "E")) {
//
// Find normF(A).
// SSQ(1) is scale
// SSQ(2) is sum-of-squares
// For better accuracy, sum each column separately.
//
ssq[0] = 0.;
ssq[1] = 1.;
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
colssq[0] = 0.;
colssq[1] = 1.;
dlassq_(m, &a[j * a_dim1 + 1], &c__1, colssq, &colssq[1]);
dcombssq_(ssq, colssq);
// L90:
}
value = ssq[0] * sqrt(ssq[1]);
}
ret_val = value;
return ret_val;
//
// End of DLANGE
//
} // dlange_
-125
View File
@@ -1,125 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DLAPY2 returns sqrt(x2+y2).
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DLAPY2 + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dlapy2.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dlapy2.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dlapy2.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// DOUBLE PRECISION FUNCTION DLAPY2( X, Y )
//
// .. Scalar Arguments ..
// DOUBLE PRECISION X, Y
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DLAPY2 returns sqrt(x**2+y**2), taking care not to cause unnecessary
//> overflow.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] X
//> \verbatim
//> X is DOUBLE PRECISION
//> \endverbatim
//>
//> \param[in] Y
//> \verbatim
//> Y is DOUBLE PRECISION
//> X and Y specify the values x and y.
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date June 2017
//
//> \ingroup OTHERauxiliary
//
// =====================================================================
double dlapy2_(double *x, double *y)
{
// System generated locals
double ret_val, d__1;
// Local variables
int x_is_nan__, y_is_nan__;
double w, z__, xabs, yabs;
extern int disnan_(double *);
//
// -- LAPACK auxiliary routine (version 3.7.1) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// June 2017
//
// .. Scalar Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. External Functions ..
// ..
// .. Intrinsic Functions ..
// ..
// .. Executable Statements ..
//
x_is_nan__ = disnan_(x);
y_is_nan__ = disnan_(y);
if (x_is_nan__) {
ret_val = *x;
}
if (y_is_nan__) {
ret_val = *y;
}
if (! (x_is_nan__ || y_is_nan__)) {
xabs = abs(*x);
yabs = abs(*y);
w = max(xabs,yabs);
z__ = min(xabs,yabs);
if (z__ == 0.) {
ret_val = w;
} else {
// Computing 2nd power
d__1 = z__ / w;
ret_val = w * sqrt(d__1 * d__1 + 1.);
}
}
return ret_val;
//
// End of DLAPY2
//
} // dlapy2_
-768
View File
@@ -1,768 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DGER
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
// Definition:
// ===========
//
// SUBROUTINE DGER(M,N,ALPHA,X,INCX,Y,INCY,A,LDA)
//
// .. Scalar Arguments ..
// DOUBLE PRECISION ALPHA
// INTEGER INCX,INCY,LDA,M,N
// ..
// .. Array Arguments ..
// DOUBLE PRECISION A(LDA,*),X(*),Y(*)
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DGER performs the rank 1 operation
//>
//> A := alpha*x*y**T + A,
//>
//> where alpha is a scalar, x is an m element vector, y is an n element
//> vector and A is an m by n matrix.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> On entry, M specifies the number of rows of the matrix A.
//> M must be at least zero.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> On entry, N specifies the number of columns of the matrix A.
//> N must be at least zero.
//> \endverbatim
//>
//> \param[in] ALPHA
//> \verbatim
//> ALPHA is DOUBLE PRECISION.
//> On entry, ALPHA specifies the scalar alpha.
//> \endverbatim
//>
//> \param[in] X
//> \verbatim
//> X is DOUBLE PRECISION array, dimension at least
//> ( 1 + ( m - 1 )*abs( INCX ) ).
//> Before entry, the incremented array X must contain the m
//> element vector x.
//> \endverbatim
//>
//> \param[in] INCX
//> \verbatim
//> INCX is INTEGER
//> On entry, INCX specifies the increment for the elements of
//> X. INCX must not be zero.
//> \endverbatim
//>
//> \param[in] Y
//> \verbatim
//> Y is DOUBLE PRECISION array, dimension at least
//> ( 1 + ( n - 1 )*abs( INCY ) ).
//> Before entry, the incremented array Y must contain the n
//> element vector y.
//> \endverbatim
//>
//> \param[in] INCY
//> \verbatim
//> INCY is INTEGER
//> On entry, INCY specifies the increment for the elements of
//> Y. INCY must not be zero.
//> \endverbatim
//>
//> \param[in,out] A
//> \verbatim
//> A is DOUBLE PRECISION array, dimension ( LDA, N )
//> Before entry, the leading m by n part of the array A must
//> contain the matrix of coefficients. On exit, A is
//> overwritten by the updated matrix.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> On entry, LDA specifies the first dimension of A as declared
//> in the calling (sub) program. LDA must be at least
//> max( 1, m ).
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup double_blas_level2
//
//> \par Further Details:
// =====================
//>
//> \verbatim
//>
//> Level 2 Blas routine.
//>
//> -- Written on 22-October-1986.
//> Jack Dongarra, Argonne National Lab.
//> Jeremy Du Croz, Nag Central Office.
//> Sven Hammarling, Nag Central Office.
//> Richard Hanson, Sandia National Labs.
//> \endverbatim
//>
// =====================================================================
/* Subroutine */ int dger_(int *m, int *n, double *alpha, double *x, int *
incx, double *y, int *incy, double *a, int *lda)
{
// System generated locals
int a_dim1, a_offset, i__1, i__2;
// Local variables
int i__, j, ix, jy, kx, info;
double temp;
extern /* Subroutine */ int xerbla_(char *, int *);
//
// -- Reference BLAS level2 routine (version 3.7.0) --
// -- Reference BLAS is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. External Subroutines ..
// ..
// .. Intrinsic Functions ..
// ..
//
// Test the input parameters.
//
// Parameter adjustments
--x;
--y;
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
// Function Body
info = 0;
if (*m < 0) {
info = 1;
} else if (*n < 0) {
info = 2;
} else if (*incx == 0) {
info = 5;
} else if (*incy == 0) {
info = 7;
} else if (*lda < max(1,*m)) {
info = 9;
}
if (info != 0) {
xerbla_("DGER ", &info);
return 0;
}
//
// Quick return if possible.
//
if (*m == 0 || *n == 0 || *alpha == 0.) {
return 0;
}
//
// Start the operations. In this version the elements of A are
// accessed sequentially with one pass through A.
//
if (*incy > 0) {
jy = 1;
} else {
jy = 1 - (*n - 1) * *incy;
}
if (*incx == 1) {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
if (y[jy] != 0.) {
temp = *alpha * y[jy];
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
a[i__ + j * a_dim1] += x[i__] * temp;
// L10:
}
}
jy += *incy;
// L20:
}
} else {
if (*incx > 0) {
kx = 1;
} else {
kx = 1 - (*m - 1) * *incx;
}
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
if (y[jy] != 0.) {
temp = *alpha * y[jy];
ix = kx;
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
a[i__ + j * a_dim1] += x[ix] * temp;
ix += *incx;
// L30:
}
}
jy += *incy;
// L40:
}
}
return 0;
//
// End of DGER .
//
} // dger_
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
//> \brief \b DLARF applies an elementary reflector to a general rectangular matrix.
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DLARF + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dlarf.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dlarf.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dlarf.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// SUBROUTINE DLARF( SIDE, M, N, V, INCV, TAU, C, LDC, WORK )
//
// .. Scalar Arguments ..
// CHARACTER SIDE
// INTEGER INCV, LDC, M, N
// DOUBLE PRECISION TAU
// ..
// .. Array Arguments ..
// DOUBLE PRECISION C( LDC, * ), V( * ), WORK( * )
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DLARF applies a real elementary reflector H to a real m by n matrix
//> C, from either the left or the right. H is represented in the form
//>
//> H = I - tau * v * v**T
//>
//> where tau is a real scalar and v is a real vector.
//>
//> If tau = 0, then H is taken to be the unit matrix.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] SIDE
//> \verbatim
//> SIDE is CHARACTER*1
//> = 'L': form H * C
//> = 'R': form C * H
//> \endverbatim
//>
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> The number of rows of the matrix C.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> The number of columns of the matrix C.
//> \endverbatim
//>
//> \param[in] V
//> \verbatim
//> V is DOUBLE PRECISION array, dimension
//> (1 + (M-1)*abs(INCV)) if SIDE = 'L'
//> or (1 + (N-1)*abs(INCV)) if SIDE = 'R'
//> The vector v in the representation of H. V is not used if
//> TAU = 0.
//> \endverbatim
//>
//> \param[in] INCV
//> \verbatim
//> INCV is INTEGER
//> The increment between elements of v. INCV <> 0.
//> \endverbatim
//>
//> \param[in] TAU
//> \verbatim
//> TAU is DOUBLE PRECISION
//> The value tau in the representation of H.
//> \endverbatim
//>
//> \param[in,out] C
//> \verbatim
//> C is DOUBLE PRECISION array, dimension (LDC,N)
//> On entry, the m by n matrix C.
//> On exit, C is overwritten by the matrix H * C if SIDE = 'L',
//> or C * H if SIDE = 'R'.
//> \endverbatim
//>
//> \param[in] LDC
//> \verbatim
//> LDC is INTEGER
//> The leading dimension of the array C. LDC >= max(1,M).
//> \endverbatim
//>
//> \param[out] WORK
//> \verbatim
//> WORK is DOUBLE PRECISION array, dimension
//> (N) if SIDE = 'L'
//> or (M) if SIDE = 'R'
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup doubleOTHERauxiliary
//
// =====================================================================
/* Subroutine */ int dlarf_(char *side, int *m, int *n, double *v, int *incv,
double *tau, double *c__, int *ldc, double *work)
{
// Table of constant values
double c_b4 = 1.;
double c_b5 = 0.;
int c__1 = 1;
// System generated locals
int c_dim1, c_offset;
double d__1;
// Local variables
int i__;
int applyleft;
extern /* Subroutine */ int dger_(int *, int *, double *, double *, int *,
double *, int *, double *, int *);
extern int lsame_(char *, char *);
extern /* Subroutine */ int dgemv_(char *, int *, int *, double *, double
*, int *, double *, int *, double *, double *, int *);
int lastc, lastv;
extern int iladlc_(int *, int *, double *, int *), iladlr_(int *, int *,
double *, int *);
//
// -- LAPACK auxiliary routine (version 3.7.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. External Subroutines ..
// ..
// .. External Functions ..
// ..
// .. Executable Statements ..
//
// Parameter adjustments
--v;
c_dim1 = *ldc;
c_offset = 1 + c_dim1;
c__ -= c_offset;
--work;
// Function Body
applyleft = lsame_(side, "L");
lastv = 0;
lastc = 0;
if (*tau != 0.) {
// Set up variables for scanning V. LASTV begins pointing to the end
// of V.
if (applyleft) {
lastv = *m;
} else {
lastv = *n;
}
if (*incv > 0) {
i__ = (lastv - 1) * *incv + 1;
} else {
i__ = 1;
}
// Look for the last non-zero row in V.
while(lastv > 0 && v[i__] == 0.) {
--lastv;
i__ -= *incv;
}
if (applyleft) {
// Scan for the last non-zero column in C(1:lastv,:).
lastc = iladlc_(&lastv, n, &c__[c_offset], ldc);
} else {
// Scan for the last non-zero row in C(:,1:lastv).
lastc = iladlr_(m, &lastv, &c__[c_offset], ldc);
}
}
// Note that lastc.eq.0 renders the BLAS operations null; no special
// case is needed at this level.
if (applyleft) {
//
// Form H * C
//
if (lastv > 0) {
//
// w(1:lastc,1) := C(1:lastv,1:lastc)**T * v(1:lastv,1)
//
dgemv_("Transpose", &lastv, &lastc, &c_b4, &c__[c_offset], ldc, &
v[1], incv, &c_b5, &work[1], &c__1);
//
// C(1:lastv,1:lastc) := C(...) - v(1:lastv,1) * w(1:lastc,1)**T
//
d__1 = -(*tau);
dger_(&lastv, &lastc, &d__1, &v[1], incv, &work[1], &c__1, &c__[
c_offset], ldc);
}
} else {
//
// Form C * H
//
if (lastv > 0) {
//
// w(1:lastc,1) := C(1:lastc,1:lastv) * v(1:lastv,1)
//
dgemv_("No transpose", &lastc, &lastv, &c_b4, &c__[c_offset], ldc,
&v[1], incv, &c_b5, &work[1], &c__1);
//
// C(1:lastc,1:lastv) := C(...) - w(1:lastc,1) * v(1:lastv,1)**T
//
d__1 = -(*tau);
dger_(&lastc, &lastv, &d__1, &work[1], &c__1, &v[1], incv, &c__[
c_offset], ldc);
}
}
return 0;
//
// End of DLARF
//
} // dlarf_
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
//> \brief \b ILADLC scans a matrix for its last non-zero column.
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download ILADLC + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/iladlc.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/iladlc.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/iladlc.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// INTEGER FUNCTION ILADLC( M, N, A, LDA )
//
// .. Scalar Arguments ..
// INTEGER M, N, LDA
// ..
// .. Array Arguments ..
// DOUBLE PRECISION A( LDA, * )
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> ILADLC scans A for its last non-zero column.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> The number of rows of the matrix A.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> The number of columns of the matrix A.
//> \endverbatim
//>
//> \param[in] A
//> \verbatim
//> A is DOUBLE PRECISION array, dimension (LDA,N)
//> The m by n matrix A.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> The leading dimension of the array A. LDA >= max(1,M).
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup OTHERauxiliary
//
// =====================================================================
int iladlc_(int *m, int *n, double *a, int *lda)
{
// System generated locals
int a_dim1, a_offset, ret_val, i__1;
// Local variables
int i__;
//
// -- LAPACK auxiliary routine (version 3.7.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. Executable Statements ..
//
// Quick test for the common case where one corner is non-zero.
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
// Function Body
if (*n == 0) {
ret_val = *n;
} else if (a[*n * a_dim1 + 1] != 0. || a[*m + *n * a_dim1] != 0.) {
ret_val = *n;
} else {
// Now scan each column from the end, returning with the first non-zero.
for (ret_val = *n; ret_val >= 1; --ret_val) {
i__1 = *m;
for (i__ = 1; i__ <= i__1; ++i__) {
if (a[i__ + ret_val * a_dim1] != 0.) {
return ret_val;
}
}
}
}
return ret_val;
} // iladlc_
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
//> \brief \b ILADLR scans a matrix for its last non-zero row.
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download ILADLR + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/iladlr.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/iladlr.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/iladlr.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// INTEGER FUNCTION ILADLR( M, N, A, LDA )
//
// .. Scalar Arguments ..
// INTEGER M, N, LDA
// ..
// .. Array Arguments ..
// DOUBLE PRECISION A( LDA, * )
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> ILADLR scans A for its last non-zero row.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> The number of rows of the matrix A.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> The number of columns of the matrix A.
//> \endverbatim
//>
//> \param[in] A
//> \verbatim
//> A is DOUBLE PRECISION array, dimension (LDA,N)
//> The m by n matrix A.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> The leading dimension of the array A. LDA >= max(1,M).
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup OTHERauxiliary
//
// =====================================================================
int iladlr_(int *m, int *n, double *a, int *lda)
{
// System generated locals
int a_dim1, a_offset, ret_val, i__1;
// Local variables
int i__, j;
//
// -- LAPACK auxiliary routine (version 3.7.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. Executable Statements ..
//
// Quick test for the common case where one corner is non-zero.
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
// Function Body
if (*m == 0) {
ret_val = *m;
} else if (a[*m + a_dim1] != 0. || a[*m + *n * a_dim1] != 0.) {
ret_val = *m;
} else {
// Scan up each column tracking the last zero row seen.
ret_val = 0;
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__ = *m;
while(a[max(i__,1) + j * a_dim1] == 0. && i__ >= 1) {
--i__;
}
ret_val = max(ret_val,i__);
}
}
return ret_val;
} // iladlr_
-824
View File
@@ -1,824 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DLARFB applies a block reflector or its transpose to a general rectangular matrix.
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DLARFB + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dlarfb.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dlarfb.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dlarfb.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// SUBROUTINE DLARFB( SIDE, TRANS, DIRECT, STOREV, M, N, K, V, LDV,
// T, LDT, C, LDC, WORK, LDWORK )
//
// .. Scalar Arguments ..
// CHARACTER DIRECT, SIDE, STOREV, TRANS
// INTEGER K, LDC, LDT, LDV, LDWORK, M, N
// ..
// .. Array Arguments ..
// DOUBLE PRECISION C( LDC, * ), T( LDT, * ), V( LDV, * ),
// $ WORK( LDWORK, * )
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DLARFB applies a real block reflector H or its transpose H**T to a
//> real m by n matrix C, from either the left or the right.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] SIDE
//> \verbatim
//> SIDE is CHARACTER*1
//> = 'L': apply H or H**T from the Left
//> = 'R': apply H or H**T from the Right
//> \endverbatim
//>
//> \param[in] TRANS
//> \verbatim
//> TRANS is CHARACTER*1
//> = 'N': apply H (No transpose)
//> = 'T': apply H**T (Transpose)
//> \endverbatim
//>
//> \param[in] DIRECT
//> \verbatim
//> DIRECT is CHARACTER*1
//> Indicates how H is formed from a product of elementary
//> reflectors
//> = 'F': H = H(1) H(2) . . . H(k) (Forward)
//> = 'B': H = H(k) . . . H(2) H(1) (Backward)
//> \endverbatim
//>
//> \param[in] STOREV
//> \verbatim
//> STOREV is CHARACTER*1
//> Indicates how the vectors which define the elementary
//> reflectors are stored:
//> = 'C': Columnwise
//> = 'R': Rowwise
//> \endverbatim
//>
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> The number of rows of the matrix C.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> The number of columns of the matrix C.
//> \endverbatim
//>
//> \param[in] K
//> \verbatim
//> K is INTEGER
//> The order of the matrix T (= the number of elementary
//> reflectors whose product defines the block reflector).
//> If SIDE = 'L', M >= K >= 0;
//> if SIDE = 'R', N >= K >= 0.
//> \endverbatim
//>
//> \param[in] V
//> \verbatim
//> V is DOUBLE PRECISION array, dimension
//> (LDV,K) if STOREV = 'C'
//> (LDV,M) if STOREV = 'R' and SIDE = 'L'
//> (LDV,N) if STOREV = 'R' and SIDE = 'R'
//> The matrix V. See Further Details.
//> \endverbatim
//>
//> \param[in] LDV
//> \verbatim
//> LDV is INTEGER
//> The leading dimension of the array V.
//> If STOREV = 'C' and SIDE = 'L', LDV >= max(1,M);
//> if STOREV = 'C' and SIDE = 'R', LDV >= max(1,N);
//> if STOREV = 'R', LDV >= K.
//> \endverbatim
//>
//> \param[in] T
//> \verbatim
//> T is DOUBLE PRECISION array, dimension (LDT,K)
//> The triangular k by k matrix T in the representation of the
//> block reflector.
//> \endverbatim
//>
//> \param[in] LDT
//> \verbatim
//> LDT is INTEGER
//> The leading dimension of the array T. LDT >= K.
//> \endverbatim
//>
//> \param[in,out] C
//> \verbatim
//> C is DOUBLE PRECISION array, dimension (LDC,N)
//> On entry, the m by n matrix C.
//> On exit, C is overwritten by H*C or H**T*C or C*H or C*H**T.
//> \endverbatim
//>
//> \param[in] LDC
//> \verbatim
//> LDC is INTEGER
//> The leading dimension of the array C. LDC >= max(1,M).
//> \endverbatim
//>
//> \param[out] WORK
//> \verbatim
//> WORK is DOUBLE PRECISION array, dimension (LDWORK,K)
//> \endverbatim
//>
//> \param[in] LDWORK
//> \verbatim
//> LDWORK is INTEGER
//> The leading dimension of the array WORK.
//> If SIDE = 'L', LDWORK >= max(1,N);
//> if SIDE = 'R', LDWORK >= max(1,M).
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date June 2013
//
//> \ingroup doubleOTHERauxiliary
//
//> \par Further Details:
// =====================
//>
//> \verbatim
//>
//> The shape of the matrix V and the storage of the vectors which define
//> the H(i) is best illustrated by the following example with n = 5 and
//> k = 3. The elements equal to 1 are not stored; the corresponding
//> array elements are modified but restored on exit. The rest of the
//> array is not used.
//>
//> DIRECT = 'F' and STOREV = 'C': DIRECT = 'F' and STOREV = 'R':
//>
//> V = ( 1 ) V = ( 1 v1 v1 v1 v1 )
//> ( v1 1 ) ( 1 v2 v2 v2 )
//> ( v1 v2 1 ) ( 1 v3 v3 )
//> ( v1 v2 v3 )
//> ( v1 v2 v3 )
//>
//> DIRECT = 'B' and STOREV = 'C': DIRECT = 'B' and STOREV = 'R':
//>
//> V = ( v1 v2 v3 ) V = ( v1 v1 1 )
//> ( v1 v2 v3 ) ( v2 v2 v2 1 )
//> ( 1 v2 v3 ) ( v3 v3 v3 v3 1 )
//> ( 1 v3 )
//> ( 1 )
//> \endverbatim
//>
// =====================================================================
/* Subroutine */ int dlarfb_(char *side, char *trans, char *direct, char *
storev, int *m, int *n, int *k, double *v, int *ldv, double *t, int *
ldt, double *c__, int *ldc, double *work, int *ldwork)
{
// Table of constant values
int c__1 = 1;
double c_b14 = 1.;
double c_b25 = -1.;
// System generated locals
int c_dim1, c_offset, t_dim1, t_offset, v_dim1, v_offset, work_dim1,
work_offset, i__1, i__2;
// Local variables
int i__, j;
extern /* Subroutine */ int dgemm_(char *, char *, int *, int *, int *,
double *, double *, int *, double *, int *, double *, double *,
int *);
extern int lsame_(char *, char *);
extern /* Subroutine */ int dcopy_(int *, double *, int *, double *, int *
), dtrmm_(char *, char *, char *, char *, int *, int *, double *,
double *, int *, double *, int *);
char transt[1+1]={'\0'};
//
// -- LAPACK auxiliary routine (version 3.7.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// June 2013
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. External Functions ..
// ..
// .. External Subroutines ..
// ..
// .. Executable Statements ..
//
// Quick return if possible
//
// Parameter adjustments
v_dim1 = *ldv;
v_offset = 1 + v_dim1;
v -= v_offset;
t_dim1 = *ldt;
t_offset = 1 + t_dim1;
t -= t_offset;
c_dim1 = *ldc;
c_offset = 1 + c_dim1;
c__ -= c_offset;
work_dim1 = *ldwork;
work_offset = 1 + work_dim1;
work -= work_offset;
// Function Body
if (*m <= 0 || *n <= 0) {
return 0;
}
if (lsame_(trans, "N")) {
*(unsigned char *)transt = 'T';
} else {
*(unsigned char *)transt = 'N';
}
if (lsame_(storev, "C")) {
if (lsame_(direct, "F")) {
//
// Let V = ( V1 ) (first K rows)
// ( V2 )
// where V1 is unit lower triangular.
//
if (lsame_(side, "L")) {
//
// Form H * C or H**T * C where C = ( C1 )
// ( C2 )
//
// W := C**T * V = (C1**T * V1 + C2**T * V2) (stored in WORK)
//
// W := C1**T
//
i__1 = *k;
for (j = 1; j <= i__1; ++j) {
dcopy_(n, &c__[j + c_dim1], ldc, &work[j * work_dim1 + 1],
&c__1);
// L10:
}
//
// W := W * V1
//
dtrmm_("Right", "Lower", "No transpose", "Unit", n, k, &c_b14,
&v[v_offset], ldv, &work[work_offset], ldwork);
if (*m > *k) {
//
// W := W + C2**T * V2
//
i__1 = *m - *k;
dgemm_("Transpose", "No transpose", n, k, &i__1, &c_b14, &
c__[*k + 1 + c_dim1], ldc, &v[*k + 1 + v_dim1],
ldv, &c_b14, &work[work_offset], ldwork);
}
//
// W := W * T**T or W * T
//
dtrmm_("Right", "Upper", transt, "Non-unit", n, k, &c_b14, &t[
t_offset], ldt, &work[work_offset], ldwork);
//
// C := C - V * W**T
//
if (*m > *k) {
//
// C2 := C2 - V2 * W**T
//
i__1 = *m - *k;
dgemm_("No transpose", "Transpose", &i__1, n, k, &c_b25, &
v[*k + 1 + v_dim1], ldv, &work[work_offset],
ldwork, &c_b14, &c__[*k + 1 + c_dim1], ldc);
}
//
// W := W * V1**T
//
dtrmm_("Right", "Lower", "Transpose", "Unit", n, k, &c_b14, &
v[v_offset], ldv, &work[work_offset], ldwork);
//
// C1 := C1 - W**T
//
i__1 = *k;
for (j = 1; j <= i__1; ++j) {
i__2 = *n;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[j + i__ * c_dim1] -= work[i__ + j * work_dim1];
// L20:
}
// L30:
}
} else if (lsame_(side, "R")) {
//
// Form C * H or C * H**T where C = ( C1 C2 )
//
// W := C * V = (C1*V1 + C2*V2) (stored in WORK)
//
// W := C1
//
i__1 = *k;
for (j = 1; j <= i__1; ++j) {
dcopy_(m, &c__[j * c_dim1 + 1], &c__1, &work[j *
work_dim1 + 1], &c__1);
// L40:
}
//
// W := W * V1
//
dtrmm_("Right", "Lower", "No transpose", "Unit", m, k, &c_b14,
&v[v_offset], ldv, &work[work_offset], ldwork);
if (*n > *k) {
//
// W := W + C2 * V2
//
i__1 = *n - *k;
dgemm_("No transpose", "No transpose", m, k, &i__1, &
c_b14, &c__[(*k + 1) * c_dim1 + 1], ldc, &v[*k +
1 + v_dim1], ldv, &c_b14, &work[work_offset],
ldwork);
}
//
// W := W * T or W * T**T
//
dtrmm_("Right", "Upper", trans, "Non-unit", m, k, &c_b14, &t[
t_offset], ldt, &work[work_offset], ldwork);
//
// C := C - W * V**T
//
if (*n > *k) {
//
// C2 := C2 - W * V2**T
//
i__1 = *n - *k;
dgemm_("No transpose", "Transpose", m, &i__1, k, &c_b25, &
work[work_offset], ldwork, &v[*k + 1 + v_dim1],
ldv, &c_b14, &c__[(*k + 1) * c_dim1 + 1], ldc);
}
//
// W := W * V1**T
//
dtrmm_("Right", "Lower", "Transpose", "Unit", m, k, &c_b14, &
v[v_offset], ldv, &work[work_offset], ldwork);
//
// C1 := C1 - W
//
i__1 = *k;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[i__ + j * c_dim1] -= work[i__ + j * work_dim1];
// L50:
}
// L60:
}
}
} else {
//
// Let V = ( V1 )
// ( V2 ) (last K rows)
// where V2 is unit upper triangular.
//
if (lsame_(side, "L")) {
//
// Form H * C or H**T * C where C = ( C1 )
// ( C2 )
//
// W := C**T * V = (C1**T * V1 + C2**T * V2) (stored in WORK)
//
// W := C2**T
//
i__1 = *k;
for (j = 1; j <= i__1; ++j) {
dcopy_(n, &c__[*m - *k + j + c_dim1], ldc, &work[j *
work_dim1 + 1], &c__1);
// L70:
}
//
// W := W * V2
//
dtrmm_("Right", "Upper", "No transpose", "Unit", n, k, &c_b14,
&v[*m - *k + 1 + v_dim1], ldv, &work[work_offset],
ldwork);
if (*m > *k) {
//
// W := W + C1**T * V1
//
i__1 = *m - *k;
dgemm_("Transpose", "No transpose", n, k, &i__1, &c_b14, &
c__[c_offset], ldc, &v[v_offset], ldv, &c_b14, &
work[work_offset], ldwork);
}
//
// W := W * T**T or W * T
//
dtrmm_("Right", "Lower", transt, "Non-unit", n, k, &c_b14, &t[
t_offset], ldt, &work[work_offset], ldwork);
//
// C := C - V * W**T
//
if (*m > *k) {
//
// C1 := C1 - V1 * W**T
//
i__1 = *m - *k;
dgemm_("No transpose", "Transpose", &i__1, n, k, &c_b25, &
v[v_offset], ldv, &work[work_offset], ldwork, &
c_b14, &c__[c_offset], ldc);
}
//
// W := W * V2**T
//
dtrmm_("Right", "Upper", "Transpose", "Unit", n, k, &c_b14, &
v[*m - *k + 1 + v_dim1], ldv, &work[work_offset],
ldwork);
//
// C2 := C2 - W**T
//
i__1 = *k;
for (j = 1; j <= i__1; ++j) {
i__2 = *n;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[*m - *k + j + i__ * c_dim1] -= work[i__ + j *
work_dim1];
// L80:
}
// L90:
}
} else if (lsame_(side, "R")) {
//
// Form C * H or C * H**T where C = ( C1 C2 )
//
// W := C * V = (C1*V1 + C2*V2) (stored in WORK)
//
// W := C2
//
i__1 = *k;
for (j = 1; j <= i__1; ++j) {
dcopy_(m, &c__[(*n - *k + j) * c_dim1 + 1], &c__1, &work[
j * work_dim1 + 1], &c__1);
// L100:
}
//
// W := W * V2
//
dtrmm_("Right", "Upper", "No transpose", "Unit", m, k, &c_b14,
&v[*n - *k + 1 + v_dim1], ldv, &work[work_offset],
ldwork);
if (*n > *k) {
//
// W := W + C1 * V1
//
i__1 = *n - *k;
dgemm_("No transpose", "No transpose", m, k, &i__1, &
c_b14, &c__[c_offset], ldc, &v[v_offset], ldv, &
c_b14, &work[work_offset], ldwork);
}
//
// W := W * T or W * T**T
//
dtrmm_("Right", "Lower", trans, "Non-unit", m, k, &c_b14, &t[
t_offset], ldt, &work[work_offset], ldwork);
//
// C := C - W * V**T
//
if (*n > *k) {
//
// C1 := C1 - W * V1**T
//
i__1 = *n - *k;
dgemm_("No transpose", "Transpose", m, &i__1, k, &c_b25, &
work[work_offset], ldwork, &v[v_offset], ldv, &
c_b14, &c__[c_offset], ldc);
}
//
// W := W * V2**T
//
dtrmm_("Right", "Upper", "Transpose", "Unit", m, k, &c_b14, &
v[*n - *k + 1 + v_dim1], ldv, &work[work_offset],
ldwork);
//
// C2 := C2 - W
//
i__1 = *k;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[i__ + (*n - *k + j) * c_dim1] -= work[i__ + j *
work_dim1];
// L110:
}
// L120:
}
}
}
} else if (lsame_(storev, "R")) {
if (lsame_(direct, "F")) {
//
// Let V = ( V1 V2 ) (V1: first K columns)
// where V1 is unit upper triangular.
//
if (lsame_(side, "L")) {
//
// Form H * C or H**T * C where C = ( C1 )
// ( C2 )
//
// W := C**T * V**T = (C1**T * V1**T + C2**T * V2**T) (stored in WORK)
//
// W := C1**T
//
i__1 = *k;
for (j = 1; j <= i__1; ++j) {
dcopy_(n, &c__[j + c_dim1], ldc, &work[j * work_dim1 + 1],
&c__1);
// L130:
}
//
// W := W * V1**T
//
dtrmm_("Right", "Upper", "Transpose", "Unit", n, k, &c_b14, &
v[v_offset], ldv, &work[work_offset], ldwork);
if (*m > *k) {
//
// W := W + C2**T * V2**T
//
i__1 = *m - *k;
dgemm_("Transpose", "Transpose", n, k, &i__1, &c_b14, &
c__[*k + 1 + c_dim1], ldc, &v[(*k + 1) * v_dim1 +
1], ldv, &c_b14, &work[work_offset], ldwork);
}
//
// W := W * T**T or W * T
//
dtrmm_("Right", "Upper", transt, "Non-unit", n, k, &c_b14, &t[
t_offset], ldt, &work[work_offset], ldwork);
//
// C := C - V**T * W**T
//
if (*m > *k) {
//
// C2 := C2 - V2**T * W**T
//
i__1 = *m - *k;
dgemm_("Transpose", "Transpose", &i__1, n, k, &c_b25, &v[(
*k + 1) * v_dim1 + 1], ldv, &work[work_offset],
ldwork, &c_b14, &c__[*k + 1 + c_dim1], ldc);
}
//
// W := W * V1
//
dtrmm_("Right", "Upper", "No transpose", "Unit", n, k, &c_b14,
&v[v_offset], ldv, &work[work_offset], ldwork);
//
// C1 := C1 - W**T
//
i__1 = *k;
for (j = 1; j <= i__1; ++j) {
i__2 = *n;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[j + i__ * c_dim1] -= work[i__ + j * work_dim1];
// L140:
}
// L150:
}
} else if (lsame_(side, "R")) {
//
// Form C * H or C * H**T where C = ( C1 C2 )
//
// W := C * V**T = (C1*V1**T + C2*V2**T) (stored in WORK)
//
// W := C1
//
i__1 = *k;
for (j = 1; j <= i__1; ++j) {
dcopy_(m, &c__[j * c_dim1 + 1], &c__1, &work[j *
work_dim1 + 1], &c__1);
// L160:
}
//
// W := W * V1**T
//
dtrmm_("Right", "Upper", "Transpose", "Unit", m, k, &c_b14, &
v[v_offset], ldv, &work[work_offset], ldwork);
if (*n > *k) {
//
// W := W + C2 * V2**T
//
i__1 = *n - *k;
dgemm_("No transpose", "Transpose", m, k, &i__1, &c_b14, &
c__[(*k + 1) * c_dim1 + 1], ldc, &v[(*k + 1) *
v_dim1 + 1], ldv, &c_b14, &work[work_offset],
ldwork);
}
//
// W := W * T or W * T**T
//
dtrmm_("Right", "Upper", trans, "Non-unit", m, k, &c_b14, &t[
t_offset], ldt, &work[work_offset], ldwork);
//
// C := C - W * V
//
if (*n > *k) {
//
// C2 := C2 - W * V2
//
i__1 = *n - *k;
dgemm_("No transpose", "No transpose", m, &i__1, k, &
c_b25, &work[work_offset], ldwork, &v[(*k + 1) *
v_dim1 + 1], ldv, &c_b14, &c__[(*k + 1) * c_dim1
+ 1], ldc);
}
//
// W := W * V1
//
dtrmm_("Right", "Upper", "No transpose", "Unit", m, k, &c_b14,
&v[v_offset], ldv, &work[work_offset], ldwork);
//
// C1 := C1 - W
//
i__1 = *k;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[i__ + j * c_dim1] -= work[i__ + j * work_dim1];
// L170:
}
// L180:
}
}
} else {
//
// Let V = ( V1 V2 ) (V2: last K columns)
// where V2 is unit lower triangular.
//
if (lsame_(side, "L")) {
//
// Form H * C or H**T * C where C = ( C1 )
// ( C2 )
//
// W := C**T * V**T = (C1**T * V1**T + C2**T * V2**T) (stored in WORK)
//
// W := C2**T
//
i__1 = *k;
for (j = 1; j <= i__1; ++j) {
dcopy_(n, &c__[*m - *k + j + c_dim1], ldc, &work[j *
work_dim1 + 1], &c__1);
// L190:
}
//
// W := W * V2**T
//
dtrmm_("Right", "Lower", "Transpose", "Unit", n, k, &c_b14, &
v[(*m - *k + 1) * v_dim1 + 1], ldv, &work[work_offset]
, ldwork);
if (*m > *k) {
//
// W := W + C1**T * V1**T
//
i__1 = *m - *k;
dgemm_("Transpose", "Transpose", n, k, &i__1, &c_b14, &
c__[c_offset], ldc, &v[v_offset], ldv, &c_b14, &
work[work_offset], ldwork);
}
//
// W := W * T**T or W * T
//
dtrmm_("Right", "Lower", transt, "Non-unit", n, k, &c_b14, &t[
t_offset], ldt, &work[work_offset], ldwork);
//
// C := C - V**T * W**T
//
if (*m > *k) {
//
// C1 := C1 - V1**T * W**T
//
i__1 = *m - *k;
dgemm_("Transpose", "Transpose", &i__1, n, k, &c_b25, &v[
v_offset], ldv, &work[work_offset], ldwork, &
c_b14, &c__[c_offset], ldc);
}
//
// W := W * V2
//
dtrmm_("Right", "Lower", "No transpose", "Unit", n, k, &c_b14,
&v[(*m - *k + 1) * v_dim1 + 1], ldv, &work[
work_offset], ldwork);
//
// C2 := C2 - W**T
//
i__1 = *k;
for (j = 1; j <= i__1; ++j) {
i__2 = *n;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[*m - *k + j + i__ * c_dim1] -= work[i__ + j *
work_dim1];
// L200:
}
// L210:
}
} else if (lsame_(side, "R")) {
//
// Form C * H or C * H' where C = ( C1 C2 )
//
// W := C * V**T = (C1*V1**T + C2*V2**T) (stored in WORK)
//
// W := C2
//
i__1 = *k;
for (j = 1; j <= i__1; ++j) {
dcopy_(m, &c__[(*n - *k + j) * c_dim1 + 1], &c__1, &work[
j * work_dim1 + 1], &c__1);
// L220:
}
//
// W := W * V2**T
//
dtrmm_("Right", "Lower", "Transpose", "Unit", m, k, &c_b14, &
v[(*n - *k + 1) * v_dim1 + 1], ldv, &work[work_offset]
, ldwork);
if (*n > *k) {
//
// W := W + C1 * V1**T
//
i__1 = *n - *k;
dgemm_("No transpose", "Transpose", m, k, &i__1, &c_b14, &
c__[c_offset], ldc, &v[v_offset], ldv, &c_b14, &
work[work_offset], ldwork);
}
//
// W := W * T or W * T**T
//
dtrmm_("Right", "Lower", trans, "Non-unit", m, k, &c_b14, &t[
t_offset], ldt, &work[work_offset], ldwork);
//
// C := C - W * V
//
if (*n > *k) {
//
// C1 := C1 - W * V1
//
i__1 = *n - *k;
dgemm_("No transpose", "No transpose", m, &i__1, k, &
c_b25, &work[work_offset], ldwork, &v[v_offset],
ldv, &c_b14, &c__[c_offset], ldc);
}
//
// W := W * V2
//
dtrmm_("Right", "Lower", "No transpose", "Unit", m, k, &c_b14,
&v[(*n - *k + 1) * v_dim1 + 1], ldv, &work[
work_offset], ldwork);
//
// C1 := C1 - W
//
i__1 = *k;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[i__ + (*n - *k + j) * c_dim1] -= work[i__ + j *
work_dim1];
// L230:
}
// L240:
}
}
}
}
return 0;
//
// End of DLARFB
//
} // dlarfb_
-216
View File
@@ -1,216 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DLARFG generates an elementary reflector (Householder matrix).
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DLARFG + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dlarfg.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dlarfg.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dlarfg.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// SUBROUTINE DLARFG( N, ALPHA, X, INCX, TAU )
//
// .. Scalar Arguments ..
// INTEGER INCX, N
// DOUBLE PRECISION ALPHA, TAU
// ..
// .. Array Arguments ..
// DOUBLE PRECISION X( * )
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DLARFG generates a real elementary reflector H of order n, such
//> that
//>
//> H * ( alpha ) = ( beta ), H**T * H = I.
//> ( x ) ( 0 )
//>
//> where alpha and beta are scalars, and x is an (n-1)-element real
//> vector. H is represented in the form
//>
//> H = I - tau * ( 1 ) * ( 1 v**T ) ,
//> ( v )
//>
//> where tau is a real scalar and v is a real (n-1)-element
//> vector.
//>
//> If the elements of x are all zero, then tau = 0 and H is taken to be
//> the unit matrix.
//>
//> Otherwise 1 <= tau <= 2.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> The order of the elementary reflector.
//> \endverbatim
//>
//> \param[in,out] ALPHA
//> \verbatim
//> ALPHA is DOUBLE PRECISION
//> On entry, the value alpha.
//> On exit, it is overwritten with the value beta.
//> \endverbatim
//>
//> \param[in,out] X
//> \verbatim
//> X is DOUBLE PRECISION array, dimension
//> (1+(N-2)*abs(INCX))
//> On entry, the vector x.
//> On exit, it is overwritten with the vector v.
//> \endverbatim
//>
//> \param[in] INCX
//> \verbatim
//> INCX is INTEGER
//> The increment between elements of X. INCX > 0.
//> \endverbatim
//>
//> \param[out] TAU
//> \verbatim
//> TAU is DOUBLE PRECISION
//> The value tau.
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date November 2017
//
//> \ingroup doubleOTHERauxiliary
//
// =====================================================================
/* Subroutine */ int dlarfg_(int *n, double *alpha, double *x, int *incx,
double *tau)
{
// System generated locals
int i__1;
double d__1;
// Local variables
int j, knt;
double beta;
extern double dnrm2_(int *, double *, int *);
extern /* Subroutine */ int dscal_(int *, double *, double *, int *);
double xnorm;
extern double dlapy2_(double *, double *), dlamch_(char *);
double safmin, rsafmn;
//
// -- LAPACK auxiliary routine (version 3.8.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// November 2017
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. External Functions ..
// ..
// .. Intrinsic Functions ..
// ..
// .. External Subroutines ..
// ..
// .. Executable Statements ..
//
// Parameter adjustments
--x;
// Function Body
if (*n <= 1) {
*tau = 0.;
return 0;
}
i__1 = *n - 1;
xnorm = dnrm2_(&i__1, &x[1], incx);
if (xnorm == 0.) {
//
// H = I
//
*tau = 0.;
} else {
//
// general case
//
d__1 = dlapy2_(alpha, &xnorm);
beta = -d_sign(&d__1, alpha);
safmin = dlamch_("S") / dlamch_("E");
knt = 0;
if (abs(beta) < safmin) {
//
// XNORM, BETA may be inaccurate; scale X and recompute them
//
rsafmn = 1. / safmin;
L10:
++knt;
i__1 = *n - 1;
dscal_(&i__1, &rsafmn, &x[1], incx);
beta *= rsafmn;
*alpha *= rsafmn;
if (abs(beta) < safmin && knt < 20) {
goto L10;
}
//
// New BETA is at most 1, at least SAFMIN
//
i__1 = *n - 1;
xnorm = dnrm2_(&i__1, &x[1], incx);
d__1 = dlapy2_(alpha, &xnorm);
beta = -d_sign(&d__1, alpha);
}
*tau = (beta - *alpha) / beta;
i__1 = *n - 1;
d__1 = 1. / (*alpha - beta);
dscal_(&i__1, &d__1, &x[1], incx);
//
// If ALPHA is subnormal, it may lose relative accuracy
//
i__1 = knt;
for (j = 1; j <= i__1; ++j) {
beta *= safmin;
// L20:
}
*alpha = beta;
}
return 0;
//
// End of DLARFG
//
} // dlarfg_
-389
View File
@@ -1,389 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DLARFT forms the triangular factor T of a block reflector H = I - vtvH
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DLARFT + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dlarft.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dlarft.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dlarft.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// SUBROUTINE DLARFT( DIRECT, STOREV, N, K, V, LDV, TAU, T, LDT )
//
// .. Scalar Arguments ..
// CHARACTER DIRECT, STOREV
// INTEGER K, LDT, LDV, N
// ..
// .. Array Arguments ..
// DOUBLE PRECISION T( LDT, * ), TAU( * ), V( LDV, * )
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DLARFT forms the triangular factor T of a real block reflector H
//> of order n, which is defined as a product of k elementary reflectors.
//>
//> If DIRECT = 'F', H = H(1) H(2) . . . H(k) and T is upper triangular;
//>
//> If DIRECT = 'B', H = H(k) . . . H(2) H(1) and T is lower triangular.
//>
//> If STOREV = 'C', the vector which defines the elementary reflector
//> H(i) is stored in the i-th column of the array V, and
//>
//> H = I - V * T * V**T
//>
//> If STOREV = 'R', the vector which defines the elementary reflector
//> H(i) is stored in the i-th row of the array V, and
//>
//> H = I - V**T * T * V
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] DIRECT
//> \verbatim
//> DIRECT is CHARACTER*1
//> Specifies the order in which the elementary reflectors are
//> multiplied to form the block reflector:
//> = 'F': H = H(1) H(2) . . . H(k) (Forward)
//> = 'B': H = H(k) . . . H(2) H(1) (Backward)
//> \endverbatim
//>
//> \param[in] STOREV
//> \verbatim
//> STOREV is CHARACTER*1
//> Specifies how the vectors which define the elementary
//> reflectors are stored (see also Further Details):
//> = 'C': columnwise
//> = 'R': rowwise
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> The order of the block reflector H. N >= 0.
//> \endverbatim
//>
//> \param[in] K
//> \verbatim
//> K is INTEGER
//> The order of the triangular factor T (= the number of
//> elementary reflectors). K >= 1.
//> \endverbatim
//>
//> \param[in] V
//> \verbatim
//> V is DOUBLE PRECISION array, dimension
//> (LDV,K) if STOREV = 'C'
//> (LDV,N) if STOREV = 'R'
//> The matrix V. See further details.
//> \endverbatim
//>
//> \param[in] LDV
//> \verbatim
//> LDV is INTEGER
//> The leading dimension of the array V.
//> If STOREV = 'C', LDV >= max(1,N); if STOREV = 'R', LDV >= K.
//> \endverbatim
//>
//> \param[in] TAU
//> \verbatim
//> TAU is DOUBLE PRECISION array, dimension (K)
//> TAU(i) must contain the scalar factor of the elementary
//> reflector H(i).
//> \endverbatim
//>
//> \param[out] T
//> \verbatim
//> T is DOUBLE PRECISION array, dimension (LDT,K)
//> The k by k triangular factor T of the block reflector.
//> If DIRECT = 'F', T is upper triangular; if DIRECT = 'B', T is
//> lower triangular. The rest of the array is not used.
//> \endverbatim
//>
//> \param[in] LDT
//> \verbatim
//> LDT is INTEGER
//> The leading dimension of the array T. LDT >= K.
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup doubleOTHERauxiliary
//
//> \par Further Details:
// =====================
//>
//> \verbatim
//>
//> The shape of the matrix V and the storage of the vectors which define
//> the H(i) is best illustrated by the following example with n = 5 and
//> k = 3. The elements equal to 1 are not stored.
//>
//> DIRECT = 'F' and STOREV = 'C': DIRECT = 'F' and STOREV = 'R':
//>
//> V = ( 1 ) V = ( 1 v1 v1 v1 v1 )
//> ( v1 1 ) ( 1 v2 v2 v2 )
//> ( v1 v2 1 ) ( 1 v3 v3 )
//> ( v1 v2 v3 )
//> ( v1 v2 v3 )
//>
//> DIRECT = 'B' and STOREV = 'C': DIRECT = 'B' and STOREV = 'R':
//>
//> V = ( v1 v2 v3 ) V = ( v1 v1 1 )
//> ( v1 v2 v3 ) ( v2 v2 v2 1 )
//> ( 1 v2 v3 ) ( v3 v3 v3 v3 1 )
//> ( 1 v3 )
//> ( 1 )
//> \endverbatim
//>
// =====================================================================
/* Subroutine */ int dlarft_(char *direct, char *storev, int *n, int *k,
double *v, int *ldv, double *tau, double *t, int *ldt)
{
// Table of constant values
int c__1 = 1;
double c_b7 = 1.;
// System generated locals
int t_dim1, t_offset, v_dim1, v_offset, i__1, i__2, i__3;
double d__1;
// Local variables
int i__, j, prevlastv;
extern int lsame_(char *, char *);
extern /* Subroutine */ int dgemv_(char *, int *, int *, double *, double
*, int *, double *, int *, double *, double *, int *);
int lastv;
extern /* Subroutine */ int dtrmv_(char *, char *, char *, int *, double *
, int *, double *, int *);
//
// -- LAPACK auxiliary routine (version 3.7.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. External Subroutines ..
// ..
// .. External Functions ..
// ..
// .. Executable Statements ..
//
// Quick return if possible
//
// Parameter adjustments
v_dim1 = *ldv;
v_offset = 1 + v_dim1;
v -= v_offset;
--tau;
t_dim1 = *ldt;
t_offset = 1 + t_dim1;
t -= t_offset;
// Function Body
if (*n == 0) {
return 0;
}
if (lsame_(direct, "F")) {
prevlastv = *n;
i__1 = *k;
for (i__ = 1; i__ <= i__1; ++i__) {
prevlastv = max(i__,prevlastv);
if (tau[i__] == 0.) {
//
// H(i) = I
//
i__2 = i__;
for (j = 1; j <= i__2; ++j) {
t[j + i__ * t_dim1] = 0.;
}
} else {
//
// general case
//
if (lsame_(storev, "C")) {
// Skip any trailing zeros.
i__2 = i__ + 1;
for (lastv = *n; lastv >= i__2; --lastv) {
if (v[lastv + i__ * v_dim1] != 0.) {
break;
}
}
i__2 = i__ - 1;
for (j = 1; j <= i__2; ++j) {
t[j + i__ * t_dim1] = -tau[i__] * v[i__ + j * v_dim1];
}
j = min(lastv,prevlastv);
//
// T(1:i-1,i) := - tau(i) * V(i:j,1:i-1)**T * V(i:j,i)
//
i__2 = j - i__;
i__3 = i__ - 1;
d__1 = -tau[i__];
dgemv_("Transpose", &i__2, &i__3, &d__1, &v[i__ + 1 +
v_dim1], ldv, &v[i__ + 1 + i__ * v_dim1], &c__1, &
c_b7, &t[i__ * t_dim1 + 1], &c__1);
} else {
// Skip any trailing zeros.
i__2 = i__ + 1;
for (lastv = *n; lastv >= i__2; --lastv) {
if (v[i__ + lastv * v_dim1] != 0.) {
break;
}
}
i__2 = i__ - 1;
for (j = 1; j <= i__2; ++j) {
t[j + i__ * t_dim1] = -tau[i__] * v[j + i__ * v_dim1];
}
j = min(lastv,prevlastv);
//
// T(1:i-1,i) := - tau(i) * V(1:i-1,i:j) * V(i,i:j)**T
//
i__2 = i__ - 1;
i__3 = j - i__;
d__1 = -tau[i__];
dgemv_("No transpose", &i__2, &i__3, &d__1, &v[(i__ + 1) *
v_dim1 + 1], ldv, &v[i__ + (i__ + 1) * v_dim1],
ldv, &c_b7, &t[i__ * t_dim1 + 1], &c__1);
}
//
// T(1:i-1,i) := T(1:i-1,1:i-1) * T(1:i-1,i)
//
i__2 = i__ - 1;
dtrmv_("Upper", "No transpose", "Non-unit", &i__2, &t[
t_offset], ldt, &t[i__ * t_dim1 + 1], &c__1);
t[i__ + i__ * t_dim1] = tau[i__];
if (i__ > 1) {
prevlastv = max(prevlastv,lastv);
} else {
prevlastv = lastv;
}
}
}
} else {
prevlastv = 1;
for (i__ = *k; i__ >= 1; --i__) {
if (tau[i__] == 0.) {
//
// H(i) = I
//
i__1 = *k;
for (j = i__; j <= i__1; ++j) {
t[j + i__ * t_dim1] = 0.;
}
} else {
//
// general case
//
if (i__ < *k) {
if (lsame_(storev, "C")) {
// Skip any leading zeros.
i__1 = i__ - 1;
for (lastv = 1; lastv <= i__1; ++lastv) {
if (v[lastv + i__ * v_dim1] != 0.) {
break;
}
}
i__1 = *k;
for (j = i__ + 1; j <= i__1; ++j) {
t[j + i__ * t_dim1] = -tau[i__] * v[*n - *k + i__
+ j * v_dim1];
}
j = max(lastv,prevlastv);
//
// T(i+1:k,i) = -tau(i) * V(j:n-k+i,i+1:k)**T * V(j:n-k+i,i)
//
i__1 = *n - *k + i__ - j;
i__2 = *k - i__;
d__1 = -tau[i__];
dgemv_("Transpose", &i__1, &i__2, &d__1, &v[j + (i__
+ 1) * v_dim1], ldv, &v[j + i__ * v_dim1], &
c__1, &c_b7, &t[i__ + 1 + i__ * t_dim1], &
c__1);
} else {
// Skip any leading zeros.
i__1 = i__ - 1;
for (lastv = 1; lastv <= i__1; ++lastv) {
if (v[i__ + lastv * v_dim1] != 0.) {
break;
}
}
i__1 = *k;
for (j = i__ + 1; j <= i__1; ++j) {
t[j + i__ * t_dim1] = -tau[i__] * v[j + (*n - *k
+ i__) * v_dim1];
}
j = max(lastv,prevlastv);
//
// T(i+1:k,i) = -tau(i) * V(i+1:k,j:n-k+i) * V(i,j:n-k+i)**T
//
i__1 = *k - i__;
i__2 = *n - *k + i__ - j;
d__1 = -tau[i__];
dgemv_("No transpose", &i__1, &i__2, &d__1, &v[i__ +
1 + j * v_dim1], ldv, &v[i__ + j * v_dim1],
ldv, &c_b7, &t[i__ + 1 + i__ * t_dim1], &c__1)
;
}
//
// T(i+1:k,i) := T(i+1:k,i+1:k) * T(i+1:k,i)
//
i__1 = *k - i__;
dtrmv_("Lower", "No transpose", "Non-unit", &i__1, &t[i__
+ 1 + (i__ + 1) * t_dim1], ldt, &t[i__ + 1 + i__ *
t_dim1], &c__1);
if (i__ > 1) {
prevlastv = min(prevlastv,lastv);
} else {
prevlastv = lastv;
}
}
t[i__ + i__ * t_dim1] = tau[i__];
}
}
}
return 0;
//
// End of DLARFT
//
} // dlarft_
-236
View File
@@ -1,236 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DLARTG generates a plane rotation with real cosine and real sine.
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DLARTG + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dlartg.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dlartg.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dlartg.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// SUBROUTINE DLARTG( F, G, CS, SN, R )
//
// .. Scalar Arguments ..
// DOUBLE PRECISION CS, F, G, R, SN
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DLARTG generate a plane rotation so that
//>
//> [ CS SN ] . [ F ] = [ R ] where CS**2 + SN**2 = 1.
//> [ -SN CS ] [ G ] [ 0 ]
//>
//> This is a slower, more accurate version of the BLAS1 routine DROTG,
//> with the following other differences:
//> F and G are unchanged on return.
//> If G=0, then CS=1 and SN=0.
//> If F=0 and (G .ne. 0), then CS=0 and SN=1 without doing any
//> floating point operations (saves work in DBDSQR when
//> there are zeros on the diagonal).
//>
//> If F exceeds G in magnitude, CS will be positive.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] F
//> \verbatim
//> F is DOUBLE PRECISION
//> The first component of vector to be rotated.
//> \endverbatim
//>
//> \param[in] G
//> \verbatim
//> G is DOUBLE PRECISION
//> The second component of vector to be rotated.
//> \endverbatim
//>
//> \param[out] CS
//> \verbatim
//> CS is DOUBLE PRECISION
//> The cosine of the rotation.
//> \endverbatim
//>
//> \param[out] SN
//> \verbatim
//> SN is DOUBLE PRECISION
//> The sine of the rotation.
//> \endverbatim
//>
//> \param[out] R
//> \verbatim
//> R is DOUBLE PRECISION
//> The nonzero component of the rotated vector.
//>
//> This version has a few statements commented out for thread safety
//> (machine parameters are computed on each entry). 10 feb 03, SJH.
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup OTHERauxiliary
//
// =====================================================================
/* Subroutine */ int dlartg_(double *f, double *g, double *cs, double *sn,
double *r__)
{
// System generated locals
int i__1;
double d__1, d__2;
// Local variables
int i__;
double f1, g1, eps, scale;
int count;
double safmn2, safmx2;
extern double dlamch_(char *);
double safmin;
//
// -- LAPACK auxiliary routine (version 3.7.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// LOGICAL FIRST
// ..
// .. External Functions ..
// ..
// .. Intrinsic Functions ..
// ..
// .. Save statement ..
// SAVE FIRST, SAFMX2, SAFMIN, SAFMN2
// ..
// .. Data statements ..
// DATA FIRST / .TRUE. /
// ..
// .. Executable Statements ..
//
// IF( FIRST ) THEN
safmin = dlamch_("S");
eps = dlamch_("E");
d__1 = dlamch_("B");
i__1 = (int) (log(safmin / eps) / log(dlamch_("B")) / 2.);
safmn2 = pow_di(&d__1, &i__1);
safmx2 = 1. / safmn2;
// FIRST = .FALSE.
// END IF
if (*g == 0.) {
*cs = 1.;
*sn = 0.;
*r__ = *f;
} else if (*f == 0.) {
*cs = 0.;
*sn = 1.;
*r__ = *g;
} else {
f1 = *f;
g1 = *g;
// Computing MAX
d__1 = abs(f1), d__2 = abs(g1);
scale = max(d__1,d__2);
if (scale >= safmx2) {
count = 0;
L10:
++count;
f1 *= safmn2;
g1 *= safmn2;
// Computing MAX
d__1 = abs(f1), d__2 = abs(g1);
scale = max(d__1,d__2);
if (scale >= safmx2) {
goto L10;
}
// Computing 2nd power
d__1 = f1;
// Computing 2nd power
d__2 = g1;
*r__ = sqrt(d__1 * d__1 + d__2 * d__2);
*cs = f1 / *r__;
*sn = g1 / *r__;
i__1 = count;
for (i__ = 1; i__ <= i__1; ++i__) {
*r__ *= safmx2;
// L20:
}
} else if (scale <= safmn2) {
count = 0;
L30:
++count;
f1 *= safmx2;
g1 *= safmx2;
// Computing MAX
d__1 = abs(f1), d__2 = abs(g1);
scale = max(d__1,d__2);
if (scale <= safmn2) {
goto L30;
}
// Computing 2nd power
d__1 = f1;
// Computing 2nd power
d__2 = g1;
*r__ = sqrt(d__1 * d__1 + d__2 * d__2);
*cs = f1 / *r__;
*sn = g1 / *r__;
i__1 = count;
for (i__ = 1; i__ <= i__1; ++i__) {
*r__ *= safmn2;
// L40:
}
} else {
// Computing 2nd power
d__1 = f1;
// Computing 2nd power
d__2 = g1;
*r__ = sqrt(d__1 * d__1 + d__2 * d__2);
*cs = f1 / *r__;
*sn = g1 / *r__;
}
if (abs(*f) > abs(*g) && *cs < 0.) {
*cs = -(*cs);
*sn = -(*sn);
*r__ = -(*r__);
}
}
return 0;
//
// End of DLARTG
//
} // dlartg_
-413
View File
@@ -1,413 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DLASCL multiplies a general rectangular matrix by a real scalar defined as cto/cfrom.
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DLASCL + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dlascl.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dlascl.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dlascl.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// SUBROUTINE DLASCL( TYPE, KL, KU, CFROM, CTO, M, N, A, LDA, INFO )
//
// .. Scalar Arguments ..
// CHARACTER TYPE
// INTEGER INFO, KL, KU, LDA, M, N
// DOUBLE PRECISION CFROM, CTO
// ..
// .. Array Arguments ..
// DOUBLE PRECISION A( LDA, * )
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DLASCL multiplies the M by N real matrix A by the real scalar
//> CTO/CFROM. This is done without over/underflow as long as the final
//> result CTO*A(I,J)/CFROM does not over/underflow. TYPE specifies that
//> A may be full, upper triangular, lower triangular, upper Hessenberg,
//> or banded.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] TYPE
//> \verbatim
//> TYPE is CHARACTER*1
//> TYPE indices the storage type of the input matrix.
//> = 'G': A is a full matrix.
//> = 'L': A is a lower triangular matrix.
//> = 'U': A is an upper triangular matrix.
//> = 'H': A is an upper Hessenberg matrix.
//> = 'B': A is a symmetric band matrix with lower bandwidth KL
//> and upper bandwidth KU and with the only the lower
//> half stored.
//> = 'Q': A is a symmetric band matrix with lower bandwidth KL
//> and upper bandwidth KU and with the only the upper
//> half stored.
//> = 'Z': A is a band matrix with lower bandwidth KL and upper
//> bandwidth KU. See DGBTRF for storage details.
//> \endverbatim
//>
//> \param[in] KL
//> \verbatim
//> KL is INTEGER
//> The lower bandwidth of A. Referenced only if TYPE = 'B',
//> 'Q' or 'Z'.
//> \endverbatim
//>
//> \param[in] KU
//> \verbatim
//> KU is INTEGER
//> The upper bandwidth of A. Referenced only if TYPE = 'B',
//> 'Q' or 'Z'.
//> \endverbatim
//>
//> \param[in] CFROM
//> \verbatim
//> CFROM is DOUBLE PRECISION
//> \endverbatim
//>
//> \param[in] CTO
//> \verbatim
//> CTO is DOUBLE PRECISION
//>
//> The matrix A is multiplied by CTO/CFROM. A(I,J) is computed
//> without over/underflow if the final result CTO*A(I,J)/CFROM
//> can be represented without over/underflow. CFROM must be
//> nonzero.
//> \endverbatim
//>
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> The number of rows of the matrix A. M >= 0.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> The number of columns of the matrix A. N >= 0.
//> \endverbatim
//>
//> \param[in,out] A
//> \verbatim
//> A is DOUBLE PRECISION array, dimension (LDA,N)
//> The matrix to be multiplied by CTO/CFROM. See TYPE for the
//> storage type.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> The leading dimension of the array A.
//> If TYPE = 'G', 'L', 'U', 'H', LDA >= max(1,M);
//> TYPE = 'B', LDA >= KL+1;
//> TYPE = 'Q', LDA >= KU+1;
//> TYPE = 'Z', LDA >= 2*KL+KU+1.
//> \endverbatim
//>
//> \param[out] INFO
//> \verbatim
//> INFO is INTEGER
//> 0 - successful exit
//> <0 - if INFO = -i, the i-th argument had an illegal value.
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date June 2016
//
//> \ingroup OTHERauxiliary
//
// =====================================================================
/* Subroutine */ int dlascl_(char *type__, int *kl, int *ku, double *cfrom,
double *cto, int *m, int *n, double *a, int *lda, int *info)
{
// System generated locals
int a_dim1, a_offset, i__1, i__2, i__3, i__4, i__5;
// Local variables
int i__, j, k1, k2, k3, k4;
double mul, cto1;
int done;
double ctoc;
extern int lsame_(char *, char *);
int itype;
double cfrom1;
extern double dlamch_(char *);
double cfromc;
extern int disnan_(double *);
extern /* Subroutine */ int xerbla_(char *, int *);
double bignum, smlnum;
//
// -- LAPACK auxiliary routine (version 3.7.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// June 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. External Functions ..
// ..
// .. Intrinsic Functions ..
// ..
// .. External Subroutines ..
// ..
// .. Executable Statements ..
//
// Test the input arguments
//
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
// Function Body
*info = 0;
if (lsame_(type__, "G")) {
itype = 0;
} else if (lsame_(type__, "L")) {
itype = 1;
} else if (lsame_(type__, "U")) {
itype = 2;
} else if (lsame_(type__, "H")) {
itype = 3;
} else if (lsame_(type__, "B")) {
itype = 4;
} else if (lsame_(type__, "Q")) {
itype = 5;
} else if (lsame_(type__, "Z")) {
itype = 6;
} else {
itype = -1;
}
if (itype == -1) {
*info = -1;
} else if (*cfrom == 0. || disnan_(cfrom)) {
*info = -4;
} else if (disnan_(cto)) {
*info = -5;
} else if (*m < 0) {
*info = -6;
} else if (*n < 0 || itype == 4 && *n != *m || itype == 5 && *n != *m) {
*info = -7;
} else if (itype <= 3 && *lda < max(1,*m)) {
*info = -9;
} else if (itype >= 4) {
// Computing MAX
i__1 = *m - 1;
if (*kl < 0 || *kl > max(i__1,0)) {
*info = -2;
} else /* if(complicated condition) */ {
// Computing MAX
i__1 = *n - 1;
if (*ku < 0 || *ku > max(i__1,0) || (itype == 4 || itype == 5) &&
*kl != *ku) {
*info = -3;
} else if (itype == 4 && *lda < *kl + 1 || itype == 5 && *lda < *
ku + 1 || itype == 6 && *lda < (*kl << 1) + *ku + 1) {
*info = -9;
}
}
}
if (*info != 0) {
i__1 = -(*info);
xerbla_("DLASCL", &i__1);
return 0;
}
//
// Quick return if possible
//
if (*n == 0 || *m == 0) {
return 0;
}
//
// Get machine parameters
//
smlnum = dlamch_("S");
bignum = 1. / smlnum;
cfromc = *cfrom;
ctoc = *cto;
L10:
cfrom1 = cfromc * smlnum;
if (cfrom1 == cfromc) {
// CFROMC is an inf. Multiply by a correctly signed zero for
// finite CTOC, or a NaN if CTOC is infinite.
mul = ctoc / cfromc;
done = TRUE_;
cto1 = ctoc;
} else {
cto1 = ctoc / bignum;
if (cto1 == ctoc) {
// CTOC is either 0 or an inf. In both cases, CTOC itself
// serves as the correct multiplication factor.
mul = ctoc;
done = TRUE_;
cfromc = 1.;
} else if (abs(cfrom1) > abs(ctoc) && ctoc != 0.) {
mul = smlnum;
done = FALSE_;
cfromc = cfrom1;
} else if (abs(cto1) > abs(cfromc)) {
mul = bignum;
done = FALSE_;
ctoc = cto1;
} else {
mul = ctoc / cfromc;
done = TRUE_;
}
}
if (itype == 0) {
//
// Full matrix
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
a[i__ + j * a_dim1] *= mul;
// L20:
}
// L30:
}
} else if (itype == 1) {
//
// Lower triangular matrix
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = j; i__ <= i__2; ++i__) {
a[i__ + j * a_dim1] *= mul;
// L40:
}
// L50:
}
} else if (itype == 2) {
//
// Upper triangular matrix
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = min(j,*m);
for (i__ = 1; i__ <= i__2; ++i__) {
a[i__ + j * a_dim1] *= mul;
// L60:
}
// L70:
}
} else if (itype == 3) {
//
// Upper Hessenberg matrix
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
// Computing MIN
i__3 = j + 1;
i__2 = min(i__3,*m);
for (i__ = 1; i__ <= i__2; ++i__) {
a[i__ + j * a_dim1] *= mul;
// L80:
}
// L90:
}
} else if (itype == 4) {
//
// Lower half of a symmetric band matrix
//
k3 = *kl + 1;
k4 = *n + 1;
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
// Computing MIN
i__3 = k3, i__4 = k4 - j;
i__2 = min(i__3,i__4);
for (i__ = 1; i__ <= i__2; ++i__) {
a[i__ + j * a_dim1] *= mul;
// L100:
}
// L110:
}
} else if (itype == 5) {
//
// Upper half of a symmetric band matrix
//
k1 = *ku + 2;
k3 = *ku + 1;
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
// Computing MAX
i__2 = k1 - j;
i__3 = k3;
for (i__ = max(i__2,1); i__ <= i__3; ++i__) {
a[i__ + j * a_dim1] *= mul;
// L120:
}
// L130:
}
} else if (itype == 6) {
//
// Band matrix
//
k1 = *kl + *ku + 2;
k2 = *kl + 1;
k3 = (*kl << 1) + *ku + 1;
k4 = *kl + *ku + 1 + *m;
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
// Computing MAX
i__3 = k1 - j;
// Computing MIN
i__4 = k3, i__5 = k4 - j;
i__2 = min(i__4,i__5);
for (i__ = max(i__3,k2); i__ <= i__2; ++i__) {
a[i__ + j * a_dim1] *= mul;
// L140:
}
// L150:
}
}
if (! done) {
goto L10;
}
return 0;
//
// End of DLASCL
//
} // dlascl_
-209
View File
@@ -1,209 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DLASET initializes the off-diagonal elements and the diagonal elements of a matrix to given values.
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DLASET + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dlaset.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dlaset.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dlaset.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// SUBROUTINE DLASET( UPLO, M, N, ALPHA, BETA, A, LDA )
//
// .. Scalar Arguments ..
// CHARACTER UPLO
// INTEGER LDA, M, N
// DOUBLE PRECISION ALPHA, BETA
// ..
// .. Array Arguments ..
// DOUBLE PRECISION A( LDA, * )
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DLASET initializes an m-by-n matrix A to BETA on the diagonal and
//> ALPHA on the offdiagonals.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] UPLO
//> \verbatim
//> UPLO is CHARACTER*1
//> Specifies the part of the matrix A to be set.
//> = 'U': Upper triangular part is set; the strictly lower
//> triangular part of A is not changed.
//> = 'L': Lower triangular part is set; the strictly upper
//> triangular part of A is not changed.
//> Otherwise: All of the matrix A is set.
//> \endverbatim
//>
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> The number of rows of the matrix A. M >= 0.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> The number of columns of the matrix A. N >= 0.
//> \endverbatim
//>
//> \param[in] ALPHA
//> \verbatim
//> ALPHA is DOUBLE PRECISION
//> The constant to which the offdiagonal elements are to be set.
//> \endverbatim
//>
//> \param[in] BETA
//> \verbatim
//> BETA is DOUBLE PRECISION
//> The constant to which the diagonal elements are to be set.
//> \endverbatim
//>
//> \param[out] A
//> \verbatim
//> A is DOUBLE PRECISION array, dimension (LDA,N)
//> On exit, the leading m-by-n submatrix of A is set as follows:
//>
//> if UPLO = 'U', A(i,j) = ALPHA, 1<=i<=j-1, 1<=j<=n,
//> if UPLO = 'L', A(i,j) = ALPHA, j+1<=i<=m, 1<=j<=n,
//> otherwise, A(i,j) = ALPHA, 1<=i<=m, 1<=j<=n, i.ne.j,
//>
//> and, for all UPLO, A(i,i) = BETA, 1<=i<=min(m,n).
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> The leading dimension of the array A. LDA >= max(1,M).
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup OTHERauxiliary
//
// =====================================================================
/* Subroutine */ int dlaset_(char *uplo, int *m, int *n, double *alpha,
double *beta, double *a, int *lda)
{
// System generated locals
int a_dim1, a_offset, i__1, i__2, i__3;
// Local variables
int i__, j;
extern int lsame_(char *, char *);
//
// -- LAPACK auxiliary routine (version 3.7.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
//=====================================================================
//
// .. Local Scalars ..
// ..
// .. External Functions ..
// ..
// .. Intrinsic Functions ..
// ..
// .. Executable Statements ..
//
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
// Function Body
if (lsame_(uplo, "U")) {
//
// Set the strictly upper triangular or trapezoidal part of the
// array to ALPHA.
//
i__1 = *n;
for (j = 2; j <= i__1; ++j) {
// Computing MIN
i__3 = j - 1;
i__2 = min(i__3,*m);
for (i__ = 1; i__ <= i__2; ++i__) {
a[i__ + j * a_dim1] = *alpha;
// L10:
}
// L20:
}
} else if (lsame_(uplo, "L")) {
//
// Set the strictly lower triangular or trapezoidal part of the
// array to ALPHA.
//
i__1 = min(*m,*n);
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = j + 1; i__ <= i__2; ++i__) {
a[i__ + j * a_dim1] = *alpha;
// L30:
}
// L40:
}
} else {
//
// Set the leading m-by-n submatrix to ALPHA.
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
a[i__ + j * a_dim1] = *alpha;
// L50:
}
// L60:
}
}
//
// Set the first min(M,N) diagonal elements to BETA.
//
i__1 = min(*m,*n);
for (i__ = 1; i__ <= i__1; ++i__) {
a[i__ + i__ * a_dim1] = *beta;
// L70:
}
return 0;
//
// End of DLASET
//
} // dlaset_
-172
View File
@@ -1,172 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DLASSQ updates a sum of squares represented in scaled form.
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DLASSQ + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dlassq.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dlassq.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dlassq.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// SUBROUTINE DLASSQ( N, X, INCX, SCALE, SUMSQ )
//
// .. Scalar Arguments ..
// INTEGER INCX, N
// DOUBLE PRECISION SCALE, SUMSQ
// ..
// .. Array Arguments ..
// DOUBLE PRECISION X( * )
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DLASSQ returns the values scl and smsq such that
//>
//> ( scl**2 )*smsq = x( 1 )**2 +...+ x( n )**2 + ( scale**2 )*sumsq,
//>
//> where x( i ) = X( 1 + ( i - 1 )*INCX ). The value of sumsq is
//> assumed to be non-negative and scl returns the value
//>
//> scl = max( scale, abs( x( i ) ) ).
//>
//> scale and sumsq must be supplied in SCALE and SUMSQ and
//> scl and smsq are overwritten on SCALE and SUMSQ respectively.
//>
//> The routine makes only one pass through the vector x.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> The number of elements to be used from the vector X.
//> \endverbatim
//>
//> \param[in] X
//> \verbatim
//> X is DOUBLE PRECISION array, dimension (1+(N-1)*INCX)
//> The vector for which a scaled sum of squares is computed.
//> x( i ) = X( 1 + ( i - 1 )*INCX ), 1 <= i <= n.
//> \endverbatim
//>
//> \param[in] INCX
//> \verbatim
//> INCX is INTEGER
//> The increment between successive values of the vector X.
//> INCX > 0.
//> \endverbatim
//>
//> \param[in,out] SCALE
//> \verbatim
//> SCALE is DOUBLE PRECISION
//> On entry, the value scale in the equation above.
//> On exit, SCALE is overwritten with scl , the scaling factor
//> for the sum of squares.
//> \endverbatim
//>
//> \param[in,out] SUMSQ
//> \verbatim
//> SUMSQ is DOUBLE PRECISION
//> On entry, the value sumsq in the equation above.
//> On exit, SUMSQ is overwritten with smsq , the basic sum of
//> squares from which scl has been factored out.
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup OTHERauxiliary
//
// =====================================================================
/* Subroutine */ int dlassq_(int *n, double *x, int *incx, double *scale,
double *sumsq)
{
// System generated locals
int i__1, i__2;
double d__1;
// Local variables
int ix;
double absxi;
extern int disnan_(double *);
//
// -- LAPACK auxiliary routine (version 3.7.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
//=====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. External Functions ..
// ..
// .. Intrinsic Functions ..
// ..
// .. Executable Statements ..
//
// Parameter adjustments
--x;
// Function Body
if (*n > 0) {
i__1 = (*n - 1) * *incx + 1;
i__2 = *incx;
for (ix = 1; i__2 < 0 ? ix >= i__1 : ix <= i__1; ix += i__2) {
absxi = (d__1 = x[ix], abs(d__1));
if (absxi > 0. || disnan_(&absxi)) {
if (*scale < absxi) {
// Computing 2nd power
d__1 = *scale / absxi;
*sumsq = *sumsq * (d__1 * d__1) + 1;
*scale = absxi;
} else {
// Computing 2nd power
d__1 = absxi / *scale;
*sumsq += d__1 * d__1;
}
}
// L10:
}
}
return 0;
//
// End of DLASSQ
//
} // dlassq_
-149
View File
@@ -1,149 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DNRM2
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
// Definition:
// ===========
//
// DOUBLE PRECISION FUNCTION DNRM2(N,X,INCX)
//
// .. Scalar Arguments ..
// INTEGER INCX,N
// ..
// .. Array Arguments ..
// DOUBLE PRECISION X(*)
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DNRM2 returns the euclidean norm of a vector via the function
//> name, so that
//>
//> DNRM2 := sqrt( x'*x )
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> number of elements in input vector(s)
//> \endverbatim
//>
//> \param[in] X
//> \verbatim
//> X is DOUBLE PRECISION array, dimension ( 1 + ( N - 1 )*abs( INCX ) )
//> \endverbatim
//>
//> \param[in] INCX
//> \verbatim
//> INCX is INTEGER
//> storage spacing between elements of DX
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date November 2017
//
//> \ingroup double_blas_level1
//
//> \par Further Details:
// =====================
//>
//> \verbatim
//>
//> -- This version written on 25-October-1982.
//> Modified on 14-October-1993 to inline the call to DLASSQ.
//> Sven Hammarling, Nag Ltd.
//> \endverbatim
//>
// =====================================================================
double dnrm2_(int *n, double *x, int *incx)
{
// System generated locals
int i__1, i__2;
double ret_val, d__1;
// Local variables
int ix;
double ssq, norm, scale, absxi;
//
// -- Reference BLAS level1 routine (version 3.8.0) --
// -- Reference BLAS is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// November 2017
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. Intrinsic Functions ..
// ..
// Parameter adjustments
--x;
// Function Body
if (*n < 1 || *incx < 1) {
norm = 0.;
} else if (*n == 1) {
norm = abs(x[1]);
} else {
scale = 0.;
ssq = 1.;
// The following loop is equivalent to this call to the LAPACK
// auxiliary routine:
// CALL DLASSQ( N, X, INCX, SCALE, SSQ )
//
i__1 = (*n - 1) * *incx + 1;
i__2 = *incx;
for (ix = 1; i__2 < 0 ? ix >= i__1 : ix <= i__1; ix += i__2) {
if (x[ix] != 0.) {
absxi = (d__1 = x[ix], abs(d__1));
if (scale < absxi) {
// Computing 2nd power
d__1 = scale / absxi;
ssq = ssq * (d__1 * d__1) + 1.;
scale = absxi;
} else {
// Computing 2nd power
d__1 = absxi / scale;
ssq += d__1 * d__1;
}
}
// L10:
}
norm = scale * sqrt(ssq);
}
ret_val = norm;
return ret_val;
//
// End of DNRM2.
//
} // dnrm2_
-571
View File
@@ -1,571 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DORG2R generates all or part of the orthogonal matrix Q from a QR factorization determined by sgeqrf (unblocked algorithm).
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DORG2R + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dorg2r.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dorg2r.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dorg2r.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// SUBROUTINE DORG2R( M, N, K, A, LDA, TAU, WORK, INFO )
//
// .. Scalar Arguments ..
// INTEGER INFO, K, LDA, M, N
// ..
// .. Array Arguments ..
// DOUBLE PRECISION A( LDA, * ), TAU( * ), WORK( * )
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DORG2R generates an m by n real matrix Q with orthonormal columns,
//> which is defined as the first n columns of a product of k elementary
//> reflectors of order m
//>
//> Q = H(1) H(2) . . . H(k)
//>
//> as returned by DGEQRF.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> The number of rows of the matrix Q. M >= 0.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> The number of columns of the matrix Q. M >= N >= 0.
//> \endverbatim
//>
//> \param[in] K
//> \verbatim
//> K is INTEGER
//> The number of elementary reflectors whose product defines the
//> matrix Q. N >= K >= 0.
//> \endverbatim
//>
//> \param[in,out] A
//> \verbatim
//> A is DOUBLE PRECISION array, dimension (LDA,N)
//> On entry, the i-th column must contain the vector which
//> defines the elementary reflector H(i), for i = 1,2,...,k, as
//> returned by DGEQRF in the first k columns of its array
//> argument A.
//> On exit, the m-by-n matrix Q.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> The first dimension of the array A. LDA >= max(1,M).
//> \endverbatim
//>
//> \param[in] TAU
//> \verbatim
//> TAU is DOUBLE PRECISION array, dimension (K)
//> TAU(i) must contain the scalar factor of the elementary
//> reflector H(i), as returned by DGEQRF.
//> \endverbatim
//>
//> \param[out] WORK
//> \verbatim
//> WORK is DOUBLE PRECISION array, dimension (N)
//> \endverbatim
//>
//> \param[out] INFO
//> \verbatim
//> INFO is INTEGER
//> = 0: successful exit
//> < 0: if INFO = -i, the i-th argument has an illegal value
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup doubleOTHERcomputational
//
// =====================================================================
/* Subroutine */ int dorg2r_(int *m, int *n, int *k, double *a, int *lda,
double *tau, double *work, int *info)
{
// Table of constant values
int c__1 = 1;
// System generated locals
int a_dim1, a_offset, i__1, i__2;
double d__1;
// Local variables
int i__, j, l;
extern /* Subroutine */ int dscal_(int *, double *, double *, int *),
dlarf_(char *, int *, int *, double *, int *, double *, double *,
int *, double *), xerbla_(char *, int *);
//
// -- LAPACK computational routine (version 3.7.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. External Subroutines ..
// ..
// .. Intrinsic Functions ..
// ..
// .. Executable Statements ..
//
// Test the input arguments
//
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
--tau;
--work;
// Function Body
*info = 0;
if (*m < 0) {
*info = -1;
} else if (*n < 0 || *n > *m) {
*info = -2;
} else if (*k < 0 || *k > *n) {
*info = -3;
} else if (*lda < max(1,*m)) {
*info = -5;
}
if (*info != 0) {
i__1 = -(*info);
xerbla_("DORG2R", &i__1);
return 0;
}
//
// Quick return if possible
//
if (*n <= 0) {
return 0;
}
//
// Initialise columns k+1:n to columns of the unit matrix
//
i__1 = *n;
for (j = *k + 1; j <= i__1; ++j) {
i__2 = *m;
for (l = 1; l <= i__2; ++l) {
a[l + j * a_dim1] = 0.;
// L10:
}
a[j + j * a_dim1] = 1.;
// L20:
}
for (i__ = *k; i__ >= 1; --i__) {
//
// Apply H(i) to A(i:m,i:n) from the left
//
if (i__ < *n) {
a[i__ + i__ * a_dim1] = 1.;
i__1 = *m - i__ + 1;
i__2 = *n - i__;
dlarf_("Left", &i__1, &i__2, &a[i__ + i__ * a_dim1], &c__1, &tau[
i__], &a[i__ + (i__ + 1) * a_dim1], lda, &work[1]);
}
if (i__ < *m) {
i__1 = *m - i__;
d__1 = -tau[i__];
dscal_(&i__1, &d__1, &a[i__ + 1 + i__ * a_dim1], &c__1);
}
a[i__ + i__ * a_dim1] = 1. - tau[i__];
//
// Set A(1:i-1,i) to zero
//
i__1 = i__ - 1;
for (l = 1; l <= i__1; ++l) {
a[l + i__ * a_dim1] = 0.;
// L30:
}
// L40:
}
return 0;
//
// End of DORG2R
//
} // dorg2r_
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
//> \brief \b DORGQR
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DORGQR + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dorgqr.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dorgqr.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dorgqr.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// SUBROUTINE DORGQR( M, N, K, A, LDA, TAU, WORK, LWORK, INFO )
//
// .. Scalar Arguments ..
// INTEGER INFO, K, LDA, LWORK, M, N
// ..
// .. Array Arguments ..
// DOUBLE PRECISION A( LDA, * ), TAU( * ), WORK( * )
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DORGQR generates an M-by-N real matrix Q with orthonormal columns,
//> which is defined as the first N columns of a product of K elementary
//> reflectors of order M
//>
//> Q = H(1) H(2) . . . H(k)
//>
//> as returned by DGEQRF.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> The number of rows of the matrix Q. M >= 0.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> The number of columns of the matrix Q. M >= N >= 0.
//> \endverbatim
//>
//> \param[in] K
//> \verbatim
//> K is INTEGER
//> The number of elementary reflectors whose product defines the
//> matrix Q. N >= K >= 0.
//> \endverbatim
//>
//> \param[in,out] A
//> \verbatim
//> A is DOUBLE PRECISION array, dimension (LDA,N)
//> On entry, the i-th column must contain the vector which
//> defines the elementary reflector H(i), for i = 1,2,...,k, as
//> returned by DGEQRF in the first k columns of its array
//> argument A.
//> On exit, the M-by-N matrix Q.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> The first dimension of the array A. LDA >= max(1,M).
//> \endverbatim
//>
//> \param[in] TAU
//> \verbatim
//> TAU is DOUBLE PRECISION array, dimension (K)
//> TAU(i) must contain the scalar factor of the elementary
//> reflector H(i), as returned by DGEQRF.
//> \endverbatim
//>
//> \param[out] WORK
//> \verbatim
//> WORK is DOUBLE PRECISION array, dimension (MAX(1,LWORK))
//> On exit, if INFO = 0, WORK(1) returns the optimal LWORK.
//> \endverbatim
//>
//> \param[in] LWORK
//> \verbatim
//> LWORK is INTEGER
//> The dimension of the array WORK. LWORK >= max(1,N).
//> For optimum performance LWORK >= N*NB, where NB is the
//> optimal blocksize.
//>
//> If LWORK = -1, then a workspace query is assumed; the routine
//> only calculates the optimal size of the WORK array, returns
//> this value as the first entry of the WORK array, and no error
//> message related to LWORK is issued by XERBLA.
//> \endverbatim
//>
//> \param[out] INFO
//> \verbatim
//> INFO is INTEGER
//> = 0: successful exit
//> < 0: if INFO = -i, the i-th argument has an illegal value
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup doubleOTHERcomputational
//
// =====================================================================
/* Subroutine */ int dorgqr_(int *m, int *n, int *k, double *a, int *lda,
double *tau, double *work, int *lwork, int *info)
{
// Table of constant values
int c__1 = 1;
int c_n1 = -1;
int c__3 = 3;
int c__2 = 2;
// System generated locals
int a_dim1, a_offset, i__1, i__2, i__3;
// Local variables
int i__, j, l, ib, nb, ki, kk, nx, iws, nbmin, iinfo;
extern /* Subroutine */ int dorg2r_(int *, int *, int *, double *, int *,
double *, double *, int *), dlarfb_(char *, char *, char *, char *
, int *, int *, int *, double *, int *, double *, int *, double *,
int *, double *, int *), dlarft_(char *, char *, int *, int *,
double *, int *, double *, double *, int *), xerbla_(char *, int *
);
extern int ilaenv_(int *, char *, char *, int *, int *, int *, int *);
int ldwork, lwkopt;
int lquery;
//
// -- LAPACK computational routine (version 3.7.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. External Subroutines ..
// ..
// .. Intrinsic Functions ..
// ..
// .. External Functions ..
// ..
// .. Executable Statements ..
//
// Test the input arguments
//
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
--tau;
--work;
// Function Body
*info = 0;
nb = ilaenv_(&c__1, "DORGQR", " ", m, n, k, &c_n1);
lwkopt = max(1,*n) * nb;
work[1] = (double) lwkopt;
lquery = *lwork == -1;
if (*m < 0) {
*info = -1;
} else if (*n < 0 || *n > *m) {
*info = -2;
} else if (*k < 0 || *k > *n) {
*info = -3;
} else if (*lda < max(1,*m)) {
*info = -5;
} else if (*lwork < max(1,*n) && ! lquery) {
*info = -8;
}
if (*info != 0) {
i__1 = -(*info);
xerbla_("DORGQR", &i__1);
return 0;
} else if (lquery) {
return 0;
}
//
// Quick return if possible
//
if (*n <= 0) {
work[1] = 1.;
return 0;
}
nbmin = 2;
nx = 0;
iws = *n;
if (nb > 1 && nb < *k) {
//
// Determine when to cross over from blocked to unblocked code.
//
// Computing MAX
i__1 = 0, i__2 = ilaenv_(&c__3, "DORGQR", " ", m, n, k, &c_n1);
nx = max(i__1,i__2);
if (nx < *k) {
//
// Determine if workspace is large enough for blocked code.
//
ldwork = *n;
iws = ldwork * nb;
if (*lwork < iws) {
//
// Not enough workspace to use optimal NB: reduce NB and
// determine the minimum value of NB.
//
nb = *lwork / ldwork;
// Computing MAX
i__1 = 2, i__2 = ilaenv_(&c__2, "DORGQR", " ", m, n, k, &c_n1)
;
nbmin = max(i__1,i__2);
}
}
}
if (nb >= nbmin && nb < *k && nx < *k) {
//
// Use blocked code after the last block.
// The first kk columns are handled by the block method.
//
ki = (*k - nx - 1) / nb * nb;
// Computing MIN
i__1 = *k, i__2 = ki + nb;
kk = min(i__1,i__2);
//
// Set A(1:kk,kk+1:n) to zero.
//
i__1 = *n;
for (j = kk + 1; j <= i__1; ++j) {
i__2 = kk;
for (i__ = 1; i__ <= i__2; ++i__) {
a[i__ + j * a_dim1] = 0.;
// L10:
}
// L20:
}
} else {
kk = 0;
}
//
// Use unblocked code for the last or only block.
//
if (kk < *n) {
i__1 = *m - kk;
i__2 = *n - kk;
i__3 = *k - kk;
dorg2r_(&i__1, &i__2, &i__3, &a[kk + 1 + (kk + 1) * a_dim1], lda, &
tau[kk + 1], &work[1], &iinfo);
}
if (kk > 0) {
//
// Use blocked code
//
i__1 = -nb;
for (i__ = ki + 1; i__1 < 0 ? i__ >= 1 : i__ <= 1; i__ += i__1) {
// Computing MIN
i__2 = nb, i__3 = *k - i__ + 1;
ib = min(i__2,i__3);
if (i__ + ib <= *n) {
//
// Form the triangular factor of the block reflector
// H = H(i) H(i+1) . . . H(i+ib-1)
//
i__2 = *m - i__ + 1;
dlarft_("Forward", "Columnwise", &i__2, &ib, &a[i__ + i__ *
a_dim1], lda, &tau[i__], &work[1], &ldwork);
//
// Apply H to A(i:m,i+ib:n) from the left
//
i__2 = *m - i__ + 1;
i__3 = *n - i__ - ib + 1;
dlarfb_("Left", "No transpose", "Forward", "Columnwise", &
i__2, &i__3, &ib, &a[i__ + i__ * a_dim1], lda, &work[
1], &ldwork, &a[i__ + (i__ + ib) * a_dim1], lda, &
work[ib + 1], &ldwork);
}
//
// Apply H to rows i:m of current block
//
i__2 = *m - i__ + 1;
dorg2r_(&i__2, &ib, &ib, &a[i__ + i__ * a_dim1], lda, &tau[i__], &
work[1], &iinfo);
//
// Set rows 1:i-1 of current block to zero
//
i__2 = i__ + ib - 1;
for (j = i__; j <= i__2; ++j) {
i__3 = i__ - 1;
for (l = 1; l <= i__3; ++l) {
a[l + j * a_dim1] = 0.;
// L30:
}
// L40:
}
// L50:
}
}
work[1] = (double) iws;
return 0;
//
// End of DORGQR
//
} // dorgqr_
-684
View File
@@ -1,684 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DORM2R multiplies a general matrix by the orthogonal matrix from a QR factorization determined by sgeqrf (unblocked algorithm).
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DORM2R + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dorm2r.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dorm2r.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dorm2r.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// SUBROUTINE DORM2R( SIDE, TRANS, M, N, K, A, LDA, TAU, C, LDC,
// WORK, INFO )
//
// .. Scalar Arguments ..
// CHARACTER SIDE, TRANS
// INTEGER INFO, K, LDA, LDC, M, N
// ..
// .. Array Arguments ..
// DOUBLE PRECISION A( LDA, * ), C( LDC, * ), TAU( * ), WORK( * )
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DORM2R overwrites the general real m by n matrix C with
//>
//> Q * C if SIDE = 'L' and TRANS = 'N', or
//>
//> Q**T* C if SIDE = 'L' and TRANS = 'T', or
//>
//> C * Q if SIDE = 'R' and TRANS = 'N', or
//>
//> C * Q**T if SIDE = 'R' and TRANS = 'T',
//>
//> where Q is a real orthogonal matrix defined as the product of k
//> elementary reflectors
//>
//> Q = H(1) H(2) . . . H(k)
//>
//> as returned by DGEQRF. Q is of order m if SIDE = 'L' and of order n
//> if SIDE = 'R'.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] SIDE
//> \verbatim
//> SIDE is CHARACTER*1
//> = 'L': apply Q or Q**T from the Left
//> = 'R': apply Q or Q**T from the Right
//> \endverbatim
//>
//> \param[in] TRANS
//> \verbatim
//> TRANS is CHARACTER*1
//> = 'N': apply Q (No transpose)
//> = 'T': apply Q**T (Transpose)
//> \endverbatim
//>
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> The number of rows of the matrix C. M >= 0.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> The number of columns of the matrix C. N >= 0.
//> \endverbatim
//>
//> \param[in] K
//> \verbatim
//> K is INTEGER
//> The number of elementary reflectors whose product defines
//> the matrix Q.
//> If SIDE = 'L', M >= K >= 0;
//> if SIDE = 'R', N >= K >= 0.
//> \endverbatim
//>
//> \param[in] A
//> \verbatim
//> A is DOUBLE PRECISION array, dimension (LDA,K)
//> The i-th column must contain the vector which defines the
//> elementary reflector H(i), for i = 1,2,...,k, as returned by
//> DGEQRF in the first k columns of its array argument A.
//> A is modified by the routine but restored on exit.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> The leading dimension of the array A.
//> If SIDE = 'L', LDA >= max(1,M);
//> if SIDE = 'R', LDA >= max(1,N).
//> \endverbatim
//>
//> \param[in] TAU
//> \verbatim
//> TAU is DOUBLE PRECISION array, dimension (K)
//> TAU(i) must contain the scalar factor of the elementary
//> reflector H(i), as returned by DGEQRF.
//> \endverbatim
//>
//> \param[in,out] C
//> \verbatim
//> C is DOUBLE PRECISION array, dimension (LDC,N)
//> On entry, the m by n matrix C.
//> On exit, C is overwritten by Q*C or Q**T*C or C*Q**T or C*Q.
//> \endverbatim
//>
//> \param[in] LDC
//> \verbatim
//> LDC is INTEGER
//> The leading dimension of the array C. LDC >= max(1,M).
//> \endverbatim
//>
//> \param[out] WORK
//> \verbatim
//> WORK is DOUBLE PRECISION array, dimension
//> (N) if SIDE = 'L',
//> (M) if SIDE = 'R'
//> \endverbatim
//>
//> \param[out] INFO
//> \verbatim
//> INFO is INTEGER
//> = 0: successful exit
//> < 0: if INFO = -i, the i-th argument had an illegal value
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup doubleOTHERcomputational
//
// =====================================================================
/* Subroutine */ int dorm2r_(char *side, char *trans, int *m, int *n, int *k,
double *a, int *lda, double *tau, double *c__, int *ldc, double *work,
int *info)
{
// Table of constant values
int c__1 = 1;
// System generated locals
int a_dim1, a_offset, c_dim1, c_offset, i__1, i__2;
// Local variables
int i__, i1, i2, i3, ic, jc, mi, ni, nq;
double aii;
int left;
extern /* Subroutine */ int dlarf_(char *, int *, int *, double *, int *,
double *, double *, int *, double *);
extern int lsame_(char *, char *);
extern /* Subroutine */ int xerbla_(char *, int *);
int notran;
//
// -- LAPACK computational routine (version 3.7.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. External Functions ..
// ..
// .. External Subroutines ..
// ..
// .. Intrinsic Functions ..
// ..
// .. Executable Statements ..
//
// Test the input arguments
//
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
--tau;
c_dim1 = *ldc;
c_offset = 1 + c_dim1;
c__ -= c_offset;
--work;
// Function Body
*info = 0;
left = lsame_(side, "L");
notran = lsame_(trans, "N");
//
// NQ is the order of Q
//
if (left) {
nq = *m;
} else {
nq = *n;
}
if (! left && ! lsame_(side, "R")) {
*info = -1;
} else if (! notran && ! lsame_(trans, "T")) {
*info = -2;
} else if (*m < 0) {
*info = -3;
} else if (*n < 0) {
*info = -4;
} else if (*k < 0 || *k > nq) {
*info = -5;
} else if (*lda < max(1,nq)) {
*info = -7;
} else if (*ldc < max(1,*m)) {
*info = -10;
}
if (*info != 0) {
i__1 = -(*info);
xerbla_("DORM2R", &i__1);
return 0;
}
//
// Quick return if possible
//
if (*m == 0 || *n == 0 || *k == 0) {
return 0;
}
if (left && ! notran || ! left && notran) {
i1 = 1;
i2 = *k;
i3 = 1;
} else {
i1 = *k;
i2 = 1;
i3 = -1;
}
if (left) {
ni = *n;
jc = 1;
} else {
mi = *m;
ic = 1;
}
i__1 = i2;
i__2 = i3;
for (i__ = i1; i__2 < 0 ? i__ >= i__1 : i__ <= i__1; i__ += i__2) {
if (left) {
//
// H(i) is applied to C(i:m,1:n)
//
mi = *m - i__ + 1;
ic = i__;
} else {
//
// H(i) is applied to C(1:m,i:n)
//
ni = *n - i__ + 1;
jc = i__;
}
//
// Apply H(i)
//
aii = a[i__ + i__ * a_dim1];
a[i__ + i__ * a_dim1] = 1.;
dlarf_(side, &mi, &ni, &a[i__ + i__ * a_dim1], &c__1, &tau[i__], &c__[
ic + jc * c_dim1], ldc, &work[1]);
a[i__ + i__ * a_dim1] = aii;
// L10:
}
return 0;
//
// End of DORM2R
//
} // dorm2r_
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
//> \brief \b DORMQR
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
//> \htmlonly
//> Download DORMQR + dependencies
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dormqr.f">
//> [TGZ]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dormqr.f">
//> [ZIP]</a>
//> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dormqr.f">
//> [TXT]</a>
//> \endhtmlonly
//
// Definition:
// ===========
//
// SUBROUTINE DORMQR( SIDE, TRANS, M, N, K, A, LDA, TAU, C, LDC,
// WORK, LWORK, INFO )
//
// .. Scalar Arguments ..
// CHARACTER SIDE, TRANS
// INTEGER INFO, K, LDA, LDC, LWORK, M, N
// ..
// .. Array Arguments ..
// DOUBLE PRECISION A( LDA, * ), C( LDC, * ), TAU( * ), WORK( * )
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DORMQR overwrites the general real M-by-N matrix C with
//>
//> SIDE = 'L' SIDE = 'R'
//> TRANS = 'N': Q * C C * Q
//> TRANS = 'T': Q**T * C C * Q**T
//>
//> where Q is a real orthogonal matrix defined as the product of k
//> elementary reflectors
//>
//> Q = H(1) H(2) . . . H(k)
//>
//> as returned by DGEQRF. Q is of order M if SIDE = 'L' and of order N
//> if SIDE = 'R'.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] SIDE
//> \verbatim
//> SIDE is CHARACTER*1
//> = 'L': apply Q or Q**T from the Left;
//> = 'R': apply Q or Q**T from the Right.
//> \endverbatim
//>
//> \param[in] TRANS
//> \verbatim
//> TRANS is CHARACTER*1
//> = 'N': No transpose, apply Q;
//> = 'T': Transpose, apply Q**T.
//> \endverbatim
//>
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> The number of rows of the matrix C. M >= 0.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> The number of columns of the matrix C. N >= 0.
//> \endverbatim
//>
//> \param[in] K
//> \verbatim
//> K is INTEGER
//> The number of elementary reflectors whose product defines
//> the matrix Q.
//> If SIDE = 'L', M >= K >= 0;
//> if SIDE = 'R', N >= K >= 0.
//> \endverbatim
//>
//> \param[in] A
//> \verbatim
//> A is DOUBLE PRECISION array, dimension (LDA,K)
//> The i-th column must contain the vector which defines the
//> elementary reflector H(i), for i = 1,2,...,k, as returned by
//> DGEQRF in the first k columns of its array argument A.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> The leading dimension of the array A.
//> If SIDE = 'L', LDA >= max(1,M);
//> if SIDE = 'R', LDA >= max(1,N).
//> \endverbatim
//>
//> \param[in] TAU
//> \verbatim
//> TAU is DOUBLE PRECISION array, dimension (K)
//> TAU(i) must contain the scalar factor of the elementary
//> reflector H(i), as returned by DGEQRF.
//> \endverbatim
//>
//> \param[in,out] C
//> \verbatim
//> C is DOUBLE PRECISION array, dimension (LDC,N)
//> On entry, the M-by-N matrix C.
//> On exit, C is overwritten by Q*C or Q**T*C or C*Q**T or C*Q.
//> \endverbatim
//>
//> \param[in] LDC
//> \verbatim
//> LDC is INTEGER
//> The leading dimension of the array C. LDC >= max(1,M).
//> \endverbatim
//>
//> \param[out] WORK
//> \verbatim
//> WORK is DOUBLE PRECISION array, dimension (MAX(1,LWORK))
//> On exit, if INFO = 0, WORK(1) returns the optimal LWORK.
//> \endverbatim
//>
//> \param[in] LWORK
//> \verbatim
//> LWORK is INTEGER
//> The dimension of the array WORK.
//> If SIDE = 'L', LWORK >= max(1,N);
//> if SIDE = 'R', LWORK >= max(1,M).
//> For good performance, LWORK should generally be larger.
//>
//> If LWORK = -1, then a workspace query is assumed; the routine
//> only calculates the optimal size of the WORK array, returns
//> this value as the first entry of the WORK array, and no error
//> message related to LWORK is issued by XERBLA.
//> \endverbatim
//>
//> \param[out] INFO
//> \verbatim
//> INFO is INTEGER
//> = 0: successful exit
//> < 0: if INFO = -i, the i-th argument had an illegal value
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup doubleOTHERcomputational
//
// =====================================================================
/* Subroutine */ int dormqr_(char *side, char *trans, int *m, int *n, int *k,
double *a, int *lda, double *tau, double *c__, int *ldc, double *work,
int *lwork, int *info)
{
// Table of constant values
int c__1 = 1;
int c_n1 = -1;
int c__2 = 2;
int c__65 = 65;
// System generated locals
address a__1[2];
int a_dim1, a_offset, c_dim1, c_offset, i__1, i__2, i__3[2], i__4, i__5;
char ch__1[2+1]={'\0'};
// Local variables
int i__, i1, i2, i3, ib, ic, jc, nb, mi, ni, nq, nw, iwt;
int left;
extern int lsame_(char *, char *);
int nbmin, iinfo;
extern /* Subroutine */ int dorm2r_(char *, char *, int *, int *, int *,
double *, int *, double *, double *, int *, double *, int *),
dlarfb_(char *, char *, char *, char *, int *, int *, int *,
double *, int *, double *, int *, double *, int *, double *, int *
), dlarft_(char *, char *, int *, int *, double *, int *, double *
, double *, int *), xerbla_(char *, int *);
extern int ilaenv_(int *, char *, char *, int *, int *, int *, int *);
int notran;
int ldwork, lwkopt;
int lquery;
//
// -- LAPACK computational routine (version 3.7.0) --
// -- LAPACK is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. External Functions ..
// ..
// .. External Subroutines ..
// ..
// .. Intrinsic Functions ..
// ..
// .. Executable Statements ..
//
// Test the input arguments
//
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
--tau;
c_dim1 = *ldc;
c_offset = 1 + c_dim1;
c__ -= c_offset;
--work;
// Function Body
*info = 0;
left = lsame_(side, "L");
notran = lsame_(trans, "N");
lquery = *lwork == -1;
//
// NQ is the order of Q and NW is the minimum dimension of WORK
//
if (left) {
nq = *m;
nw = *n;
} else {
nq = *n;
nw = *m;
}
if (! left && ! lsame_(side, "R")) {
*info = -1;
} else if (! notran && ! lsame_(trans, "T")) {
*info = -2;
} else if (*m < 0) {
*info = -3;
} else if (*n < 0) {
*info = -4;
} else if (*k < 0 || *k > nq) {
*info = -5;
} else if (*lda < max(1,nq)) {
*info = -7;
} else if (*ldc < max(1,*m)) {
*info = -10;
} else if (*lwork < max(1,nw) && ! lquery) {
*info = -12;
}
if (*info == 0) {
//
// Compute the workspace requirements
//
// Computing MIN
// Writing concatenation
i__3[0] = 1, a__1[0] = side;
i__3[1] = 1, a__1[1] = trans;
s_cat(ch__1, a__1, i__3, &c__2);
i__1 = 64, i__2 = ilaenv_(&c__1, "DORMQR", ch__1, m, n, k, &c_n1);
nb = min(i__1,i__2);
lwkopt = max(1,nw) * nb + 4160;
work[1] = (double) lwkopt;
}
if (*info != 0) {
i__1 = -(*info);
xerbla_("DORMQR", &i__1);
return 0;
} else if (lquery) {
return 0;
}
//
// Quick return if possible
//
if (*m == 0 || *n == 0 || *k == 0) {
work[1] = 1.;
return 0;
}
nbmin = 2;
ldwork = nw;
if (nb > 1 && nb < *k) {
if (*lwork < nw * nb + 4160) {
nb = (*lwork - 4160) / ldwork;
// Computing MAX
// Writing concatenation
i__3[0] = 1, a__1[0] = side;
i__3[1] = 1, a__1[1] = trans;
s_cat(ch__1, a__1, i__3, &c__2);
i__1 = 2, i__2 = ilaenv_(&c__2, "DORMQR", ch__1, m, n, k, &c_n1);
nbmin = max(i__1,i__2);
}
}
if (nb < nbmin || nb >= *k) {
//
// Use unblocked code
//
dorm2r_(side, trans, m, n, k, &a[a_offset], lda, &tau[1], &c__[
c_offset], ldc, &work[1], &iinfo);
} else {
//
// Use blocked code
//
iwt = nw * nb + 1;
if (left && ! notran || ! left && notran) {
i1 = 1;
i2 = *k;
i3 = nb;
} else {
i1 = (*k - 1) / nb * nb + 1;
i2 = 1;
i3 = -nb;
}
if (left) {
ni = *n;
jc = 1;
} else {
mi = *m;
ic = 1;
}
i__1 = i2;
i__2 = i3;
for (i__ = i1; i__2 < 0 ? i__ >= i__1 : i__ <= i__1; i__ += i__2) {
// Computing MIN
i__4 = nb, i__5 = *k - i__ + 1;
ib = min(i__4,i__5);
//
// Form the triangular factor of the block reflector
// H = H(i) H(i+1) . . . H(i+ib-1)
//
i__4 = nq - i__ + 1;
dlarft_("Forward", "Columnwise", &i__4, &ib, &a[i__ + i__ *
a_dim1], lda, &tau[i__], &work[iwt], &c__65);
if (left) {
//
// H or H**T is applied to C(i:m,1:n)
//
mi = *m - i__ + 1;
ic = i__;
} else {
//
// H or H**T is applied to C(1:m,i:n)
//
ni = *n - i__ + 1;
jc = i__;
}
//
// Apply H or H**T
//
dlarfb_(side, trans, "Forward", "Columnwise", &mi, &ni, &ib, &a[
i__ + i__ * a_dim1], lda, &work[iwt], &c__65, &c__[ic +
jc * c_dim1], ldc, &work[1], &ldwork);
// L10:
}
}
work[1] = (double) lwkopt;
return 0;
//
// End of DORMQR
//
} // dormqr_
-164
View File
@@ -1,164 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DROT
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
// Definition:
// ===========
//
// SUBROUTINE DROT(N,DX,INCX,DY,INCY,C,S)
//
// .. Scalar Arguments ..
// DOUBLE PRECISION C,S
// INTEGER INCX,INCY,N
// ..
// .. Array Arguments ..
// DOUBLE PRECISION DX(*),DY(*)
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DROT applies a plane rotation.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> number of elements in input vector(s)
//> \endverbatim
//>
//> \param[in,out] DX
//> \verbatim
//> DX is DOUBLE PRECISION array, dimension ( 1 + ( N - 1 )*abs( INCX ) )
//> \endverbatim
//>
//> \param[in] INCX
//> \verbatim
//> INCX is INTEGER
//> storage spacing between elements of DX
//> \endverbatim
//>
//> \param[in,out] DY
//> \verbatim
//> DY is DOUBLE PRECISION array, dimension ( 1 + ( N - 1 )*abs( INCY ) )
//> \endverbatim
//>
//> \param[in] INCY
//> \verbatim
//> INCY is INTEGER
//> storage spacing between elements of DY
//> \endverbatim
//>
//> \param[in] C
//> \verbatim
//> C is DOUBLE PRECISION
//> \endverbatim
//>
//> \param[in] S
//> \verbatim
//> S is DOUBLE PRECISION
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date November 2017
//
//> \ingroup double_blas_level1
//
//> \par Further Details:
// =====================
//>
//> \verbatim
//>
//> jack dongarra, linpack, 3/11/78.
//> modified 12/3/93, array(1) declarations changed to array(*)
//> \endverbatim
//>
// =====================================================================
/* Subroutine */ int drot_(int *n, double *dx, int *incx, double *dy, int *
incy, double *c__, double *s)
{
// System generated locals
int i__1;
// Local variables
int i__, ix, iy;
double dtemp;
//
// -- Reference BLAS level1 routine (version 3.8.0) --
// -- Reference BLAS is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// November 2017
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Local Scalars ..
// ..
// Parameter adjustments
--dy;
--dx;
// Function Body
if (*n <= 0) {
return 0;
}
if (*incx == 1 && *incy == 1) {
//
// code for both increments equal to 1
//
i__1 = *n;
for (i__ = 1; i__ <= i__1; ++i__) {
dtemp = *c__ * dx[i__] + *s * dy[i__];
dy[i__] = *c__ * dy[i__] - *s * dx[i__];
dx[i__] = dtemp;
}
} else {
//
// code for unequal increments or equal increments not equal
// to 1
//
ix = 1;
iy = 1;
if (*incx < 0) {
ix = (-(*n) + 1) * *incx + 1;
}
if (*incy < 0) {
iy = (-(*n) + 1) * *incy + 1;
}
i__1 = *n;
for (i__ = 1; i__ <= i__1; ++i__) {
dtemp = *c__ * dx[ix] + *s * dy[iy];
dy[iy] = *c__ * dy[iy] - *s * dx[ix];
dx[ix] = dtemp;
ix += *incx;
iy += *incy;
}
}
return 0;
} // drot_
-155
View File
@@ -1,155 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DSCAL
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
// Definition:
// ===========
//
// SUBROUTINE DSCAL(N,DA,DX,INCX)
//
// .. Scalar Arguments ..
// DOUBLE PRECISION DA
// INTEGER INCX,N
// ..
// .. Array Arguments ..
// DOUBLE PRECISION DX(*)
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DSCAL scales a vector by a constant.
//> uses unrolled loops for increment equal to 1.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> number of elements in input vector(s)
//> \endverbatim
//>
//> \param[in] DA
//> \verbatim
//> DA is DOUBLE PRECISION
//> On entry, DA specifies the scalar alpha.
//> \endverbatim
//>
//> \param[in,out] DX
//> \verbatim
//> DX is DOUBLE PRECISION array, dimension ( 1 + ( N - 1 )*abs( INCX ) )
//> \endverbatim
//>
//> \param[in] INCX
//> \verbatim
//> INCX is INTEGER
//> storage spacing between elements of DX
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date November 2017
//
//> \ingroup double_blas_level1
//
//> \par Further Details:
// =====================
//>
//> \verbatim
//>
//> jack dongarra, linpack, 3/11/78.
//> modified 3/93 to return if incx .le. 0.
//> modified 12/3/93, array(1) declarations changed to array(*)
//> \endverbatim
//>
// =====================================================================
/* Subroutine */ int dscal_(int *n, double *da, double *dx, int *incx)
{
// System generated locals
int i__1, i__2;
// Local variables
int i__, m, mp1, nincx;
//
// -- Reference BLAS level1 routine (version 3.8.0) --
// -- Reference BLAS is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// November 2017
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Local Scalars ..
// ..
// .. Intrinsic Functions ..
// ..
// Parameter adjustments
--dx;
// Function Body
if (*n <= 0 || *incx <= 0) {
return 0;
}
if (*incx == 1) {
//
// code for increment equal to 1
//
//
// clean-up loop
//
m = *n % 5;
if (m != 0) {
i__1 = m;
for (i__ = 1; i__ <= i__1; ++i__) {
dx[i__] = *da * dx[i__];
}
if (*n < 5) {
return 0;
}
}
mp1 = m + 1;
i__1 = *n;
for (i__ = mp1; i__ <= i__1; i__ += 5) {
dx[i__] = *da * dx[i__];
dx[i__ + 1] = *da * dx[i__ + 1];
dx[i__ + 2] = *da * dx[i__ + 2];
dx[i__ + 3] = *da * dx[i__ + 3];
dx[i__ + 4] = *da * dx[i__ + 4];
}
} else {
//
// code for increment not equal to 1
//
nincx = *n * *incx;
i__1 = nincx;
i__2 = *incx;
for (i__ = 1; i__2 < 0 ? i__ >= i__1 : i__ <= i__1; i__ += i__2) {
dx[i__] = *da * dx[i__];
}
}
return 0;
} // dscal_
-178
View File
@@ -1,178 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DSWAP
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
// Definition:
// ===========
//
// SUBROUTINE DSWAP(N,DX,INCX,DY,INCY)
//
// .. Scalar Arguments ..
// INTEGER INCX,INCY,N
// ..
// .. Array Arguments ..
// DOUBLE PRECISION DX(*),DY(*)
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DSWAP interchanges two vectors.
//> uses unrolled loops for increments equal to 1.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> number of elements in input vector(s)
//> \endverbatim
//>
//> \param[in,out] DX
//> \verbatim
//> DX is DOUBLE PRECISION array, dimension ( 1 + ( N - 1 )*abs( INCX ) )
//> \endverbatim
//>
//> \param[in] INCX
//> \verbatim
//> INCX is INTEGER
//> storage spacing between elements of DX
//> \endverbatim
//>
//> \param[in,out] DY
//> \verbatim
//> DY is DOUBLE PRECISION array, dimension ( 1 + ( N - 1 )*abs( INCY ) )
//> \endverbatim
//>
//> \param[in] INCY
//> \verbatim
//> INCY is INTEGER
//> storage spacing between elements of DY
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date November 2017
//
//> \ingroup double_blas_level1
//
//> \par Further Details:
// =====================
//>
//> \verbatim
//>
//> jack dongarra, linpack, 3/11/78.
//> modified 12/3/93, array(1) declarations changed to array(*)
//> \endverbatim
//>
// =====================================================================
/* Subroutine */ int dswap_(int *n, double *dx, int *incx, double *dy, int *
incy)
{
// System generated locals
int i__1;
// Local variables
int i__, m, ix, iy, mp1;
double dtemp;
//
// -- Reference BLAS level1 routine (version 3.8.0) --
// -- Reference BLAS is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// November 2017
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Local Scalars ..
// ..
// .. Intrinsic Functions ..
// ..
// Parameter adjustments
--dy;
--dx;
// Function Body
if (*n <= 0) {
return 0;
}
if (*incx == 1 && *incy == 1) {
//
// code for both increments equal to 1
//
//
// clean-up loop
//
m = *n % 3;
if (m != 0) {
i__1 = m;
for (i__ = 1; i__ <= i__1; ++i__) {
dtemp = dx[i__];
dx[i__] = dy[i__];
dy[i__] = dtemp;
}
if (*n < 3) {
return 0;
}
}
mp1 = m + 1;
i__1 = *n;
for (i__ = mp1; i__ <= i__1; i__ += 3) {
dtemp = dx[i__];
dx[i__] = dy[i__];
dy[i__] = dtemp;
dtemp = dx[i__ + 1];
dx[i__ + 1] = dy[i__ + 1];
dy[i__ + 1] = dtemp;
dtemp = dx[i__ + 2];
dx[i__ + 2] = dy[i__ + 2];
dy[i__ + 2] = dtemp;
}
} else {
//
// code for unequal increments or equal increments not equal
// to 1
//
ix = 1;
iy = 1;
if (*incx < 0) {
ix = (-(*n) + 1) * *incx + 1;
}
if (*incy < 0) {
iy = (-(*n) + 1) * *incy + 1;
}
i__1 = *n;
for (i__ = 1; i__ <= i__1; ++i__) {
dtemp = dx[ix];
dx[ix] = dy[iy];
dy[iy] = dtemp;
ix += *incx;
iy += *incy;
}
}
return 0;
} // dswap_
-509
View File
@@ -1,509 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DTRMM
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
// Definition:
// ===========
//
// SUBROUTINE DTRMM(SIDE,UPLO,TRANSA,DIAG,M,N,ALPHA,A,LDA,B,LDB)
//
// .. Scalar Arguments ..
// DOUBLE PRECISION ALPHA
// INTEGER LDA,LDB,M,N
// CHARACTER DIAG,SIDE,TRANSA,UPLO
// ..
// .. Array Arguments ..
// DOUBLE PRECISION A(LDA,*),B(LDB,*)
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DTRMM performs one of the matrix-matrix operations
//>
//> B := alpha*op( A )*B, or B := alpha*B*op( A ),
//>
//> where alpha is a scalar, B is an m by n matrix, A is a unit, or
//> non-unit, upper or lower triangular matrix and op( A ) is one of
//>
//> op( A ) = A or op( A ) = A**T.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] SIDE
//> \verbatim
//> SIDE is CHARACTER*1
//> On entry, SIDE specifies whether op( A ) multiplies B from
//> the left or right as follows:
//>
//> SIDE = 'L' or 'l' B := alpha*op( A )*B.
//>
//> SIDE = 'R' or 'r' B := alpha*B*op( A ).
//> \endverbatim
//>
//> \param[in] UPLO
//> \verbatim
//> UPLO is CHARACTER*1
//> On entry, UPLO specifies whether the matrix A is an upper or
//> lower triangular matrix as follows:
//>
//> UPLO = 'U' or 'u' A is an upper triangular matrix.
//>
//> UPLO = 'L' or 'l' A is a lower triangular matrix.
//> \endverbatim
//>
//> \param[in] TRANSA
//> \verbatim
//> TRANSA is CHARACTER*1
//> On entry, TRANSA specifies the form of op( A ) to be used in
//> the matrix multiplication as follows:
//>
//> TRANSA = 'N' or 'n' op( A ) = A.
//>
//> TRANSA = 'T' or 't' op( A ) = A**T.
//>
//> TRANSA = 'C' or 'c' op( A ) = A**T.
//> \endverbatim
//>
//> \param[in] DIAG
//> \verbatim
//> DIAG is CHARACTER*1
//> On entry, DIAG specifies whether or not A is unit triangular
//> as follows:
//>
//> DIAG = 'U' or 'u' A is assumed to be unit triangular.
//>
//> DIAG = 'N' or 'n' A is not assumed to be unit
//> triangular.
//> \endverbatim
//>
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> On entry, M specifies the number of rows of B. M must be at
//> least zero.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> On entry, N specifies the number of columns of B. N must be
//> at least zero.
//> \endverbatim
//>
//> \param[in] ALPHA
//> \verbatim
//> ALPHA is DOUBLE PRECISION.
//> On entry, ALPHA specifies the scalar alpha. When alpha is
//> zero then A is not referenced and B need not be set before
//> entry.
//> \endverbatim
//>
//> \param[in] A
//> \verbatim
//> A is DOUBLE PRECISION array, dimension ( LDA, k ), where k is m
//> when SIDE = 'L' or 'l' and is n when SIDE = 'R' or 'r'.
//> Before entry with UPLO = 'U' or 'u', the leading k by k
//> upper triangular part of the array A must contain the upper
//> triangular matrix and the strictly lower triangular part of
//> A is not referenced.
//> Before entry with UPLO = 'L' or 'l', the leading k by k
//> lower triangular part of the array A must contain the lower
//> triangular matrix and the strictly upper triangular part of
//> A is not referenced.
//> Note that when DIAG = 'U' or 'u', the diagonal elements of
//> A are not referenced either, but are assumed to be unity.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> On entry, LDA specifies the first dimension of A as declared
//> in the calling (sub) program. When SIDE = 'L' or 'l' then
//> LDA must be at least max( 1, m ), when SIDE = 'R' or 'r'
//> then LDA must be at least max( 1, n ).
//> \endverbatim
//>
//> \param[in,out] B
//> \verbatim
//> B is DOUBLE PRECISION array, dimension ( LDB, N )
//> Before entry, the leading m by n part of the array B must
//> contain the matrix B, and on exit is overwritten by the
//> transformed matrix.
//> \endverbatim
//>
//> \param[in] LDB
//> \verbatim
//> LDB is INTEGER
//> On entry, LDB specifies the first dimension of B as declared
//> in the calling (sub) program. LDB must be at least
//> max( 1, m ).
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup double_blas_level3
//
//> \par Further Details:
// =====================
//>
//> \verbatim
//>
//> Level 3 Blas routine.
//>
//> -- Written on 8-February-1989.
//> Jack Dongarra, Argonne National Laboratory.
//> Iain Duff, AERE Harwell.
//> Jeremy Du Croz, Numerical Algorithms Group Ltd.
//> Sven Hammarling, Numerical Algorithms Group Ltd.
//> \endverbatim
//>
// =====================================================================
/* Subroutine */ int dtrmm_(char *side, char *uplo, char *transa, char *diag,
int *m, int *n, double *alpha, double *a, int *lda, double *b, int *
ldb)
{
// System generated locals
int a_dim1, a_offset, b_dim1, b_offset, i__1, i__2, i__3;
// Local variables
int i__, j, k, info;
double temp;
int lside;
extern int lsame_(char *, char *);
int nrowa;
int upper;
extern /* Subroutine */ int xerbla_(char *, int *);
int nounit;
//
// -- Reference BLAS level3 routine (version 3.7.0) --
// -- Reference BLAS is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. External Functions ..
// ..
// .. External Subroutines ..
// ..
// .. Intrinsic Functions ..
// ..
// .. Local Scalars ..
// ..
// .. Parameters ..
// ..
//
// Test the input parameters.
//
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
b_dim1 = *ldb;
b_offset = 1 + b_dim1;
b -= b_offset;
// Function Body
lside = lsame_(side, "L");
if (lside) {
nrowa = *m;
} else {
nrowa = *n;
}
nounit = lsame_(diag, "N");
upper = lsame_(uplo, "U");
info = 0;
if (! lside && ! lsame_(side, "R")) {
info = 1;
} else if (! upper && ! lsame_(uplo, "L")) {
info = 2;
} else if (! lsame_(transa, "N") && ! lsame_(transa, "T") && ! lsame_(
transa, "C")) {
info = 3;
} else if (! lsame_(diag, "U") && ! lsame_(diag, "N")) {
info = 4;
} else if (*m < 0) {
info = 5;
} else if (*n < 0) {
info = 6;
} else if (*lda < max(1,nrowa)) {
info = 9;
} else if (*ldb < max(1,*m)) {
info = 11;
}
if (info != 0) {
xerbla_("DTRMM ", &info);
return 0;
}
//
// Quick return if possible.
//
if (*m == 0 || *n == 0) {
return 0;
}
//
// And when alpha.eq.zero.
//
if (*alpha == 0.) {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
b[i__ + j * b_dim1] = 0.;
// L10:
}
// L20:
}
return 0;
}
//
// Start the operations.
//
if (lside) {
if (lsame_(transa, "N")) {
//
// Form B := alpha*A*B.
//
if (upper) {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (k = 1; k <= i__2; ++k) {
if (b[k + j * b_dim1] != 0.) {
temp = *alpha * b[k + j * b_dim1];
i__3 = k - 1;
for (i__ = 1; i__ <= i__3; ++i__) {
b[i__ + j * b_dim1] += temp * a[i__ + k *
a_dim1];
// L30:
}
if (nounit) {
temp *= a[k + k * a_dim1];
}
b[k + j * b_dim1] = temp;
}
// L40:
}
// L50:
}
} else {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
for (k = *m; k >= 1; --k) {
if (b[k + j * b_dim1] != 0.) {
temp = *alpha * b[k + j * b_dim1];
b[k + j * b_dim1] = temp;
if (nounit) {
b[k + j * b_dim1] *= a[k + k * a_dim1];
}
i__2 = *m;
for (i__ = k + 1; i__ <= i__2; ++i__) {
b[i__ + j * b_dim1] += temp * a[i__ + k *
a_dim1];
// L60:
}
}
// L70:
}
// L80:
}
}
} else {
//
// Form B := alpha*A**T*B.
//
if (upper) {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
for (i__ = *m; i__ >= 1; --i__) {
temp = b[i__ + j * b_dim1];
if (nounit) {
temp *= a[i__ + i__ * a_dim1];
}
i__2 = i__ - 1;
for (k = 1; k <= i__2; ++k) {
temp += a[k + i__ * a_dim1] * b[k + j * b_dim1];
// L90:
}
b[i__ + j * b_dim1] = *alpha * temp;
// L100:
}
// L110:
}
} else {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp = b[i__ + j * b_dim1];
if (nounit) {
temp *= a[i__ + i__ * a_dim1];
}
i__3 = *m;
for (k = i__ + 1; k <= i__3; ++k) {
temp += a[k + i__ * a_dim1] * b[k + j * b_dim1];
// L120:
}
b[i__ + j * b_dim1] = *alpha * temp;
// L130:
}
// L140:
}
}
}
} else {
if (lsame_(transa, "N")) {
//
// Form B := alpha*B*A.
//
if (upper) {
for (j = *n; j >= 1; --j) {
temp = *alpha;
if (nounit) {
temp *= a[j + j * a_dim1];
}
i__1 = *m;
for (i__ = 1; i__ <= i__1; ++i__) {
b[i__ + j * b_dim1] = temp * b[i__ + j * b_dim1];
// L150:
}
i__1 = j - 1;
for (k = 1; k <= i__1; ++k) {
if (a[k + j * a_dim1] != 0.) {
temp = *alpha * a[k + j * a_dim1];
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
b[i__ + j * b_dim1] += temp * b[i__ + k *
b_dim1];
// L160:
}
}
// L170:
}
// L180:
}
} else {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
temp = *alpha;
if (nounit) {
temp *= a[j + j * a_dim1];
}
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
b[i__ + j * b_dim1] = temp * b[i__ + j * b_dim1];
// L190:
}
i__2 = *n;
for (k = j + 1; k <= i__2; ++k) {
if (a[k + j * a_dim1] != 0.) {
temp = *alpha * a[k + j * a_dim1];
i__3 = *m;
for (i__ = 1; i__ <= i__3; ++i__) {
b[i__ + j * b_dim1] += temp * b[i__ + k *
b_dim1];
// L200:
}
}
// L210:
}
// L220:
}
}
} else {
//
// Form B := alpha*B*A**T.
//
if (upper) {
i__1 = *n;
for (k = 1; k <= i__1; ++k) {
i__2 = k - 1;
for (j = 1; j <= i__2; ++j) {
if (a[j + k * a_dim1] != 0.) {
temp = *alpha * a[j + k * a_dim1];
i__3 = *m;
for (i__ = 1; i__ <= i__3; ++i__) {
b[i__ + j * b_dim1] += temp * b[i__ + k *
b_dim1];
// L230:
}
}
// L240:
}
temp = *alpha;
if (nounit) {
temp *= a[k + k * a_dim1];
}
if (temp != 1.) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
b[i__ + k * b_dim1] = temp * b[i__ + k * b_dim1];
// L250:
}
}
// L260:
}
} else {
for (k = *n; k >= 1; --k) {
i__1 = *n;
for (j = k + 1; j <= i__1; ++j) {
if (a[j + k * a_dim1] != 0.) {
temp = *alpha * a[j + k * a_dim1];
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
b[i__ + j * b_dim1] += temp * b[i__ + k *
b_dim1];
// L270:
}
}
// L280:
}
temp = *alpha;
if (nounit) {
temp *= a[k + k * a_dim1];
}
if (temp != 1.) {
i__1 = *m;
for (i__ = 1; i__ <= i__1; ++i__) {
b[i__ + k * b_dim1] = temp * b[i__ + k * b_dim1];
// L290:
}
}
// L300:
}
}
}
}
return 0;
//
// End of DTRMM .
//
} // dtrmm_
-396
View File
@@ -1,396 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b DTRMV
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
// Definition:
// ===========
//
// SUBROUTINE DTRMV(UPLO,TRANS,DIAG,N,A,LDA,X,INCX)
//
// .. Scalar Arguments ..
// INTEGER INCX,LDA,N
// CHARACTER DIAG,TRANS,UPLO
// ..
// .. Array Arguments ..
// DOUBLE PRECISION A(LDA,*),X(*)
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> DTRMV performs one of the matrix-vector operations
//>
//> x := A*x, or x := A**T*x,
//>
//> where x is an n element vector and A is an n by n unit, or non-unit,
//> upper or lower triangular matrix.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] UPLO
//> \verbatim
//> UPLO is CHARACTER*1
//> On entry, UPLO specifies whether the matrix is an upper or
//> lower triangular matrix as follows:
//>
//> UPLO = 'U' or 'u' A is an upper triangular matrix.
//>
//> UPLO = 'L' or 'l' A is a lower triangular matrix.
//> \endverbatim
//>
//> \param[in] TRANS
//> \verbatim
//> TRANS is CHARACTER*1
//> On entry, TRANS specifies the operation to be performed as
//> follows:
//>
//> TRANS = 'N' or 'n' x := A*x.
//>
//> TRANS = 'T' or 't' x := A**T*x.
//>
//> TRANS = 'C' or 'c' x := A**T*x.
//> \endverbatim
//>
//> \param[in] DIAG
//> \verbatim
//> DIAG is CHARACTER*1
//> On entry, DIAG specifies whether or not A is unit
//> triangular as follows:
//>
//> DIAG = 'U' or 'u' A is assumed to be unit triangular.
//>
//> DIAG = 'N' or 'n' A is not assumed to be unit
//> triangular.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> On entry, N specifies the order of the matrix A.
//> N must be at least zero.
//> \endverbatim
//>
//> \param[in] A
//> \verbatim
//> A is DOUBLE PRECISION array, dimension ( LDA, N )
//> Before entry with UPLO = 'U' or 'u', the leading n by n
//> upper triangular part of the array A must contain the upper
//> triangular matrix and the strictly lower triangular part of
//> A is not referenced.
//> Before entry with UPLO = 'L' or 'l', the leading n by n
//> lower triangular part of the array A must contain the lower
//> triangular matrix and the strictly upper triangular part of
//> A is not referenced.
//> Note that when DIAG = 'U' or 'u', the diagonal elements of
//> A are not referenced either, but are assumed to be unity.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> On entry, LDA specifies the first dimension of A as declared
//> in the calling (sub) program. LDA must be at least
//> max( 1, n ).
//> \endverbatim
//>
//> \param[in,out] X
//> \verbatim
//> X is DOUBLE PRECISION array, dimension at least
//> ( 1 + ( n - 1 )*abs( INCX ) ).
//> Before entry, the incremented array X must contain the n
//> element vector x. On exit, X is overwritten with the
//> transformed vector x.
//> \endverbatim
//>
//> \param[in] INCX
//> \verbatim
//> INCX is INTEGER
//> On entry, INCX specifies the increment for the elements of
//> X. INCX must not be zero.
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup double_blas_level2
//
//> \par Further Details:
// =====================
//>
//> \verbatim
//>
//> Level 2 Blas routine.
//> The vector and matrix arguments are not referenced when N = 0, or M = 0
//>
//> -- Written on 22-October-1986.
//> Jack Dongarra, Argonne National Lab.
//> Jeremy Du Croz, Nag Central Office.
//> Sven Hammarling, Nag Central Office.
//> Richard Hanson, Sandia National Labs.
//> \endverbatim
//>
// =====================================================================
/* Subroutine */ int dtrmv_(char *uplo, char *trans, char *diag, int *n,
double *a, int *lda, double *x, int *incx)
{
// System generated locals
int a_dim1, a_offset, i__1, i__2;
// Local variables
int i__, j, ix, jx, kx, info;
double temp;
extern int lsame_(char *, char *);
extern /* Subroutine */ int xerbla_(char *, int *);
int nounit;
//
// -- Reference BLAS level2 routine (version 3.7.0) --
// -- Reference BLAS is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. Parameters ..
// ..
// .. Local Scalars ..
// ..
// .. External Functions ..
// ..
// .. External Subroutines ..
// ..
// .. Intrinsic Functions ..
// ..
//
// Test the input parameters.
//
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
--x;
// Function Body
info = 0;
if (! lsame_(uplo, "U") && ! lsame_(uplo, "L")) {
info = 1;
} else if (! lsame_(trans, "N") && ! lsame_(trans, "T") && ! lsame_(trans,
"C")) {
info = 2;
} else if (! lsame_(diag, "U") && ! lsame_(diag, "N")) {
info = 3;
} else if (*n < 0) {
info = 4;
} else if (*lda < max(1,*n)) {
info = 6;
} else if (*incx == 0) {
info = 8;
}
if (info != 0) {
xerbla_("DTRMV ", &info);
return 0;
}
//
// Quick return if possible.
//
if (*n == 0) {
return 0;
}
nounit = lsame_(diag, "N");
//
// Set up the start point in X if the increment is not unity. This
// will be ( N - 1 )*INCX too small for descending loops.
//
if (*incx <= 0) {
kx = 1 - (*n - 1) * *incx;
} else if (*incx != 1) {
kx = 1;
}
//
// Start the operations. In this version the elements of A are
// accessed sequentially with one pass through A.
//
if (lsame_(trans, "N")) {
//
// Form x := A*x.
//
if (lsame_(uplo, "U")) {
if (*incx == 1) {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
if (x[j] != 0.) {
temp = x[j];
i__2 = j - 1;
for (i__ = 1; i__ <= i__2; ++i__) {
x[i__] += temp * a[i__ + j * a_dim1];
// L10:
}
if (nounit) {
x[j] *= a[j + j * a_dim1];
}
}
// L20:
}
} else {
jx = kx;
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
if (x[jx] != 0.) {
temp = x[jx];
ix = kx;
i__2 = j - 1;
for (i__ = 1; i__ <= i__2; ++i__) {
x[ix] += temp * a[i__ + j * a_dim1];
ix += *incx;
// L30:
}
if (nounit) {
x[jx] *= a[j + j * a_dim1];
}
}
jx += *incx;
// L40:
}
}
} else {
if (*incx == 1) {
for (j = *n; j >= 1; --j) {
if (x[j] != 0.) {
temp = x[j];
i__1 = j + 1;
for (i__ = *n; i__ >= i__1; --i__) {
x[i__] += temp * a[i__ + j * a_dim1];
// L50:
}
if (nounit) {
x[j] *= a[j + j * a_dim1];
}
}
// L60:
}
} else {
kx += (*n - 1) * *incx;
jx = kx;
for (j = *n; j >= 1; --j) {
if (x[jx] != 0.) {
temp = x[jx];
ix = kx;
i__1 = j + 1;
for (i__ = *n; i__ >= i__1; --i__) {
x[ix] += temp * a[i__ + j * a_dim1];
ix -= *incx;
// L70:
}
if (nounit) {
x[jx] *= a[j + j * a_dim1];
}
}
jx -= *incx;
// L80:
}
}
}
} else {
//
// Form x := A**T*x.
//
if (lsame_(uplo, "U")) {
if (*incx == 1) {
for (j = *n; j >= 1; --j) {
temp = x[j];
if (nounit) {
temp *= a[j + j * a_dim1];
}
for (i__ = j - 1; i__ >= 1; --i__) {
temp += a[i__ + j * a_dim1] * x[i__];
// L90:
}
x[j] = temp;
// L100:
}
} else {
jx = kx + (*n - 1) * *incx;
for (j = *n; j >= 1; --j) {
temp = x[jx];
ix = jx;
if (nounit) {
temp *= a[j + j * a_dim1];
}
for (i__ = j - 1; i__ >= 1; --i__) {
ix -= *incx;
temp += a[i__ + j * a_dim1] * x[ix];
// L110:
}
x[jx] = temp;
jx -= *incx;
// L120:
}
}
} else {
if (*incx == 1) {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
temp = x[j];
if (nounit) {
temp *= a[j + j * a_dim1];
}
i__2 = *n;
for (i__ = j + 1; i__ <= i__2; ++i__) {
temp += a[i__ + j * a_dim1] * x[i__];
// L130:
}
x[j] = temp;
// L140:
}
} else {
jx = kx;
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
temp = x[jx];
ix = jx;
if (nounit) {
temp *= a[j + j * a_dim1];
}
i__2 = *n;
for (i__ = j + 1; i__ <= i__2; ++i__) {
ix += *incx;
temp += a[i__ + j * a_dim1] * x[ix];
// L150:
}
x[jx] = temp;
jx += *incx;
// L160:
}
}
}
}
return 0;
//
// End of DTRMV .
//
} // dtrmv_
-1334
View File
File diff suppressed because it is too large Load Diff
-444
View File
@@ -1,444 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b SGEMM
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
// Definition:
// ===========
//
// SUBROUTINE SGEMM(TRANSA,TRANSB,M,N,K,ALPHA,A,LDA,B,LDB,BETA,C,LDC)
//
// .. Scalar Arguments ..
// REAL ALPHA,BETA
// INTEGER K,LDA,LDB,LDC,M,N
// CHARACTER TRANSA,TRANSB
// ..
// .. Array Arguments ..
// REAL A(LDA,*),B(LDB,*),C(LDC,*)
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> SGEMM performs one of the matrix-matrix operations
//>
//> C := alpha*op( A )*op( B ) + beta*C,
//>
//> where op( X ) is one of
//>
//> op( X ) = X or op( X ) = X**T,
//>
//> alpha and beta are scalars, and A, B and C are matrices, with op( A )
//> an m by k matrix, op( B ) a k by n matrix and C an m by n matrix.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] TRANSA
//> \verbatim
//> TRANSA is CHARACTER*1
//> On entry, TRANSA specifies the form of op( A ) to be used in
//> the matrix multiplication as follows:
//>
//> TRANSA = 'N' or 'n', op( A ) = A.
//>
//> TRANSA = 'T' or 't', op( A ) = A**T.
//>
//> TRANSA = 'C' or 'c', op( A ) = A**T.
//> \endverbatim
//>
//> \param[in] TRANSB
//> \verbatim
//> TRANSB is CHARACTER*1
//> On entry, TRANSB specifies the form of op( B ) to be used in
//> the matrix multiplication as follows:
//>
//> TRANSB = 'N' or 'n', op( B ) = B.
//>
//> TRANSB = 'T' or 't', op( B ) = B**T.
//>
//> TRANSB = 'C' or 'c', op( B ) = B**T.
//> \endverbatim
//>
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> On entry, M specifies the number of rows of the matrix
//> op( A ) and of the matrix C. M must be at least zero.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> On entry, N specifies the number of columns of the matrix
//> op( B ) and the number of columns of the matrix C. N must be
//> at least zero.
//> \endverbatim
//>
//> \param[in] K
//> \verbatim
//> K is INTEGER
//> On entry, K specifies the number of columns of the matrix
//> op( A ) and the number of rows of the matrix op( B ). K must
//> be at least zero.
//> \endverbatim
//>
//> \param[in] ALPHA
//> \verbatim
//> ALPHA is REAL
//> On entry, ALPHA specifies the scalar alpha.
//> \endverbatim
//>
//> \param[in] A
//> \verbatim
//> A is REAL array, dimension ( LDA, ka ), where ka is
//> k when TRANSA = 'N' or 'n', and is m otherwise.
//> Before entry with TRANSA = 'N' or 'n', the leading m by k
//> part of the array A must contain the matrix A, otherwise
//> the leading k by m part of the array A must contain the
//> matrix A.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> On entry, LDA specifies the first dimension of A as declared
//> in the calling (sub) program. When TRANSA = 'N' or 'n' then
//> LDA must be at least max( 1, m ), otherwise LDA must be at
//> least max( 1, k ).
//> \endverbatim
//>
//> \param[in] B
//> \verbatim
//> B is REAL array, dimension ( LDB, kb ), where kb is
//> n when TRANSB = 'N' or 'n', and is k otherwise.
//> Before entry with TRANSB = 'N' or 'n', the leading k by n
//> part of the array B must contain the matrix B, otherwise
//> the leading n by k part of the array B must contain the
//> matrix B.
//> \endverbatim
//>
//> \param[in] LDB
//> \verbatim
//> LDB is INTEGER
//> On entry, LDB specifies the first dimension of B as declared
//> in the calling (sub) program. When TRANSB = 'N' or 'n' then
//> LDB must be at least max( 1, k ), otherwise LDB must be at
//> least max( 1, n ).
//> \endverbatim
//>
//> \param[in] BETA
//> \verbatim
//> BETA is REAL
//> On entry, BETA specifies the scalar beta. When BETA is
//> supplied as zero then C need not be set on input.
//> \endverbatim
//>
//> \param[in,out] C
//> \verbatim
//> C is REAL array, dimension ( LDC, N )
//> Before entry, the leading m by n part of the array C must
//> contain the matrix C, except when beta is zero, in which
//> case C need not be set on entry.
//> On exit, the array C is overwritten by the m by n matrix
//> ( alpha*op( A )*op( B ) + beta*C ).
//> \endverbatim
//>
//> \param[in] LDC
//> \verbatim
//> LDC is INTEGER
//> On entry, LDC specifies the first dimension of C as declared
//> in the calling (sub) program. LDC must be at least
//> max( 1, m ).
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup single_blas_level3
//
//> \par Further Details:
// =====================
//>
//> \verbatim
//>
//> Level 3 Blas routine.
//>
//> -- Written on 8-February-1989.
//> Jack Dongarra, Argonne National Laboratory.
//> Iain Duff, AERE Harwell.
//> Jeremy Du Croz, Numerical Algorithms Group Ltd.
//> Sven Hammarling, Numerical Algorithms Group Ltd.
//> \endverbatim
//>
// =====================================================================
/* Subroutine */ int sgemm_(char *transa, char *transb, int *m, int *n, int *
k, float *alpha, float *a, int *lda, float *b, int *ldb, float *beta,
float *c__, int *ldc)
{
// System generated locals
int a_dim1, a_offset, b_dim1, b_offset, c_dim1, c_offset, i__1, i__2,
i__3;
// Local variables
int i__, j, l, info;
int nota, notb;
float temp;
int ncola;
extern int lsame_(char *, char *);
int nrowa, nrowb;
extern /* Subroutine */ int xerbla_(char *, int *);
//
// -- Reference BLAS level3 routine (version 3.7.0) --
// -- Reference BLAS is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. External Functions ..
// ..
// .. External Subroutines ..
// ..
// .. Intrinsic Functions ..
// ..
// .. Local Scalars ..
// ..
// .. Parameters ..
// ..
//
// Set NOTA and NOTB as true if A and B respectively are not
// transposed and set NROWA, NCOLA and NROWB as the number of rows
// and columns of A and the number of rows of B respectively.
//
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
b_dim1 = *ldb;
b_offset = 1 + b_dim1;
b -= b_offset;
c_dim1 = *ldc;
c_offset = 1 + c_dim1;
c__ -= c_offset;
// Function Body
nota = lsame_(transa, "N");
notb = lsame_(transb, "N");
if (nota) {
nrowa = *m;
ncola = *k;
} else {
nrowa = *k;
ncola = *m;
}
if (notb) {
nrowb = *k;
} else {
nrowb = *n;
}
//
// Test the input parameters.
//
info = 0;
if (! nota && ! lsame_(transa, "C") && ! lsame_(transa, "T")) {
info = 1;
} else if (! notb && ! lsame_(transb, "C") && ! lsame_(transb, "T")) {
info = 2;
} else if (*m < 0) {
info = 3;
} else if (*n < 0) {
info = 4;
} else if (*k < 0) {
info = 5;
} else if (*lda < max(1,nrowa)) {
info = 8;
} else if (*ldb < max(1,nrowb)) {
info = 10;
} else if (*ldc < max(1,*m)) {
info = 13;
}
if (info != 0) {
xerbla_("SGEMM ", &info);
return 0;
}
//
// Quick return if possible.
//
if (*m == 0 || *n == 0 || (*alpha == 0.f || *k == 0) && *beta == 1.f) {
return 0;
}
//
// And if alpha.eq.zero.
//
if (*alpha == 0.f) {
if (*beta == 0.f) {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[i__ + j * c_dim1] = 0.f;
// L10:
}
// L20:
}
} else {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[i__ + j * c_dim1] = *beta * c__[i__ + j * c_dim1];
// L30:
}
// L40:
}
}
return 0;
}
//
// Start the operations.
//
if (notb) {
if (nota) {
//
// Form C := alpha*A*B + beta*C.
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
if (*beta == 0.f) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[i__ + j * c_dim1] = 0.f;
// L50:
}
} else if (*beta != 1.f) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[i__ + j * c_dim1] = *beta * c__[i__ + j * c_dim1];
// L60:
}
}
i__2 = *k;
for (l = 1; l <= i__2; ++l) {
temp = *alpha * b[l + j * b_dim1];
i__3 = *m;
for (i__ = 1; i__ <= i__3; ++i__) {
c__[i__ + j * c_dim1] += temp * a[i__ + l * a_dim1];
// L70:
}
// L80:
}
// L90:
}
} else {
//
// Form C := alpha*A**T*B + beta*C
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp = 0.f;
i__3 = *k;
for (l = 1; l <= i__3; ++l) {
temp += a[l + i__ * a_dim1] * b[l + j * b_dim1];
// L100:
}
if (*beta == 0.f) {
c__[i__ + j * c_dim1] = *alpha * temp;
} else {
c__[i__ + j * c_dim1] = *alpha * temp + *beta * c__[
i__ + j * c_dim1];
}
// L110:
}
// L120:
}
}
} else {
if (nota) {
//
// Form C := alpha*A*B**T + beta*C
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
if (*beta == 0.f) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[i__ + j * c_dim1] = 0.f;
// L130:
}
} else if (*beta != 1.f) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
c__[i__ + j * c_dim1] = *beta * c__[i__ + j * c_dim1];
// L140:
}
}
i__2 = *k;
for (l = 1; l <= i__2; ++l) {
temp = *alpha * b[j + l * b_dim1];
i__3 = *m;
for (i__ = 1; i__ <= i__3; ++i__) {
c__[i__ + j * c_dim1] += temp * a[i__ + l * a_dim1];
// L150:
}
// L160:
}
// L170:
}
} else {
//
// Form C := alpha*A**T*B**T + beta*C
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp = 0.f;
i__3 = *k;
for (l = 1; l <= i__3; ++l) {
temp += a[l + i__ * a_dim1] * b[j + l * b_dim1];
// L180:
}
if (*beta == 0.f) {
c__[i__ + j * c_dim1] = *alpha * temp;
} else {
c__[i__ + j * c_dim1] = *alpha * temp + *beta * c__[
i__ + j * c_dim1];
}
// L190:
}
// L200:
}
}
}
return 0;
//
// End of SGEMM .
//
} // sgemm_
-752
View File
@@ -1,752 +0,0 @@
/* -- translated by f2c (version 20201020 (for_lapack)). -- */
#include "f2c.h"
//> \brief \b ZGEMM
//
// =========== DOCUMENTATION ===========
//
// Online html documentation available at
// http://www.netlib.org/lapack/explore-html/
//
// Definition:
// ===========
//
// SUBROUTINE ZGEMM(TRANSA,TRANSB,M,N,K,ALPHA,A,LDA,B,LDB,BETA,C,LDC)
//
// .. Scalar Arguments ..
// COMPLEX*16 ALPHA,BETA
// INTEGER K,LDA,LDB,LDC,M,N
// CHARACTER TRANSA,TRANSB
// ..
// .. Array Arguments ..
// COMPLEX*16 A(LDA,*),B(LDB,*),C(LDC,*)
// ..
//
//
//> \par Purpose:
// =============
//>
//> \verbatim
//>
//> ZGEMM performs one of the matrix-matrix operations
//>
//> C := alpha*op( A )*op( B ) + beta*C,
//>
//> where op( X ) is one of
//>
//> op( X ) = X or op( X ) = X**T or op( X ) = X**H,
//>
//> alpha and beta are scalars, and A, B and C are matrices, with op( A )
//> an m by k matrix, op( B ) a k by n matrix and C an m by n matrix.
//> \endverbatim
//
// Arguments:
// ==========
//
//> \param[in] TRANSA
//> \verbatim
//> TRANSA is CHARACTER*1
//> On entry, TRANSA specifies the form of op( A ) to be used in
//> the matrix multiplication as follows:
//>
//> TRANSA = 'N' or 'n', op( A ) = A.
//>
//> TRANSA = 'T' or 't', op( A ) = A**T.
//>
//> TRANSA = 'C' or 'c', op( A ) = A**H.
//> \endverbatim
//>
//> \param[in] TRANSB
//> \verbatim
//> TRANSB is CHARACTER*1
//> On entry, TRANSB specifies the form of op( B ) to be used in
//> the matrix multiplication as follows:
//>
//> TRANSB = 'N' or 'n', op( B ) = B.
//>
//> TRANSB = 'T' or 't', op( B ) = B**T.
//>
//> TRANSB = 'C' or 'c', op( B ) = B**H.
//> \endverbatim
//>
//> \param[in] M
//> \verbatim
//> M is INTEGER
//> On entry, M specifies the number of rows of the matrix
//> op( A ) and of the matrix C. M must be at least zero.
//> \endverbatim
//>
//> \param[in] N
//> \verbatim
//> N is INTEGER
//> On entry, N specifies the number of columns of the matrix
//> op( B ) and the number of columns of the matrix C. N must be
//> at least zero.
//> \endverbatim
//>
//> \param[in] K
//> \verbatim
//> K is INTEGER
//> On entry, K specifies the number of columns of the matrix
//> op( A ) and the number of rows of the matrix op( B ). K must
//> be at least zero.
//> \endverbatim
//>
//> \param[in] ALPHA
//> \verbatim
//> ALPHA is COMPLEX*16
//> On entry, ALPHA specifies the scalar alpha.
//> \endverbatim
//>
//> \param[in] A
//> \verbatim
//> A is COMPLEX*16 array, dimension ( LDA, ka ), where ka is
//> k when TRANSA = 'N' or 'n', and is m otherwise.
//> Before entry with TRANSA = 'N' or 'n', the leading m by k
//> part of the array A must contain the matrix A, otherwise
//> the leading k by m part of the array A must contain the
//> matrix A.
//> \endverbatim
//>
//> \param[in] LDA
//> \verbatim
//> LDA is INTEGER
//> On entry, LDA specifies the first dimension of A as declared
//> in the calling (sub) program. When TRANSA = 'N' or 'n' then
//> LDA must be at least max( 1, m ), otherwise LDA must be at
//> least max( 1, k ).
//> \endverbatim
//>
//> \param[in] B
//> \verbatim
//> B is COMPLEX*16 array, dimension ( LDB, kb ), where kb is
//> n when TRANSB = 'N' or 'n', and is k otherwise.
//> Before entry with TRANSB = 'N' or 'n', the leading k by n
//> part of the array B must contain the matrix B, otherwise
//> the leading n by k part of the array B must contain the
//> matrix B.
//> \endverbatim
//>
//> \param[in] LDB
//> \verbatim
//> LDB is INTEGER
//> On entry, LDB specifies the first dimension of B as declared
//> in the calling (sub) program. When TRANSB = 'N' or 'n' then
//> LDB must be at least max( 1, k ), otherwise LDB must be at
//> least max( 1, n ).
//> \endverbatim
//>
//> \param[in] BETA
//> \verbatim
//> BETA is COMPLEX*16
//> On entry, BETA specifies the scalar beta. When BETA is
//> supplied as zero then C need not be set on input.
//> \endverbatim
//>
//> \param[in,out] C
//> \verbatim
//> C is COMPLEX*16 array, dimension ( LDC, N )
//> Before entry, the leading m by n part of the array C must
//> contain the matrix C, except when beta is zero, in which
//> case C need not be set on entry.
//> On exit, the array C is overwritten by the m by n matrix
//> ( alpha*op( A )*op( B ) + beta*C ).
//> \endverbatim
//>
//> \param[in] LDC
//> \verbatim
//> LDC is INTEGER
//> On entry, LDC specifies the first dimension of C as declared
//> in the calling (sub) program. LDC must be at least
//> max( 1, m ).
//> \endverbatim
//
// Authors:
// ========
//
//> \author Univ. of Tennessee
//> \author Univ. of California Berkeley
//> \author Univ. of Colorado Denver
//> \author NAG Ltd.
//
//> \date December 2016
//
//> \ingroup complex16_blas_level3
//
//> \par Further Details:
// =====================
//>
//> \verbatim
//>
//> Level 3 Blas routine.
//>
//> -- Written on 8-February-1989.
//> Jack Dongarra, Argonne National Laboratory.
//> Iain Duff, AERE Harwell.
//> Jeremy Du Croz, Numerical Algorithms Group Ltd.
//> Sven Hammarling, Numerical Algorithms Group Ltd.
//> \endverbatim
//>
// =====================================================================
/* Subroutine */ int zgemm_(char *transa, char *transb, int *m, int *n, int *
k, doublecomplex *alpha, doublecomplex *a, int *lda, doublecomplex *b,
int *ldb, doublecomplex *beta, doublecomplex *c__, int *ldc)
{
// Table of constant values
doublecomplex c_b1 = {1.,0.};
doublecomplex c_b2 = {0.,0.};
// System generated locals
int a_dim1, a_offset, b_dim1, b_offset, c_dim1, c_offset, i__1, i__2,
i__3, i__4, i__5, i__6;
doublecomplex z__1, z__2, z__3, z__4;
// Local variables
int i__, j, l, info;
int nota, notb;
doublecomplex temp;
int conja, conjb;
int ncola;
extern int lsame_(char *, char *);
int nrowa, nrowb;
extern /* Subroutine */ int xerbla_(char *, int *);
//
// -- Reference BLAS level3 routine (version 3.7.0) --
// -- Reference BLAS is a software package provided by Univ. of Tennessee, --
// -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
// December 2016
//
// .. Scalar Arguments ..
// ..
// .. Array Arguments ..
// ..
//
// =====================================================================
//
// .. External Functions ..
// ..
// .. External Subroutines ..
// ..
// .. Intrinsic Functions ..
// ..
// .. Local Scalars ..
// ..
// .. Parameters ..
// ..
//
// Set NOTA and NOTB as true if A and B respectively are not
// conjugated or transposed, set CONJA and CONJB as true if A and
// B respectively are to be transposed but not conjugated and set
// NROWA, NCOLA and NROWB as the number of rows and columns of A
// and the number of rows of B respectively.
//
// Parameter adjustments
a_dim1 = *lda;
a_offset = 1 + a_dim1;
a -= a_offset;
b_dim1 = *ldb;
b_offset = 1 + b_dim1;
b -= b_offset;
c_dim1 = *ldc;
c_offset = 1 + c_dim1;
c__ -= c_offset;
// Function Body
nota = lsame_(transa, "N");
notb = lsame_(transb, "N");
conja = lsame_(transa, "C");
conjb = lsame_(transb, "C");
if (nota) {
nrowa = *m;
ncola = *k;
} else {
nrowa = *k;
ncola = *m;
}
if (notb) {
nrowb = *k;
} else {
nrowb = *n;
}
//
// Test the input parameters.
//
info = 0;
if (! nota && ! conja && ! lsame_(transa, "T")) {
info = 1;
} else if (! notb && ! conjb && ! lsame_(transb, "T")) {
info = 2;
} else if (*m < 0) {
info = 3;
} else if (*n < 0) {
info = 4;
} else if (*k < 0) {
info = 5;
} else if (*lda < max(1,nrowa)) {
info = 8;
} else if (*ldb < max(1,nrowb)) {
info = 10;
} else if (*ldc < max(1,*m)) {
info = 13;
}
if (info != 0) {
xerbla_("ZGEMM ", &info);
return 0;
}
//
// Quick return if possible.
//
if (*m == 0 || *n == 0 || (alpha->r == 0. && alpha->i == 0. || *k == 0) &&
(beta->r == 1. && beta->i == 0.)) {
return 0;
}
//
// And when alpha.eq.zero.
//
if (alpha->r == 0. && alpha->i == 0.) {
if (beta->r == 0. && beta->i == 0.) {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
i__3 = i__ + j * c_dim1;
c__[i__3].r = 0., c__[i__3].i = 0.;
// L10:
}
// L20:
}
} else {
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
i__3 = i__ + j * c_dim1;
i__4 = i__ + j * c_dim1;
z__1.r = beta->r * c__[i__4].r - beta->i * c__[i__4].i,
z__1.i = beta->r * c__[i__4].i + beta->i * c__[
i__4].r;
c__[i__3].r = z__1.r, c__[i__3].i = z__1.i;
// L30:
}
// L40:
}
}
return 0;
}
//
// Start the operations.
//
if (notb) {
if (nota) {
//
// Form C := alpha*A*B + beta*C.
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
if (beta->r == 0. && beta->i == 0.) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
i__3 = i__ + j * c_dim1;
c__[i__3].r = 0., c__[i__3].i = 0.;
// L50:
}
} else if (beta->r != 1. || beta->i != 0.) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
i__3 = i__ + j * c_dim1;
i__4 = i__ + j * c_dim1;
z__1.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, z__1.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
c__[i__3].r = z__1.r, c__[i__3].i = z__1.i;
// L60:
}
}
i__2 = *k;
for (l = 1; l <= i__2; ++l) {
i__3 = l + j * b_dim1;
z__1.r = alpha->r * b[i__3].r - alpha->i * b[i__3].i,
z__1.i = alpha->r * b[i__3].i + alpha->i * b[i__3]
.r;
temp.r = z__1.r, temp.i = z__1.i;
i__3 = *m;
for (i__ = 1; i__ <= i__3; ++i__) {
i__4 = i__ + j * c_dim1;
i__5 = i__ + j * c_dim1;
i__6 = i__ + l * a_dim1;
z__2.r = temp.r * a[i__6].r - temp.i * a[i__6].i,
z__2.i = temp.r * a[i__6].i + temp.i * a[i__6]
.r;
z__1.r = c__[i__5].r + z__2.r, z__1.i = c__[i__5].i +
z__2.i;
c__[i__4].r = z__1.r, c__[i__4].i = z__1.i;
// L70:
}
// L80:
}
// L90:
}
} else if (conja) {
//
// Form C := alpha*A**H*B + beta*C.
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp.r = 0., temp.i = 0.;
i__3 = *k;
for (l = 1; l <= i__3; ++l) {
d_cnjg(&z__3, &a[l + i__ * a_dim1]);
i__4 = l + j * b_dim1;
z__2.r = z__3.r * b[i__4].r - z__3.i * b[i__4].i,
z__2.i = z__3.r * b[i__4].i + z__3.i * b[i__4]
.r;
z__1.r = temp.r + z__2.r, z__1.i = temp.i + z__2.i;
temp.r = z__1.r, temp.i = z__1.i;
// L100:
}
if (beta->r == 0. && beta->i == 0.) {
i__3 = i__ + j * c_dim1;
z__1.r = alpha->r * temp.r - alpha->i * temp.i,
z__1.i = alpha->r * temp.i + alpha->i *
temp.r;
c__[i__3].r = z__1.r, c__[i__3].i = z__1.i;
} else {
i__3 = i__ + j * c_dim1;
z__2.r = alpha->r * temp.r - alpha->i * temp.i,
z__2.i = alpha->r * temp.i + alpha->i *
temp.r;
i__4 = i__ + j * c_dim1;
z__3.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, z__3.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
z__1.r = z__2.r + z__3.r, z__1.i = z__2.i + z__3.i;
c__[i__3].r = z__1.r, c__[i__3].i = z__1.i;
}
// L110:
}
// L120:
}
} else {
//
// Form C := alpha*A**T*B + beta*C
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp.r = 0., temp.i = 0.;
i__3 = *k;
for (l = 1; l <= i__3; ++l) {
i__4 = l + i__ * a_dim1;
i__5 = l + j * b_dim1;
z__2.r = a[i__4].r * b[i__5].r - a[i__4].i * b[i__5]
.i, z__2.i = a[i__4].r * b[i__5].i + a[i__4]
.i * b[i__5].r;
z__1.r = temp.r + z__2.r, z__1.i = temp.i + z__2.i;
temp.r = z__1.r, temp.i = z__1.i;
// L130:
}
if (beta->r == 0. && beta->i == 0.) {
i__3 = i__ + j * c_dim1;
z__1.r = alpha->r * temp.r - alpha->i * temp.i,
z__1.i = alpha->r * temp.i + alpha->i *
temp.r;
c__[i__3].r = z__1.r, c__[i__3].i = z__1.i;
} else {
i__3 = i__ + j * c_dim1;
z__2.r = alpha->r * temp.r - alpha->i * temp.i,
z__2.i = alpha->r * temp.i + alpha->i *
temp.r;
i__4 = i__ + j * c_dim1;
z__3.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, z__3.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
z__1.r = z__2.r + z__3.r, z__1.i = z__2.i + z__3.i;
c__[i__3].r = z__1.r, c__[i__3].i = z__1.i;
}
// L140:
}
// L150:
}
}
} else if (nota) {
if (conjb) {
//
// Form C := alpha*A*B**H + beta*C.
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
if (beta->r == 0. && beta->i == 0.) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
i__3 = i__ + j * c_dim1;
c__[i__3].r = 0., c__[i__3].i = 0.;
// L160:
}
} else if (beta->r != 1. || beta->i != 0.) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
i__3 = i__ + j * c_dim1;
i__4 = i__ + j * c_dim1;
z__1.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, z__1.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
c__[i__3].r = z__1.r, c__[i__3].i = z__1.i;
// L170:
}
}
i__2 = *k;
for (l = 1; l <= i__2; ++l) {
d_cnjg(&z__2, &b[j + l * b_dim1]);
z__1.r = alpha->r * z__2.r - alpha->i * z__2.i, z__1.i =
alpha->r * z__2.i + alpha->i * z__2.r;
temp.r = z__1.r, temp.i = z__1.i;
i__3 = *m;
for (i__ = 1; i__ <= i__3; ++i__) {
i__4 = i__ + j * c_dim1;
i__5 = i__ + j * c_dim1;
i__6 = i__ + l * a_dim1;
z__2.r = temp.r * a[i__6].r - temp.i * a[i__6].i,
z__2.i = temp.r * a[i__6].i + temp.i * a[i__6]
.r;
z__1.r = c__[i__5].r + z__2.r, z__1.i = c__[i__5].i +
z__2.i;
c__[i__4].r = z__1.r, c__[i__4].i = z__1.i;
// L180:
}
// L190:
}
// L200:
}
} else {
//
// Form C := alpha*A*B**T + beta*C
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
if (beta->r == 0. && beta->i == 0.) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
i__3 = i__ + j * c_dim1;
c__[i__3].r = 0., c__[i__3].i = 0.;
// L210:
}
} else if (beta->r != 1. || beta->i != 0.) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
i__3 = i__ + j * c_dim1;
i__4 = i__ + j * c_dim1;
z__1.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, z__1.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
c__[i__3].r = z__1.r, c__[i__3].i = z__1.i;
// L220:
}
}
i__2 = *k;
for (l = 1; l <= i__2; ++l) {
i__3 = j + l * b_dim1;
z__1.r = alpha->r * b[i__3].r - alpha->i * b[i__3].i,
z__1.i = alpha->r * b[i__3].i + alpha->i * b[i__3]
.r;
temp.r = z__1.r, temp.i = z__1.i;
i__3 = *m;
for (i__ = 1; i__ <= i__3; ++i__) {
i__4 = i__ + j * c_dim1;
i__5 = i__ + j * c_dim1;
i__6 = i__ + l * a_dim1;
z__2.r = temp.r * a[i__6].r - temp.i * a[i__6].i,
z__2.i = temp.r * a[i__6].i + temp.i * a[i__6]
.r;
z__1.r = c__[i__5].r + z__2.r, z__1.i = c__[i__5].i +
z__2.i;
c__[i__4].r = z__1.r, c__[i__4].i = z__1.i;
// L230:
}
// L240:
}
// L250:
}
}
} else if (conja) {
if (conjb) {
//
// Form C := alpha*A**H*B**H + beta*C.
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp.r = 0., temp.i = 0.;
i__3 = *k;
for (l = 1; l <= i__3; ++l) {
d_cnjg(&z__3, &a[l + i__ * a_dim1]);
d_cnjg(&z__4, &b[j + l * b_dim1]);
z__2.r = z__3.r * z__4.r - z__3.i * z__4.i, z__2.i =
z__3.r * z__4.i + z__3.i * z__4.r;
z__1.r = temp.r + z__2.r, z__1.i = temp.i + z__2.i;
temp.r = z__1.r, temp.i = z__1.i;
// L260:
}
if (beta->r == 0. && beta->i == 0.) {
i__3 = i__ + j * c_dim1;
z__1.r = alpha->r * temp.r - alpha->i * temp.i,
z__1.i = alpha->r * temp.i + alpha->i *
temp.r;
c__[i__3].r = z__1.r, c__[i__3].i = z__1.i;
} else {
i__3 = i__ + j * c_dim1;
z__2.r = alpha->r * temp.r - alpha->i * temp.i,
z__2.i = alpha->r * temp.i + alpha->i *
temp.r;
i__4 = i__ + j * c_dim1;
z__3.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, z__3.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
z__1.r = z__2.r + z__3.r, z__1.i = z__2.i + z__3.i;
c__[i__3].r = z__1.r, c__[i__3].i = z__1.i;
}
// L270:
}
// L280:
}
} else {
//
// Form C := alpha*A**H*B**T + beta*C
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp.r = 0., temp.i = 0.;
i__3 = *k;
for (l = 1; l <= i__3; ++l) {
d_cnjg(&z__3, &a[l + i__ * a_dim1]);
i__4 = j + l * b_dim1;
z__2.r = z__3.r * b[i__4].r - z__3.i * b[i__4].i,
z__2.i = z__3.r * b[i__4].i + z__3.i * b[i__4]
.r;
z__1.r = temp.r + z__2.r, z__1.i = temp.i + z__2.i;
temp.r = z__1.r, temp.i = z__1.i;
// L290:
}
if (beta->r == 0. && beta->i == 0.) {
i__3 = i__ + j * c_dim1;
z__1.r = alpha->r * temp.r - alpha->i * temp.i,
z__1.i = alpha->r * temp.i + alpha->i *
temp.r;
c__[i__3].r = z__1.r, c__[i__3].i = z__1.i;
} else {
i__3 = i__ + j * c_dim1;
z__2.r = alpha->r * temp.r - alpha->i * temp.i,
z__2.i = alpha->r * temp.i + alpha->i *
temp.r;
i__4 = i__ + j * c_dim1;
z__3.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, z__3.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
z__1.r = z__2.r + z__3.r, z__1.i = z__2.i + z__3.i;
c__[i__3].r = z__1.r, c__[i__3].i = z__1.i;
}
// L300:
}
// L310:
}
}
} else {
if (conjb) {
//
// Form C := alpha*A**T*B**H + beta*C
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp.r = 0., temp.i = 0.;
i__3 = *k;
for (l = 1; l <= i__3; ++l) {
i__4 = l + i__ * a_dim1;
d_cnjg(&z__3, &b[j + l * b_dim1]);
z__2.r = a[i__4].r * z__3.r - a[i__4].i * z__3.i,
z__2.i = a[i__4].r * z__3.i + a[i__4].i *
z__3.r;
z__1.r = temp.r + z__2.r, z__1.i = temp.i + z__2.i;
temp.r = z__1.r, temp.i = z__1.i;
// L320:
}
if (beta->r == 0. && beta->i == 0.) {
i__3 = i__ + j * c_dim1;
z__1.r = alpha->r * temp.r - alpha->i * temp.i,
z__1.i = alpha->r * temp.i + alpha->i *
temp.r;
c__[i__3].r = z__1.r, c__[i__3].i = z__1.i;
} else {
i__3 = i__ + j * c_dim1;
z__2.r = alpha->r * temp.r - alpha->i * temp.i,
z__2.i = alpha->r * temp.i + alpha->i *
temp.r;
i__4 = i__ + j * c_dim1;
z__3.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, z__3.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
z__1.r = z__2.r + z__3.r, z__1.i = z__2.i + z__3.i;
c__[i__3].r = z__1.r, c__[i__3].i = z__1.i;
}
// L330:
}
// L340:
}
} else {
//
// Form C := alpha*A**T*B**T + beta*C
//
i__1 = *n;
for (j = 1; j <= i__1; ++j) {
i__2 = *m;
for (i__ = 1; i__ <= i__2; ++i__) {
temp.r = 0., temp.i = 0.;
i__3 = *k;
for (l = 1; l <= i__3; ++l) {
i__4 = l + i__ * a_dim1;
i__5 = j + l * b_dim1;
z__2.r = a[i__4].r * b[i__5].r - a[i__4].i * b[i__5]
.i, z__2.i = a[i__4].r * b[i__5].i + a[i__4]
.i * b[i__5].r;
z__1.r = temp.r + z__2.r, z__1.i = temp.i + z__2.i;
temp.r = z__1.r, temp.i = z__1.i;
// L350:
}
if (beta->r == 0. && beta->i == 0.) {
i__3 = i__ + j * c_dim1;
z__1.r = alpha->r * temp.r - alpha->i * temp.i,
z__1.i = alpha->r * temp.i + alpha->i *
temp.r;
c__[i__3].r = z__1.r, c__[i__3].i = z__1.i;
} else {
i__3 = i__ + j * c_dim1;
z__2.r = alpha->r * temp.r - alpha->i * temp.i,
z__2.i = alpha->r * temp.i + alpha->i *
temp.r;
i__4 = i__ + j * c_dim1;
z__3.r = beta->r * c__[i__4].r - beta->i * c__[i__4]
.i, z__3.i = beta->r * c__[i__4].i + beta->i *
c__[i__4].r;
z__1.r = z__2.r + z__3.r, z__1.i = z__2.i + z__3.i;
c__[i__3].r = z__1.r, c__[i__3].i = z__1.i;
}
// L360:
}
// L370:
}
}
}
return 0;
//
// End of ZGEMM .
//
} // zgemm_
+5 -5
View File
@@ -1,8 +1,8 @@
# Binaries branch name: ffmpeg/5.x_20260602
# Binaries were created for OpenCV: a0a660fcb1e58a295e6caa6aee64ed4d369b0181
ocv_update(FFMPEG_BINARIES_COMMIT "06dc20cad65dc7fcf784f70c95d46750520889a7")
ocv_update(FFMPEG_FILE_HASH_BIN32 "9cef7a78b6f7ec8cf1a3935c058cfac5")
ocv_update(FFMPEG_FILE_HASH_BIN64 "a821a1135251859655090c795af05789")
# Binaries branch name: ffmpeg/4.x_20260715
# Binaries were created for OpenCV: 6b640b424c516d27217700392a7ed6a362e040a5
ocv_update(FFMPEG_BINARIES_COMMIT "bd9418020a5c342be979c56d6e6434261959d3af")
ocv_update(FFMPEG_FILE_HASH_BIN32 "31968b434799d3dd56969b15aac3efa8")
ocv_update(FFMPEG_FILE_HASH_BIN64 "84757ed0f16ddedce99227529d3165e9")
ocv_update(FFMPEG_FILE_HASH_CMAKE "e09efc33312d1173be8a9446f3b088fe")
function(download_win_ffmpeg script_var)
-70
View File
@@ -1,70 +0,0 @@
# ----------------------------------------------------------------------------
# CMake file for the bundled HarfBuzz text-shaping library. See root CMakeLists.txt
#
# OpenCV vendors a minimal subset of HarfBuzz (core + HB_HAS_RASTER) as a
# normal static library: each .cc is a separate translation unit. The subset
# is produced by hb_extract.py (see that script to update HarfBuzz).
# ----------------------------------------------------------------------------
project(${HARFBUZZ_LIBRARY} CXX)
ocv_include_directories("${CMAKE_CURRENT_SOURCE_DIR}/src")
# Every .cc copied by hb_extract.py is a real translation unit (the script
# never copies non-TU .cc files), so a plain recursive glob is exactly right.
file(GLOB_RECURSE lib_srcs src/*.cc)
file(GLOB_RECURSE lib_hdrs src/*.h src/*.hh)
# HarfBuzz is built with HB_TINY for minimum footprint. Because we request the
# software rasterizer (HB_HAS_RASTER, defined below), HB_TINY does NOT disable
# the draw/paint/color APIs nor CFF (PostScript/.otf
# outlines) -- see hb-config.hh: HB_NO_DRAW/COLOR/PAINT and the TINY->HB_NO_CFF
# rule only trigger when no HB_HAS_* backend is requested. So CFF (CJK/Indic),
# COLR/CPAL color and our rasterizer all survive HB_TINY.
#
# Two features HB_TINY (via HB_LEAN) would otherwise drop are restored through
# HarfBuzz's config-override hook (hb-opencv-config.hh): thread-safety (HB_NO_MT)
# and variable fonts (HB_NO_VAR). No platform backends (CoreText/DirectWrite/...).
add_library(${HARFBUZZ_LIBRARY} STATIC ${OPENCV_3RDPARTY_EXCLUDE_FROM_ALL} ${lib_srcs} ${lib_hdrs})
target_compile_definitions(${HARFBUZZ_LIBRARY} PRIVATE
HB_TINY
HB_HAS_RASTER # request the software rasterizer: keeps draw/paint/color + CFF under HB_TINY
"HB_CONFIG_OVERRIDE_H=\"hb-opencv-config.hh\"")
set_target_properties(${HARFBUZZ_LIBRARY} PROPERTIES
CXX_STANDARD 17
CXX_STANDARD_REQUIRED ON
)
ocv_warnings_disable(CMAKE_CXX_FLAGS
-Wunused-variable -Wunused-function -Wunused-parameter
-Wunused-but-set-variable # clang15/gcc
-Wshadow
-Wmissing-declarations # gcc
-Wmissing-prototypes # clang
-Wimplicit-fallthrough
-Wextra-semi # clang
-Wdeprecated-copy # gcc/clang
-Wsuggest-override
-Wcast-function-type
-Wclass-memaccess # gcc
-Wexpansion-to-defined
)
ocv_warnings_disable(CMAKE_CXX_FLAGS /wd4244 /wd4267 /wd4127 /wd4146 /wd4456 /wd4459) # MSVC
set_target_properties(${HARFBUZZ_LIBRARY}
PROPERTIES OUTPUT_NAME ${HARFBUZZ_LIBRARY}
DEBUG_POSTFIX "${OPENCV_DEBUG_POSTFIX}"
COMPILE_PDB_NAME ${HARFBUZZ_LIBRARY}
COMPILE_PDB_NAME_DEBUG "${HARFBUZZ_LIBRARY}${OPENCV_DEBUG_POSTFIX}"
ARCHIVE_OUTPUT_DIRECTORY ${3P_LIBRARY_OUTPUT_PATH}
)
if(ENABLE_SOLUTION_FOLDERS)
set_target_properties(${HARFBUZZ_LIBRARY} PROPERTIES FOLDER "3rdparty")
endif()
if(NOT BUILD_SHARED_LIBS)
ocv_install_target(${HARFBUZZ_LIBRARY} EXPORT OpenCVModules ARCHIVE DESTINATION ${OPENCV_3P_LIB_INSTALL_PATH} COMPONENT dev OPTIONAL)
endif()
-42
View File
@@ -1,42 +0,0 @@
HarfBuzz is licensed under the so-called "Old MIT" license. Details follow.
For parts of HarfBuzz that are licensed under different licenses see individual
files names COPYING in subdirectories where applicable.
Copyright © 2010-2022 Google, Inc.
Copyright © 2015-2020 Ebrahim Byagowi
Copyright © 2019,2020 Facebook, Inc.
Copyright © 2012,2015 Mozilla Foundation
Copyright © 2011 Codethink Limited
Copyright © 2008,2010 Nokia Corporation and/or its subsidiary(-ies)
Copyright © 2009 Keith Stribley
Copyright © 2011 Martin Hosken and SIL International
Copyright © 2007 Chris Wilson
Copyright © 2005,2006,2020,2021,2022,2023 Behdad Esfahbod
Copyright © 2004,2007,2008,2009,2010,2013,2021,2022,2023 Red Hat, Inc.
Copyright © 1998-2005 David Turner and Werner Lemberg
Copyright © 2016 Igalia S.L.
Copyright © 2022 Matthias Clasen
Copyright © 2018,2021 Khaled Hosny
Copyright © 2018,2019,2020 Adobe, Inc
Copyright © 2013-2015 Alexei Podtelezhnikov
For full copyright notices consult the individual files in the package.
Permission is hereby granted, without written agreement and without
license or royalty fees, to use, copy, modify, and distribute this
software and its documentation for any purpose, provided that the
above copyright notice and the following two paragraphs appear in
all copies of this software.
IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
DAMAGE.
THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
-220
View File
@@ -1,220 +0,0 @@
# HarfBuzz
<div align="center">
<p><img src="HarfBuzz.png" alt="HarfBuzz Logo" width="256"/></p>
[![Linux CI Status](https://github.com/harfbuzz/harfbuzz/actions/workflows/linux.yml/badge.svg)](https://github.com/harfbuzz/harfbuzz/actions/workflows/linux.yml)
[![macoOS CI Status](https://github.com/harfbuzz/harfbuzz/actions/workflows/macos.yml/badge.svg)](https://github.com/harfbuzz/harfbuzz/actions/workflows/macos.yml)
[![Windows CI Status](https://github.com/harfbuzz/harfbuzz/actions/workflows/msvc.yml/badge.svg)](https://github.com/harfbuzz/harfbuzz/actions/workflows/msvc.yml)
[![OSS-Fuzz Status](https://oss-fuzz-build-logs.storage.googleapis.com/badges/harfbuzz.svg)](https://oss-fuzz-build-logs.storage.googleapis.com/index.html#harfbuzz)
[![Coverity Scan Build Status](https://scan.coverity.com/projects/15166/badge.svg)](https://scan.coverity.com/projects/harfbuzz)
[![Packaging status](https://repology.org/badge/tiny-repos/harfbuzz.svg)](https://repology.org/project/harfbuzz/versions)
[![OpenSSF Scorecard](https://api.securityscorecards.dev/projects/github.com/harfbuzz/harfbuzz/badge)](https://securityscorecards.dev/viewer/?uri=github.com/harfbuzz/harfbuzz)
</div>
HarfBuzz started as a text shaping engine but has grown into a
full font platform — the `ffmpeg` of text shaping. It primarily
supports [OpenType][1], but also [Apple Advanced Typography][2].
HarfBuzz shapes the majority of text on modern screens.
HarfBuzz is optimized for robustness, correctness, and performance
— in that order. Achieve all.
**[Try it live at harfbuzz-world.cc](https://harfbuzz-world.cc/)** — an interactive playground for shaping, subsetting, rasterization, vector output, and GPU rendering, all running in your browser.
Here is a quick map of its components:
### Core libraries
| Library | Description |
|---------|-------------|
| **libharfbuzz** | Text shaping, draw API, paint API. Highly configurable (see [CONFIG.md](CONFIG.md)). Optional integration backends compiled in: hb-ft (FreeType), hb-coretext (macOS), hb-uniscribe (Windows), hb-directwrite (Windows), hb-gdi (Windows), hb-glib, hb-graphite2. |
| **libharfbuzz-subset** | Font subsetting and variable-font instancing. |
### Auxiliary libraries
| Library | Description |
|---------|-------------|
| **libharfbuzz-icu** | ICU Unicode integration. |
| **libharfbuzz-cairo** | Cairo rendering integration. |
| **libharfbuzz-gobject** | GObject/GI bindings. |
### Experimental libraries
| Library | Description |
|---------|-------------|
| **libharfbuzz-raster** | Glyph rasterization to bitmaps, including color fonts. Uses hb-draw and hb-paint. |
| **libharfbuzz-vector** | Glyph output to vector formats (currently SVG), including color fonts. Uses hb-draw and hb-paint. |
| **libharfbuzz-gpu** | Encodes glyph outlines for GPU rasterization (Slug algorithm). Provides shader sources in GLSL, WGSL, MSL, and HLSL. [Live demo.](https://harfbuzz.github.io/hb-gpu-demo/) |
Notable missing feature: font hinting (including autohinting)
is not implemented. For hinted rasterization, use FreeType or
Skrifa.
For simplified builds, amalgamated sources are available:
`harfbuzz.cc` (just libharfbuzz), `harfbuzz-subset.cc` (just
libharfbuzz-subset), or `harfbuzz-world.cc` (everything, driven
by a custom `hb-features.h`). For a live in-browser playground
plus a worked example of the world.cc single-file build, see
[harfbuzz-world.cc][26].
### Command-line tools
| Tool | Description |
|------|-------------|
| **hb-shape** | Shape text and display glyph output. |
| **hb-view** | Render shaped text to an image. |
| **hb-subset** | Subset and optimize fonts. |
| **hb-info** | Display font metadata. |
| **hb-raster** | Render glyphs to bitmap images. |
| **hb-vector** | Render glyphs to vector formats (SVG). |
| **hb-gpu** | Interactive GPU text rendering. |
The canonical source tree and bug trackers are available on [github][4].
Both development and user support discussion around HarfBuzz happen on
[github][4] as well.
For license information, see [COPYING](COPYING).
## API stability
The API that comes with `hb.h` will not change incompatibly. Other, peripheral,
headers are more likely to go through minor modifications, but again, we do our
best to never change API in an incompatible way. We will never break the ABI.
The API and ABI are stable even across major version number jumps. In fact,
current HarfBuzz is API/ABI compatible all the way back to the 0.9.x series.
If one day we need to break the API/ABI, that would be called a new library.
As such, we bump the major version number only when we add major new features,
the minor version when there is new API, and the micro version when there
are bug fixes.
## Documentation
For user manual as well as API documentation, check: https://harfbuzz.github.io
## Download
Tarball releases and Win32/Win64 binary bundles are available on the
[github releases][3] page.
## Development
For build information, see [BUILD.md](BUILD.md).
For custom configurations, see [CONFIG.md](CONFIG.md).
For testing and profiling, see [TESTING.md](TESTING.md).
For using with Python, see [README.python.md](README.python.md). There is also [uharfbuzz](https://github.com/harfbuzz/uharfbuzz).
For cross-compiling to Windows from Linux or macOS, see [README.mingw.md](README.mingw.md).
To report bugs or submit patches please use [github][4] issues and pull-requests.
### Developer documents
To get a better idea of where HarfBuzz stands in the text rendering stack you
may want to read [State of Text Rendering 2024][6].
Here are a few presentation slides about HarfBuzz over the years:
- 2026 [HarfBuzz at 20!][25]
- 2016 [Ten Years of HarfBuzz][20]
- 2014 [Unicode, OpenType, and HarfBuzz: Closing the Circle][7]
- 2012 [HarfBuzz, The Free and Open Text Shaping Engine][8]
- 2009 [HarfBuzz: the Free and Open Shaping Engine][9]
More presentations and papers are available on [behdad][11]'s website.
In particular, the following _studies_ are relevant to HarfBuzz development:
- 2025 [AAT layout caches][24]
- 2025 [OpenType Layout lookup caches][23]
- 2025 [Introducing HarfRust][22]
- 2025 [Subsetting][21]
- 2025 [Caching][12]
- 2025 [`hb-decycler`][13]
- 2022 [`hb-iter`][14]
- 2022 [A C library written in C++][15]
- 2022 [The case of the slow `hb-ft` `>h_advance` function][18]
- 2022 [PackTab: A static integer table packer][16]
- 2020 [HarfBuzz OT+AAT "Unishaper"][19]
- 2014 [Building the Indic Shaper][17]
- 2012 [Memory Consumption][10]
## Name
HarfBuzz /hærfˈbɒːz/
From Persian حرف (*Harf*: letter) and باز (*Buzz*: open).
Transliteration of the Persian calque for *OpenType*.
As a noun: *The* Open Source *text shaping* engine.
As an adjective: Insincerely talkative; glib. A nod to the
GNOME project where HarfBuzz originates from.
The logo shows حرف‌باز in the IranNastaliq font, on a Damascus
steel background.
> Background: Originally there was this font format called TrueType. People and
> companies started calling their type engines all things ending in Type:
> FreeType, CoolType, ClearType, etc. And then came OpenType, which is the
> successor of TrueType. So, for my OpenType implementation, I decided to stick
> with the concept but use the Persian translation. Which is fitting given that
> Persian is written in the Arabic script, and OpenType is an extension of
> TrueType that adds support for complex script rendering, and HarfBuzz is an
> implementation of OpenType text shaping.
## Users
HarfBuzz is used in Android, Chrome, ChromeOS, Firefox, Flutter, GNOME, GTK+, KDE,
Qt, LibreOffice, OpenJDK, XeTeX, Adobe Photoshop, Illustrator, InDesign,
Microsoft Edge, Amazon Kindle, PlayStation, Godot Engine, Unreal Engine,
Figma, Canva, QuarkXPress, Scribus, smart TVs,
car displays, and many other places.
<p align="center">
<a href="https://xkcd.com/2347/" rel="nofollow">
<img src="xkcd.png" width="256" alt="xkcd-derived image">
</a>
</p>
## Distribution
<details>
<summary>Packaging status of HarfBuzz</summary>
[![Packaging status](https://repology.org/badge/vertical-allrepos/harfbuzz.svg?header=harfbuzz)](https://repology.org/project/harfbuzz/versions)
</details>
[1]: https://docs.microsoft.com/en-us/typography/opentype/spec/
[2]: https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6AATIntro.html
[3]: https://github.com/harfbuzz/harfbuzz/releases
[4]: https://github.com/harfbuzz/harfbuzz
[6]: https://behdad.org/text2024
[7]: https://docs.google.com/presentation/d/1x97pfbB1gbD53Yhz6-_yBUozQMVJ_5yMqqR_D-R7b7I/preview
[8]: https://docs.google.com/presentation/d/1ySTZaXP5XKFg0OpmHZM00v5b17GSr3ojnzJekl4U8qI/preview
[9]: https://behdad.org/doc/harfbuzz2009-slides.pdf
[10]: https://docs.google.com/document/d/12jfNpQJzeVIAxoUSpk7KziyINAa1msbGliyXqguS86M/preview
[11]: https://behdad.org/
[12]: https://docs.google.com/document/d/1_VgObf6Je0J8byMLsi7HCQHnKo2emGnx_ib_sHo-bt4/preview
[13]: https://docs.google.com/document/d/1Y-u08l9YhObRVObETZt1k8f_5lQdOix9TRH3zEXaoAw/preview
[14]: https://docs.google.com/document/d/1o-xvxCbgMe9JYFHLVnPjk01ZY_8Cj0vB9-KTI1d0nyk/preview
[15]: https://docs.google.com/document/d/18hI56KJpvXtwWbc9QSaz9zzhJwIMnrJ-zkAaKS-W-8k/preview
[16]: https://docs.google.com/document/d/1Xq3owVt61HVkJqbLFHl73il6pcTy6PdPJJ7bSouQiQw/preview
[17]: https://docs.google.com/document/d/1wMPwVNBvsIriamcyBO5aNs7Cdr8lmbwLJ8GmZBAswF4/preview
[18]: https://docs.google.com/document/d/1wskYbA-czBt57oH9gEuGf3sWbTx7bfOiEIcDs36-heo/preview
[19]: https://prezi.com/view/THNPJGFVDUCWoM20syev/
[20]: https://behdad.org/doc/harfbuzz10years-slides.pdf
[21]: https://docs.google.com/document/d/1_vZrt97OorJ0jA1YzJ29LRcGr3YGrNJANdOABjVZGEs/preview
[22]: https://docs.google.com/document/d/1aH_waagdEM5UhslQxCeFEb82ECBhPlZjy5_MwLNLBYo/preview
[23]: https://docs.google.com/document/d/1hRd5oYQJLrt0JuwWhEJWi7wh_9rbaIJkX6IR9DW7rZQ/preview
[24]: https://docs.google.com/document/d/1a3K6fHjsiWW36vSzwJwCwEBOgznunKs80PSpBbpfHiA/preview
[25]: https://docs.google.com/presentation/d/1o9Exz1c-Lr-dJjA8dcBn_Vl_Y37cupmFzmclMjBE_Bc/view
[26]: https://harfbuzz-world.cc/
-386
View File
@@ -1,386 +0,0 @@
#!/usr/bin/env python3
"""
hb_extract.py -- copy the minimal subset of HarfBuzz sources used by OpenCV.
OpenCV vendors HarfBuzz as a normal static library: every .cc is a separate
translation unit (no unity / amalgamation build). This script reproduces the
exact src/ tree that OpenCV ships, from a pristine HarfBuzz checkout.
It does three things beyond a naive copy, all to keep the vendored tree small
and buildable with OpenCV's configuration (HB_TINY + HB_HAS_RASTER, see
CMakeLists.txt):
1. Copies the translation units that upstream `harfbuzz-world.cc` would
compile for the requested HB_HAS_* sections -- EXCEPT the ones that build
to an empty object under our config (see EMPTY_TUS). It never emits
`harfbuzz-world.cc` itself (we compile each .cc directly), and never
copies non-TU .cc files found under src/OT|graph (only real TUs from the
parsed world.cc lists are copied), so CMake's file(GLOB_RECURSE src/*.cc)
picks up exactly the right set.
2. Prunes headers down to the set actually reachable by #include from the
copied TUs (+ the public headers OpenCV includes). This automatically
drops the large amount of HarfBuzz src/ that we never compile -- GPU and
WASM backends, the subset/repacker graph, platform backends (CoreText/
DirectWrite/GDI/Uniscribe/Graphite2/GObject/Cairo), the vector-paint
backend, and headers used only by the skipped empty TUs -- without a
hardcoded blocklist.
3. Generates `hb-opencv-config.hh`, the config-override header CMakeLists.txt
points HB_CONFIG_OVERRIDE_H at. It re-enables thread-safety (HB_NO_MT) and
variable fonts (HB_NO_VAR) that HB_TINY would otherwise switch off. (It
does NOT generate upstream's hb-features.h: nothing we compile includes
it.)
Usage (to refresh this very directory, just point it at a harfbuzz checkout):
python hb_extract.py path/to/harfbuzz
The defaults already produce the OpenCV subset: output '.', features
HB_HAS_RASTER, and it also copies README.md + COPYING. Override only if needed:
-o DIR output root (default '.')
-f FLAGS comma-separated HB_HAS_* (default HB_HAS_RASTER)
-a FILES extra files (default README.md,COPYING)
--list-flags print available HB_HAS_* flags and exit
-a paths are relative to the harfbuzz repo root (parent of src/):
-a README.md -> copied to <out>/README.md
-a COPYING -> copied to <out>/COPYING
-a src/hb-raster.h -> copied to <out>/src/hb-raster.h
"""
import argparse
import posixpath
import re
import shutil
import sys
from pathlib import Path
# Translation units that compile to an EMPTY object (libtool: "has no symbols")
# under OpenCV's HarfBuzz configuration (HB_TINY + HB_HAS_RASTER => AAT, legacy
# fallback shaping, math, meta, name, buffer (de)serialize/verify and the style
# API are all compiled out). They contribute nothing, so we do not vendor them.
# Re-derive this list (from the libtool warnings of a clean libharfbuzz build)
# if the HarfBuzz configuration in CMakeLists.txt changes.
EMPTY_TUS = {
"hb-aat-layout.cc",
"hb-aat-map.cc",
"hb-buffer-serialize.cc",
"hb-buffer-verify.cc",
"hb-fallback-shape.cc",
"hb-ot-math.cc",
"hb-ot-meta.cc",
"hb-ot-name.cc",
"hb-raster.cc",
"hb-style.cc",
}
# Public headers OpenCV's text engine (drawing_text.cpp) includes directly.
# Used together with the copied TUs as the roots of the header-reachability
# prune.
SEED_PUBLIC_HEADERS = ("hb.h", "hb-ot.h", "hb-raster.h")
# Generated config-override header. CMakeLists.txt builds HarfBuzz with HB_TINY
# and HB_CONFIG_OVERRIDE_H="hb-opencv-config.hh"; this file is included by
# hb-config.hh after HB_TINY/HB_LEAN expand (so the #undef takes effect) but
# before the option-closure derives dependent macros.
OPENCV_CONFIG_HH = '''\
/*
* OpenCV-specific HarfBuzz configuration override. GENERATED by hb_extract.py.
*
* HarfBuzz is built with HB_TINY (see 3rdparty/harfbuzz/CMakeLists.txt) for the
* smallest possible footprint. HB_TINY pulls in HB_LEAN + HB_MINI, which would
* disable two features that OpenCV's text engine relies on. We restore them
* here.
*
* This file is included by hb-config.hh via HB_CONFIG_OVERRIDE_H, i.e. AFTER
* HB_TINY/HB_LEAN/HB_MINI have expanded their macros (so the #undef below takes
* effect) but BEFORE the "closure of options" derives dependent macros (e.g.
* HB_NO_VAR_COMPOSITES from HB_NO_VAR). That ordering is what makes a plain
* #undef sufficient.
*
* HB_NO_MT - keep thread-safety. hb_font_t instances live inside cv::FontFace
* objects, NOT in the thread_local FontRenderEngine, so a single
* FontFace (and its hb_font_t) can be shared across threads; its
* reference counting must stay atomic.
*
* HB_NO_VAR - keep variable-font support: the 'wght' axis used for synthetic
* weights and named-instance/axis queries (hb_font_set_variations,
* hb_ot_var_*). Variable fonts are a primary reason OpenCV adopted
* HarfBuzz, so this must remain enabled.
*/
#undef HB_NO_MT
#undef HB_NO_VAR
'''
OPENCV_CONFIG_NAME = "hb-opencv-config.hh"
def parse_world_cc(path):
"""Return (core_files, sections) parsed from harfbuzz-world.cc.
core_files -- list of .cc paths (relative to src/) in the unconditional
"Core library" section.
sections -- dict HB_HAS_XXX -> list of .cc paths from that #ifdef block.
"""
lines = path.read_text(encoding="utf-8").splitlines()
include_re = re.compile(r'^\s*#include\s+"([^"]+\.cc)"')
core_files = []
sections = {}
STATE_PREAMBLE, STATE_CORE, STATE_IFDEF = "preamble", "core", "ifdef"
state = STATE_PREAMBLE
current_flag = None
ifdef_depth = 0
for line in lines:
if state == STATE_PREAMBLE:
if "/* Core library." in line:
state = STATE_CORE
elif state == STATE_CORE:
m = include_re.match(line)
if m:
core_files.append(m.group(1))
else:
m = re.match(r"^\s*#ifdef\s+(HB_HAS_\w+)", line)
if m:
current_flag = m.group(1)
sections[current_flag] = []
state = STATE_IFDEF
ifdef_depth = 1
elif state == STATE_IFDEF:
if re.match(r"^\s*#if", line):
ifdef_depth += 1
elif re.match(r"^\s*#endif", line):
ifdef_depth -= 1
if ifdef_depth == 0:
state = STATE_CORE
current_flag = None
else:
m = include_re.match(line)
if m and current_flag is not None:
sections[current_flag].append(m.group(1))
return core_files, sections
def copy_files(file_list, src_dir, out_src_dir, label=""):
skipped = []
for rel in file_list:
if Path(rel).name in EMPTY_TUS:
skipped.append(rel)
continue
src = src_dir / rel
dst = out_src_dir / rel
dst.parent.mkdir(parents=True, exist_ok=True)
if not src.exists():
print(f" [WARN] not found: {src}", file=sys.stderr)
continue
shutil.copy2(src, dst)
print(f" [{label or 'core'}] src/{rel}")
for rel in skipped:
print(f" [skip-empty] src/{rel}")
def copy_headers(src_dir, out_src_dir, public_h=True):
"""Copy all candidate headers: src/*.hh, src/*.h (public API), and *.h/*.hh
from src/OT/** and src/graph/**. The unreferenced ones are removed later by
prune_unreferenced_headers().
NOTE: .cc files under OT/ and graph/ are intentionally NOT copied here --
the only ones we need are real translation units, copied via the parsed
world.cc lists. Copying e.g. src/graph/test-classdef-graph.cc would make
CMake's file(GLOB_RECURSE src/*.cc) try to compile a non-TU file.
"""
count = 0
def _copy(p):
nonlocal count
dst = out_src_dir / p.relative_to(src_dir)
dst.parent.mkdir(parents=True, exist_ok=True)
shutil.copy2(p, dst)
count += 1
for hdr in src_dir.glob("*.hh"):
_copy(hdr)
for subdir in ("OT", "graph"):
d = src_dir / subdir
if not d.exists():
continue
for hdr in d.rglob("*"):
if hdr.is_file() and hdr.suffix in (".h", ".hh"):
_copy(hdr)
if public_h:
for hdr in src_dir.glob("*.h"):
_copy(hdr)
return count
def prune_unreferenced_headers(out_src_dir):
"""Delete every header not reachable by #include from the copied TUs and the
public seed headers, then drop any directory left empty.
HarfBuzz's amalgamated src/ ships many headers we never compile (GPU/WASM
backends, subset/repacker graph, platform backends, vector-paint, and the
headers used only by the empty TUs we skip). Reachability removes them all
without a hardcoded list. The edge regex also treats any quoted "...h"/
"...hh" literal as an include, which covers the `#include HB_STRING_ARRAY_LIST`
macro-indirection trick (e.g. hb-ot-cff1-std-str.hh, hb-ot-post-macroman.hh).
"""
files = {p.relative_to(out_src_dir).as_posix()
for p in out_src_dir.rglob("*") if p.is_file()}
headers = {f for f in files if f.endswith((".h", ".hh"))}
edge_re = re.compile(r'#\s*include\s+["<]([^">]+)[">]|"([^"]+\.hh?)"')
def resolve(inc, including):
for cand in (posixpath.normpath(posixpath.join(posixpath.dirname(including), inc)),
posixpath.normpath(inc)):
if cand in files:
return cand
return None
seeds = {f for f in files if f.endswith(".cc")}
seeds |= {h for h in SEED_PUBLIC_HEADERS if h in files}
seen, stack = set(), list(seeds)
while stack:
cur = stack.pop()
if cur in seen:
continue
seen.add(cur)
try:
txt = (out_src_dir / cur).read_text(encoding="utf-8", errors="ignore")
except OSError:
continue
for m in edge_re.finditer(txt):
inc = m.group(1) or m.group(2)
r = resolve(inc, cur)
if r and r not in seen:
stack.append(r)
needed = {f for f in seen if f.endswith((".h", ".hh"))}
for h in sorted(headers - needed):
(out_src_dir / h).unlink()
print(f" [prune] src/{h}")
# remove directories left empty by pruning (deepest first)
for d in sorted((p for p in out_src_dir.rglob("*") if p.is_dir()),
key=lambda p: len(p.parts), reverse=True):
try:
d.rmdir()
except OSError:
pass
return len(headers - needed), len(needed)
def write_opencv_config_h(out_src_dir):
"""Generate hb-opencv-config.hh (the HB_CONFIG_OVERRIDE_H header)."""
(out_src_dir / OPENCV_CONFIG_NAME).write_text(OPENCV_CONFIG_HH, encoding="utf-8")
print(f" [gen] src/{OPENCV_CONFIG_NAME}")
def main():
parser = argparse.ArgumentParser(
description="Copy the OpenCV subset of HarfBuzz sources.",
formatter_class=argparse.RawDescriptionHelpFormatter,
epilog="""
examples:
python hb_extract.py ~/work/harfbuzz -o . -f HB_HAS_RASTER -a README.md,COPYING
python hb_extract.py ~/work/harfbuzz --list-flags
""",
)
parser.add_argument("harfbuzz_dir", metavar="harfbuzz-dir",
help="path to the harfbuzz repo root (contains src/)")
parser.add_argument("-o", "--output", metavar="DIR", default=".",
help="output root directory; sources go into DIR/src/ "
"(default: current directory)")
parser.add_argument("-f", "--features", metavar="FLAGS", default="HB_HAS_RASTER",
help="comma-separated HB_HAS_* flags (default: HB_HAS_RASTER)")
parser.add_argument("-a", "--add", metavar="FILES", default="README.md,COPYING",
help="comma-separated files relative to the harfbuzz repo "
"root (default: README.md,COPYING)")
parser.add_argument("--no-public-headers", action="store_true",
help="skip copying public src/*.h API headers "
"(OT/, graph/, *.hh are always copied)")
parser.add_argument("--list-flags", action="store_true",
help="list available HB_HAS_* flags and exit")
args = parser.parse_args()
repo_dir = Path(args.harfbuzz_dir).resolve()
world_cc_path = repo_dir / "src" / "harfbuzz-world.cc"
if not world_cc_path.exists():
print(f"error: {world_cc_path} not found", file=sys.stderr)
sys.exit(1)
src_dir = world_cc_path.parent
core_files, sections = parse_world_cc(world_cc_path)
if args.list_flags:
for flag, files in sorted(sections.items()):
print(f" {flag:<22} ({len(files)} .cc files)")
return
enabled_flags = [f.strip() for f in args.features.split(",") if f.strip()]
extra_files = [f.strip() for f in args.add.split(",") if f.strip()]
out_dir = Path(args.output).resolve()
out_src_dir = out_dir / "src"
# Start from a clean src/ so removed-upstream files do not linger.
if out_src_dir.exists():
shutil.rmtree(out_src_dir)
out_src_dir.mkdir(parents=True, exist_ok=True)
print(f"\nsource : {repo_dir}")
print(f"output : {out_dir}")
print(f"flags : {enabled_flags}\n")
print("core:")
copy_files(core_files, src_dir, out_src_dir)
for flag in enabled_flags:
flag_files = sections.get(flag)
if not flag_files:
print(f"\n[{flag}]: no .cc files (flag not found in harfbuzz-world.cc)")
continue
print(f"\n[{flag}]:")
copy_files(flag_files, src_dir, out_src_dir, label=flag)
if extra_files:
print("\nextra:")
for rel in extra_files:
src = repo_dir / rel
dst = out_dir / rel
dst.parent.mkdir(parents=True, exist_ok=True)
if not src.exists():
print(f" [WARN] not found: {src}", file=sys.stderr)
continue
shutil.copy2(src, dst)
print(f" [add] {rel}")
n = copy_headers(src_dir, out_src_dir,
public_h=not args.no_public_headers)
print(f"\nheaders: {n} files copied")
print("\nprune (unreferenced headers):")
removed, kept = prune_unreferenced_headers(out_src_dir)
print(f" removed {removed}, kept {kept}")
print("\ngenerated:")
write_opencv_config_h(out_src_dir)
all_cc = sorted(out_src_dir.rglob("*.cc"))
all_h = list(out_src_dir.rglob("*.h")) + list(out_src_dir.rglob("*.hh"))
print(f"\ndone: {len(all_cc)} .cc | {len(all_h)} headers -> {out_dir}")
if __name__ == "__main__":
main()
File diff suppressed because it is too large Load Diff
File diff suppressed because it is too large Load Diff
-367
View File
@@ -1,367 +0,0 @@
/*
* Copyright © 2016 Google, Inc.
* Copyright © 2018 Ebrahim Byagowi
*
* This is part of HarfBuzz, a text shaping library.
*
* Permission is hereby granted, without written agreement and without
* license or royalty fees, to use, copy, modify, and distribute this
* software and its documentation for any purpose, provided that the
* above copyright notice and the following two paragraphs appear in
* all copies of this software.
*
* IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
* DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
* ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
* IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
* DAMAGE.
*
* THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
* BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
* FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
* ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
* PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
*
* Google Author(s): Sascha Brawer
*/
#ifndef OT_COLOR_CPAL_CPAL_HH
#define OT_COLOR_CPAL_CPAL_HH
#include "../../../hb-open-type.hh"
#include "../../../hb-ot-color.h"
#include "../../../hb-ot-name.h"
/*
* CPAL -- Color Palette
* https://docs.microsoft.com/en-us/typography/opentype/spec/cpal
*/
#define HB_OT_TAG_CPAL HB_TAG('C','P','A','L')
namespace OT {
struct CPALV1Tail
{
friend struct CPAL;
private:
hb_ot_color_palette_flags_t get_palette_flags (const void *base,
unsigned int palette_index,
unsigned int palette_count) const
{
if (!paletteFlagsZ) return HB_OT_COLOR_PALETTE_FLAG_DEFAULT;
return (hb_ot_color_palette_flags_t) (uint32_t)
(base+paletteFlagsZ).as_array (palette_count)[palette_index];
}
hb_ot_name_id_t get_palette_name_id (const void *base,
unsigned int palette_index,
unsigned int palette_count) const
{
if (!paletteLabelsZ) return HB_OT_NAME_ID_INVALID;
return (base+paletteLabelsZ).as_array (palette_count)[palette_index];
}
hb_ot_name_id_t get_color_name_id (const void *base,
unsigned int color_index,
unsigned int color_count) const
{
if (!colorLabelsZ) return HB_OT_NAME_ID_INVALID;
return (base+colorLabelsZ).as_array (color_count)[color_index];
}
public:
void collect_name_ids (const void *base,
unsigned palette_count,
unsigned color_count,
const hb_map_t *color_index_map,
hb_set_t *nameids_to_retain /* OUT */) const
{
if (paletteLabelsZ)
{
+ (base+paletteLabelsZ).as_array (palette_count)
| hb_sink (nameids_to_retain)
;
}
if (colorLabelsZ)
{
const hb_array_t<const NameID> colorLabels = (base+colorLabelsZ).as_array (color_count);
for (unsigned i = 0; i < color_count; i++)
{
if (!color_index_map->has (i)) continue;
nameids_to_retain->add (colorLabels[i]);
}
}
}
bool serialize (hb_serialize_context_t *c,
unsigned palette_count,
unsigned color_count,
const void *base,
const hb_map_t *color_index_map) const
{
TRACE_SERIALIZE (this);
auto *out = c->allocate_size<CPALV1Tail> (static_size);
if (unlikely (!out)) return_trace (false);
out->paletteFlagsZ = 0;
if (paletteFlagsZ)
out->paletteFlagsZ.serialize_copy (c, paletteFlagsZ, base, 0, hb_serialize_context_t::Head, palette_count);
out->paletteLabelsZ = 0;
if (paletteLabelsZ)
out->paletteLabelsZ.serialize_copy (c, paletteLabelsZ, base, 0, hb_serialize_context_t::Head, palette_count);
const hb_array_t<const NameID> colorLabels = (base+colorLabelsZ).as_array (color_count);
if (colorLabelsZ)
{
c->push ();
for (unsigned i = 0; i < color_count; i++)
{
if (!color_index_map->has (i)) continue;
if (!c->copy<NameID> (colorLabels[i]))
{
c->pop_discard ();
return_trace (false);
}
}
c->add_link (out->colorLabelsZ, c->pop_pack ());
}
return_trace (true);
}
bool sanitize (hb_sanitize_context_t *c,
const void *base,
unsigned int palette_count,
unsigned int color_count) const
{
TRACE_SANITIZE (this);
return_trace (c->check_struct (this) &&
(!paletteFlagsZ || (base+paletteFlagsZ).sanitize (c, palette_count)) &&
(!paletteLabelsZ || (base+paletteLabelsZ).sanitize (c, palette_count)) &&
(!colorLabelsZ || (base+colorLabelsZ).sanitize (c, color_count)));
}
protected:
// TODO(garretrieger): these offsets can hold nulls so we should not be using non-null offsets
// here. Currently they are needed since UnsizedArrayOf doesn't define null_size
NNOffset32To<UnsizedArrayOf<HBUINT32>>
paletteFlagsZ; /* Offset from the beginning of CPAL table to
* the Palette Type Array. Set to 0 if no array
* is provided. */
NNOffset32To<UnsizedArrayOf<NameID>>
paletteLabelsZ; /* Offset from the beginning of CPAL table to
* the palette labels array. Set to 0 if no
* array is provided. */
NNOffset32To<UnsizedArrayOf<NameID>>
colorLabelsZ; /* Offset from the beginning of CPAL table to
* the color labels array. Set to 0
* if no array is provided. */
public:
DEFINE_SIZE_STATIC (12);
};
typedef HBUINT32 BGRAColor;
struct CPAL
{
static constexpr hb_tag_t tableTag = HB_OT_TAG_CPAL;
bool has_data () const { return numPalettes; }
size_t get_size () const
{ return min_size + numPalettes * sizeof (colorRecordIndicesZ[0]); }
unsigned int get_palette_count () const { return numPalettes; }
unsigned int get_color_count () const { return numColors; }
hb_ot_color_palette_flags_t get_palette_flags (unsigned int palette_index) const
{ return v1 ().get_palette_flags (this, palette_index, numPalettes); }
hb_ot_name_id_t get_palette_name_id (unsigned int palette_index) const
{ return v1 ().get_palette_name_id (this, palette_index, numPalettes); }
hb_ot_name_id_t get_color_name_id (unsigned int color_index) const
{ return v1 ().get_color_name_id (this, color_index, numColors); }
hb_array_t<const BGRAColor> get_palette_colors (unsigned int palette_index) const
{
if (unlikely (palette_index >= numPalettes))
return hb_array_t<const BGRAColor> ();
unsigned int start_index = colorRecordIndicesZ[palette_index];
hb_array_t<const BGRAColor> all_colors ((this+colorRecordsZ).arrayZ, numColorRecords);
return all_colors.sub_array (start_index, numColors);
}
unsigned int get_palette_colors (unsigned int palette_index,
unsigned int start_offset,
unsigned int *color_count, /* IN/OUT. May be NULL. */
hb_color_t *colors /* OUT. May be NULL. */) const
{
if (unlikely (palette_index >= numPalettes))
{
if (color_count) *color_count = 0;
return 0;
}
unsigned int start_index = colorRecordIndicesZ[palette_index];
hb_array_t<const BGRAColor> all_colors ((this+colorRecordsZ).arrayZ, numColorRecords);
hb_array_t<const BGRAColor> palette_colors = all_colors.sub_array (start_index,
numColors);
if (color_count)
{
+ palette_colors.sub_array (start_offset, color_count)
| hb_sink (hb_array (colors, *color_count))
;
}
return numColors;
}
void collect_name_ids (const hb_map_t *color_index_map,
hb_set_t *nameids_to_retain /* OUT */) const
{
if (version == 1)
{
hb_barrier ();
v1 ().collect_name_ids (this, numPalettes, numColors, color_index_map, nameids_to_retain);
}
}
private:
const CPALV1Tail& v1 () const
{
if (version == 0) return Null (CPALV1Tail);
hb_barrier ();
return StructAfter<CPALV1Tail> (*this);
}
public:
bool serialize (hb_serialize_context_t *c,
const hb_array_t<const HBUINT16> &color_record_indices,
const hb_array_t<const BGRAColor> &color_records,
const hb_vector_t<unsigned>& first_color_index_for_layer,
const hb_map_t& first_color_to_layer_index,
const hb_set_t &retained_color_indices) const
{
TRACE_SERIALIZE (this);
// TODO(grieger): limit total final size.
for (const auto idx : color_record_indices)
{
hb_codepoint_t layer_index = first_color_to_layer_index[idx];
HBUINT16 new_idx;
new_idx = layer_index * retained_color_indices.get_population ();
if (!c->copy<HBUINT16> (new_idx)) return_trace (false);
}
c->push ();
for (unsigned first_color_index : first_color_index_for_layer)
{
for (hb_codepoint_t color_index : retained_color_indices)
{
if (!c->copy<BGRAColor> (color_records[first_color_index + color_index]))
{
c->pop_discard ();
return_trace (false);
}
}
}
c->add_link (colorRecordsZ, c->pop_pack ());
return_trace (true);
}
bool subset (hb_subset_context_t *c) const
{
TRACE_SUBSET (this);
if (!numPalettes) return_trace (false);
const hb_map_t *color_index_map = &c->plan->colr_palettes;
if (color_index_map->is_empty ()) return_trace (false);
hb_set_t retained_color_indices;
for (const auto _ : color_index_map->keys ())
{
if (_ == 0xFFFF) continue;
retained_color_indices.add (_);
}
if (retained_color_indices.is_empty ()) return_trace (false);
auto *out = c->serializer->start_embed (*this);
if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
out->version = version;
out->numColors = retained_color_indices.get_population ();
out->numPalettes = numPalettes;
hb_vector_t<unsigned> first_color_index_for_layer;
hb_map_t first_color_to_layer_index;
const hb_array_t<const HBUINT16> colorRecordIndices = colorRecordIndicesZ.as_array (numPalettes);
for (const auto first_color_record_idx : colorRecordIndices)
{
if (first_color_to_layer_index.has (first_color_record_idx)) continue;
first_color_index_for_layer.push (first_color_record_idx);
if (unlikely (!c->serializer->propagate_error (first_color_index_for_layer))) return_trace (false);
first_color_to_layer_index.set (first_color_record_idx,
first_color_index_for_layer.length - 1);
}
out->numColorRecords = first_color_index_for_layer.length
* retained_color_indices.get_population ();
const hb_array_t<const BGRAColor> color_records = (this+colorRecordsZ).as_array (numColorRecords);
if (!out->serialize (c->serializer,
colorRecordIndices,
color_records,
first_color_index_for_layer,
first_color_to_layer_index,
retained_color_indices))
return_trace (false);
if (version == 1)
{
hb_barrier ();
return_trace (v1 ().serialize (c->serializer, numPalettes, numColors, this, color_index_map));
}
return_trace (true);
}
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (c->check_struct (this) &&
hb_barrier () &&
(this+colorRecordsZ).sanitize (c, numColorRecords) &&
colorRecordIndicesZ.sanitize (c, numPalettes) &&
(version == 0 || v1 ().sanitize (c, this, numPalettes, numColors)));
}
protected:
HBUINT16 version; /* Table version number */
/* Version 0 */
HBUINT16 numColors; /* Number of colors in each palette. */
HBUINT16 numPalettes; /* Number of palettes in the table. */
HBUINT16 numColorRecords; /* Total number of color records, combined for
* all palettes. */
NNOffset32To<UnsizedArrayOf<BGRAColor>>
colorRecordsZ; /* Offset from the beginning of CPAL table to
* the first ColorRecord. */
UnsizedArrayOf<HBUINT16>
colorRecordIndicesZ; /* Index of each palettes first color record in
* the combined color record array. */
/*CPALV1Tail v1;*/
public:
DEFINE_SIZE_ARRAY (12, colorRecordIndicesZ);
};
} /* namespace OT */
#endif /* OT_COLOR_CPAL_CPAL_HH */
-449
View File
@@ -1,449 +0,0 @@
/*
* Copyright © 2018 Ebrahim Byagowi
* Copyright © 2020 Google, Inc.
*
* This is part of HarfBuzz, a text shaping library.
*
* Permission is hereby granted, without written agreement and without
* license or royalty fees, to use, copy, modify, and distribute this
* software and its documentation for any purpose, provided that the
* above copyright notice and the following two paragraphs appear in
* all copies of this software.
*
* IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
* DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
* ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
* IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
* DAMAGE.
*
* THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
* BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
* FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
* ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
* PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
*
* Google Author(s): Calder Kitagawa
*/
#ifndef OT_COLOR_SBIX_SBIX_HH
#define OT_COLOR_SBIX_SBIX_HH
#include "../../../hb-open-type.hh"
#include "../../../hb-paint.hh"
/*
* sbix -- Standard Bitmap Graphics
* https://docs.microsoft.com/en-us/typography/opentype/spec/sbix
* https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6sbix.html
*/
#define HB_OT_TAG_sbix HB_TAG('s','b','i','x')
namespace OT {
struct SBIXGlyph
{
SBIXGlyph* copy (hb_serialize_context_t *c, unsigned int data_length) const
{
TRACE_SERIALIZE (this);
SBIXGlyph* new_glyph = c->start_embed<SBIXGlyph> ();
if (unlikely (!c->extend_min (new_glyph))) return_trace (nullptr);
new_glyph->xOffset = xOffset;
new_glyph->yOffset = yOffset;
new_glyph->graphicType = graphicType;
data.copy (c, data_length);
return_trace (new_glyph);
}
HBINT16 xOffset; /* The horizontal (x-axis) offset from the left
* edge of the graphic to the glyphs origin.
* That is, the x-coordinate of the point on the
* baseline at the left edge of the glyph. */
HBINT16 yOffset; /* The vertical (y-axis) offset from the bottom
* edge of the graphic to the glyphs origin.
* That is, the y-coordinate of the point on the
* baseline at the left edge of the glyph. */
Tag graphicType; /* Indicates the format of the embedded graphic
* data: one of 'jpg ', 'png ' or 'tiff', or the
* special format 'dupe'. */
UnsizedArrayOf<HBUINT8>
data; /* The actual embedded graphic data. The total
* length is inferred from sequential entries in
* the glyphDataOffsets array and the fixed size
* (8 bytes) of the preceding fields. */
public:
DEFINE_SIZE_ARRAY (8, data);
};
struct SBIXStrike
{
static size_t get_size (unsigned num_glyphs)
{ return min_size + num_glyphs * HBUINT32::static_size; }
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (c->check_struct (this) &&
imageOffsetsZ.sanitize_shallow (c, c->get_num_glyphs () + 1));
}
hb_blob_t *get_glyph_blob (unsigned int glyph_id,
hb_blob_t *sbix_blob,
hb_tag_t file_type,
int *x_offset,
int *y_offset,
unsigned int num_glyphs,
unsigned int *strike_ppem) const
{
if (unlikely (!ppem)) return hb_blob_get_empty (); /* To get Null() object out of the way. */
unsigned int retry_count = 8;
unsigned int sbix_len = sbix_blob->length;
unsigned int strike_offset = (const char *) this - (const char *) sbix_blob->data;
assert (strike_offset < sbix_len);
retry:
if (unlikely (glyph_id >= num_glyphs ||
imageOffsetsZ[glyph_id + 1] <= imageOffsetsZ[glyph_id] ||
imageOffsetsZ[glyph_id + 1] - imageOffsetsZ[glyph_id] <= SBIXGlyph::min_size ||
(unsigned int) imageOffsetsZ[glyph_id + 1] > sbix_len - strike_offset))
return hb_blob_get_empty ();
unsigned int glyph_offset = strike_offset + (unsigned int) imageOffsetsZ[glyph_id] + SBIXGlyph::min_size;
unsigned int glyph_length = imageOffsetsZ[glyph_id + 1] - imageOffsetsZ[glyph_id] - SBIXGlyph::min_size;
const SBIXGlyph *glyph = &(this+imageOffsetsZ[glyph_id]);
if (glyph->graphicType == HB_TAG ('d','u','p','e'))
{
if (glyph_length >= 2)
{
glyph_id = *((HBUINT16 *) &glyph->data);
if (retry_count--)
goto retry;
}
return hb_blob_get_empty ();
}
if (unlikely (file_type != glyph->graphicType))
return hb_blob_get_empty ();
if (strike_ppem) *strike_ppem = ppem;
if (x_offset) *x_offset = glyph->xOffset;
if (y_offset) *y_offset = glyph->yOffset;
return hb_blob_create_sub_blob (sbix_blob, glyph_offset, glyph_length);
}
bool subset (hb_subset_context_t *c, unsigned int available_len) const
{
TRACE_SUBSET (this);
unsigned int num_output_glyphs = c->plan->num_output_glyphs ();
auto* out = c->serializer->start_embed<SBIXStrike> ();
auto snap = c->serializer->snapshot ();
if (unlikely (!c->serializer->extend (out, num_output_glyphs + 1))) return_trace (false);
out->ppem = ppem;
out->resolution = resolution;
HBUINT32 head;
head = get_size (num_output_glyphs + 1);
bool has_glyphs = false;
for (unsigned new_gid = 0; new_gid < num_output_glyphs; new_gid++)
{
hb_codepoint_t old_gid;
if (!c->plan->old_gid_for_new_gid (new_gid, &old_gid) ||
unlikely (imageOffsetsZ[old_gid].is_null () ||
imageOffsetsZ[old_gid + 1].is_null () ||
imageOffsetsZ[old_gid + 1] <= imageOffsetsZ[old_gid] ||
imageOffsetsZ[old_gid + 1] - imageOffsetsZ[old_gid] <= SBIXGlyph::min_size) ||
(unsigned int) imageOffsetsZ[old_gid + 1] > available_len)
{
out->imageOffsetsZ[new_gid] = head;
continue;
}
has_glyphs = true;
unsigned int delta = imageOffsetsZ[old_gid + 1] - imageOffsetsZ[old_gid];
unsigned int glyph_data_length = delta - SBIXGlyph::min_size;
if (!(this+imageOffsetsZ[old_gid]).copy (c->serializer, glyph_data_length))
return_trace (false);
out->imageOffsetsZ[new_gid] = head;
head += delta;
}
if (has_glyphs)
out->imageOffsetsZ[num_output_glyphs] = head;
else
c->serializer->revert (snap);
return_trace (has_glyphs);
}
public:
HBUINT16 ppem; /* The PPEM size for which this strike was designed. */
HBUINT16 resolution; /* The device pixel density (in PPI) for which this
* strike was designed. (E.g., 96 PPI, 192 PPI.) */
protected:
UnsizedArrayOf<Offset32To<SBIXGlyph>>
imageOffsetsZ; /* Offset from the beginning of the strike data header
* to bitmap data for an individual glyph ID. */
public:
DEFINE_SIZE_ARRAY (4, imageOffsetsZ);
};
struct sbix
{
static constexpr hb_tag_t tableTag = HB_OT_TAG_sbix;
bool has_data () const { return version; }
const SBIXStrike &get_strike (unsigned int i) const { return this+strikes[i]; }
struct accelerator_t
{
accelerator_t (hb_face_t *face)
{
table = hb_sanitize_context_t ().reference_table<sbix> (face);
num_glyphs = face->get_num_glyphs ();
}
~accelerator_t () { table.destroy (); }
bool has_data () const { return table->has_data (); }
bool get_extents (hb_font_t *font,
hb_codepoint_t glyph,
hb_glyph_extents_t *extents,
bool scale = true) const
{
/* We only support PNG right now, and following function checks type. */
return get_png_extents (font, glyph, extents, scale);
}
hb_blob_t *reference_png (hb_font_t *font,
hb_codepoint_t glyph_id,
int *x_offset,
int *y_offset,
unsigned int *available_ppem) const
{
return choose_strike (font).get_glyph_blob (glyph_id, table.get_blob (),
HB_TAG ('p','n','g',' '),
x_offset, y_offset,
num_glyphs, available_ppem);
}
bool paint_glyph (hb_font_t *font, hb_codepoint_t glyph, hb_paint_funcs_t *funcs, void *data) const
{
if (!has_data ())
return false;
int x_offset = 0, y_offset = 0;
unsigned int strike_ppem = 0;
hb_glyph_extents_t extents;
hb_glyph_extents_t pixel_extents;
if (!font->get_glyph_extents (glyph, &extents, false))
return false;
if (unlikely (!get_extents (font, glyph, &pixel_extents, false)))
return false;
hb_blob_t *blob = reference_png (font, glyph, &x_offset, &y_offset, &strike_ppem);
if (hb_blob_is_immutable (blob))
return false;
bool ret = funcs->image (data,
blob,
pixel_extents.width, -pixel_extents.height,
HB_PAINT_IMAGE_FORMAT_PNG,
0.f,
&extents);
hb_blob_destroy (blob);
return ret;
}
private:
const SBIXStrike &choose_strike (hb_font_t *font) const
{
unsigned count = table->strikes.len;
if (unlikely (!count))
return Null (SBIXStrike);
unsigned int requested_ppem = hb_max (font->x_ppem, font->y_ppem);
if (!requested_ppem)
requested_ppem = 1<<30; /* Choose largest strike. */
/* TODO Add DPI sensitivity as well? */
unsigned int best_i = 0;
unsigned int best_ppem = table->get_strike (0).ppem;
for (unsigned int i = 1; i < count; i++)
{
unsigned int ppem = (table->get_strike (i)).ppem;
if ((requested_ppem <= ppem && ppem < best_ppem) ||
(requested_ppem > best_ppem && ppem > best_ppem))
{
best_i = i;
best_ppem = ppem;
}
}
return table->get_strike (best_i);
}
struct PNGHeader
{
HBUINT8 signature[8];
struct
{
struct
{
HBUINT32 length;
Tag type;
} header;
HBUINT32 width;
HBUINT32 height;
HBUINT8 bitDepth;
HBUINT8 colorType;
HBUINT8 compressionMethod;
HBUINT8 filterMethod;
HBUINT8 interlaceMethod;
} IHDR;
public:
DEFINE_SIZE_STATIC (29);
};
bool get_png_extents (hb_font_t *font,
hb_codepoint_t glyph,
hb_glyph_extents_t *extents,
bool scale = true) const
{
/* Following code is safe to call even without data.
* But faster to short-circuit. */
if (!has_data ())
return false;
int x_offset = 0, y_offset = 0;
unsigned int strike_ppem = 0;
hb_blob_t *blob = reference_png (font, glyph, &x_offset, &y_offset, &strike_ppem);
const PNGHeader &png = *blob->as<PNGHeader>();
if (png.IHDR.height >= 65536 || png.IHDR.width >= 65536)
{
hb_blob_destroy (blob);
return false;
}
extents->x_bearing = x_offset;
extents->y_bearing = png.IHDR.height + y_offset;
extents->width = png.IHDR.width;
extents->height = -1 * png.IHDR.height;
/* Convert to font units. */
if (strike_ppem && scale)
{
float scale = font->face->get_upem () / (float) strike_ppem;
extents->x_bearing = roundf (extents->x_bearing * scale);
extents->y_bearing = roundf (extents->y_bearing * scale);
extents->width = roundf (extents->width * scale);
extents->height = roundf (extents->height * scale);
}
if (scale)
font->scale_glyph_extents (extents);
hb_blob_destroy (blob);
return strike_ppem;
}
private:
hb_blob_ptr_t<sbix> table;
unsigned int num_glyphs;
};
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (likely (c->check_struct (this) &&
hb_barrier () &&
version >= 1 &&
strikes.sanitize (c, this)));
}
bool
add_strike (hb_subset_context_t *c, unsigned i) const
{
if (strikes[i].is_null () || c->source_blob->length < (unsigned) strikes[i])
return false;
return (this+strikes[i]).subset (c, c->source_blob->length - (unsigned) strikes[i]);
}
bool serialize_strike_offsets (hb_subset_context_t *c) const
{
TRACE_SERIALIZE (this);
auto *out = c->serializer->start_embed<Array32OfOffset32To<SBIXStrike>> ();
if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
hb_vector_t<Offset32To<SBIXStrike>*> new_strikes;
hb_vector_t<hb_serialize_context_t::objidx_t> objidxs;
for (int i = strikes.len - 1; i >= 0; --i)
{
auto* o = out->serialize_append (c->serializer);
if (unlikely (!o)) return_trace (false);
*o = 0;
auto snap = c->serializer->snapshot ();
c->serializer->push ();
bool ret = add_strike (c, i);
if (!ret)
{
c->serializer->pop_discard ();
out->pop ();
c->serializer->revert (snap);
}
else
{
objidxs.push (c->serializer->pop_pack ());
new_strikes.push (o);
}
}
for (unsigned int i = 0; i < new_strikes.length; ++i)
c->serializer->add_link (*new_strikes[i], objidxs[new_strikes.length - 1 - i]);
return_trace (true);
}
bool subset (hb_subset_context_t* c) const
{
TRACE_SUBSET (this);
if (unlikely (!c->serializer->embed (this->version))) return_trace (false);
if (unlikely (!c->serializer->embed (this->flags))) return_trace (false);
return_trace (serialize_strike_offsets (c));
}
protected:
HBUINT16 version; /* Table version number — set to 1 */
HBUINT16 flags; /* Bit 0: Set to 1. Bit 1: Draw outlines.
* Bits 2 to 15: reserved (set to 0). */
Array32OfOffset32To<SBIXStrike>
strikes; /* Offsets from the beginning of the 'sbix'
* table to data for each individual bitmap strike. */
public:
DEFINE_SIZE_ARRAY (8, strikes);
};
struct sbix_accelerator_t : sbix::accelerator_t {
sbix_accelerator_t (hb_face_t *face) : sbix::accelerator_t (face) {}
};
} /* namespace OT */
#endif /* OT_COLOR_SBIX_SBIX_HH */
-829
View File
@@ -1,829 +0,0 @@
/*
* Copyright © 2018 Ebrahim Byagowi
*
* This is part of HarfBuzz, a text shaping library.
*
* Permission is hereby granted, without written agreement and without
* license or royalty fees, to use, copy, modify, and distribute this
* software and its documentation for any purpose, provided that the
* above copyright notice and the following two paragraphs appear in
* all copies of this software.
*
* IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
* DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
* ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
* IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
* DAMAGE.
*
* THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
* BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
* FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
* ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
* PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
*/
#ifndef OT_COLOR_SVG_SVG_HH
#define OT_COLOR_SVG_SVG_HH
#include "../../../hb-open-type.hh"
#include "../../../hb-blob.hh"
#include "../../../hb-limits.hh"
#include "../../../hb-map.hh"
#include "../../../hb-paint.hh"
#include "../../../hb-zlib.hh"
#include <ctype.h>
#include <string.h>
/*
* SVG -- SVG (Scalable Vector Graphics)
* https://docs.microsoft.com/en-us/typography/opentype/spec/svg
*/
#define HB_OT_TAG_SVG HB_TAG('S','V','G',' ')
namespace OT {
static inline hb_blob_t *
hb_ot_svg_reference_normalized_blob (hb_blob_t *image,
const char **svg,
unsigned *len)
{
hb_blob_t *blob = hb_blob_reference (image);
unsigned data_len = 0;
const char *data = hb_blob_get_data (blob, &data_len);
if (!data || !data_len)
goto fail;
if (hb_blob_is_gzip (data, data_len))
{
uint32_t expected_size = 0;
if (hb_gzip_get_uncompressed_size (data, data_len, &expected_size) &&
unlikely ((size_t) expected_size > (size_t) HB_SVG_MAX_DOCUMENT_SIZE))
goto fail;
hb_blob_t *uncompressed = hb_blob_decompress_gzip (blob,
HB_SVG_MAX_DOCUMENT_SIZE);
if (!uncompressed)
goto fail;
hb_blob_destroy (blob);
blob = uncompressed;
data = hb_blob_get_data (blob, &data_len);
if (!data || !data_len)
goto fail;
}
if (unlikely ((size_t) data_len > (size_t) HB_SVG_MAX_DOCUMENT_SIZE))
goto fail;
if (svg) *svg = data;
if (len) *len = data_len;
return blob;
fail:
hb_blob_destroy (blob);
if (svg) *svg = nullptr;
if (len) *len = 0;
return nullptr;
}
struct SVGDocumentIndexEntry
{
int cmp (hb_codepoint_t g) const
{ return g < startGlyphID ? -1 : g > endGlyphID ? 1 : 0; }
hb_codepoint_t get_start_glyph () const
{ return startGlyphID; }
hb_codepoint_t get_end_glyph () const
{ return endGlyphID; }
hb_blob_t *reference_blob (hb_blob_t *svg_blob, unsigned int index_offset) const
{
return hb_blob_create_sub_blob (svg_blob,
index_offset + (unsigned int) svgDoc,
svgDocLength);
}
bool sanitize (hb_sanitize_context_t *c, const void *base) const
{
TRACE_SANITIZE (this);
return_trace (c->check_struct (this) &&
hb_barrier () &&
svgDoc.sanitize (c, base, svgDocLength));
}
protected:
HBUINT16 startGlyphID; /* The first glyph ID in the range described by
* this index entry. */
HBUINT16 endGlyphID; /* The last glyph ID in the range described by
* this index entry. Must be >= startGlyphID. */
NNOffset32To<UnsizedArrayOf<HBUINT8>>
svgDoc; /* Offset from the beginning of the SVG Document Index
* to an SVG document. Must be non-zero. */
HBUINT32 svgDocLength; /* Length of the SVG document.
* Must be non-zero. */
public:
DEFINE_SIZE_STATIC (12);
};
struct SVG
{
static constexpr hb_tag_t tableTag = HB_OT_TAG_SVG;
struct svg_id_span_t
{
const char *p;
unsigned len;
bool operator == (const svg_id_span_t &o) const
{
return len == o.len && !memcmp (p, o.p, len);
}
uint32_t hash () const
{
uint32_t h = hb_hash (len);
for (unsigned i = 0; i < len; i++)
h = h * 33u + (unsigned char) p[i];
return h;
}
};
struct svg_defs_entry_t
{
svg_id_span_t id;
unsigned start;
unsigned end;
};
struct svg_doc_cache_t
{
hb_blob_t *blob = nullptr;
const char *svg = nullptr;
unsigned len = 0;
hb_vector_t<svg_defs_entry_t> defs_entries;
hb_codepoint_t start_glyph = HB_CODEPOINT_INVALID;
hb_codepoint_t end_glyph = HB_CODEPOINT_INVALID;
hb_vector_t<hb_pair_t<uint32_t, uint32_t>> glyph_spans;
hb_hashmap_t<svg_id_span_t, hb_pair_t<uint32_t, uint32_t>> id_spans;
};
bool has_data () const { return svgDocEntries; }
struct accelerator_t
{
accelerator_t (hb_face_t *face);
~accelerator_t ();
hb_blob_t *reference_blob_for_glyph (hb_codepoint_t glyph_id) const
{
return table->get_glyph_entry (glyph_id).reference_blob (table.get_blob (),
table->svgDocEntries);
}
unsigned get_document_count () const
{ return table->get_document_count (); }
bool get_glyph_document_index (hb_codepoint_t glyph_id, unsigned *index) const
{ return table->get_glyph_document_index (glyph_id, index); }
bool get_document_glyph_range (unsigned index,
hb_codepoint_t *start_glyph,
hb_codepoint_t *end_glyph) const
{ return table->get_document_glyph_range (index, start_glyph, end_glyph); }
bool has_data () const { return table->has_data (); }
const svg_doc_cache_t *
get_or_create_doc_cache (hb_blob_t *image,
const char *svg,
unsigned len,
unsigned doc_index,
hb_codepoint_t start_glyph,
hb_codepoint_t end_glyph) const;
const char *
doc_cache_get_svg (const svg_doc_cache_t *doc,
unsigned *len) const;
const hb_vector_t<svg_defs_entry_t> *
doc_cache_get_defs_entries (const svg_doc_cache_t *doc) const;
bool
doc_cache_get_glyph_span (const svg_doc_cache_t *doc,
hb_codepoint_t glyph,
unsigned *start,
unsigned *end) const;
bool
doc_cache_find_id_span (const svg_doc_cache_t *doc,
svg_id_span_t id,
unsigned *start,
unsigned *end) const;
bool
doc_cache_find_id_cstr (const svg_doc_cache_t *doc,
const char *id,
unsigned *start,
unsigned *end) const;
bool paint_glyph (hb_font_t *font HB_UNUSED, hb_codepoint_t glyph, hb_paint_funcs_t *funcs, void *data) const
{
if (!has_data ())
return false;
hb_blob_t *blob = reference_blob_for_glyph (glyph);
if (blob == hb_blob_get_empty ())
return false;
bool ret = funcs->image (data,
blob,
0, 0,
HB_PAINT_IMAGE_FORMAT_SVG,
0.f,
nullptr);
hb_blob_destroy (blob);
return ret;
}
private:
svg_doc_cache_t *
make_doc_cache (hb_blob_t *image,
const char *svg,
unsigned len,
hb_codepoint_t start_glyph,
hb_codepoint_t end_glyph) const;
static void destroy_doc_cache (svg_doc_cache_t *doc);
hb_blob_ptr_t<SVG> table;
mutable hb_vector_t<hb_atomic_t<svg_doc_cache_t *>> doc_caches;
public:
DEFINE_SIZE_STATIC (sizeof (hb_blob_ptr_t<SVG>) +
sizeof (hb_vector_t<hb_atomic_t<svg_doc_cache_t *>>));
};
const SVGDocumentIndexEntry &get_glyph_entry (hb_codepoint_t glyph_id) const
{ return (this+svgDocEntries).bsearch (glyph_id); }
unsigned get_document_count () const
{
if (!has_data ())
return 0;
return (this + svgDocEntries).len;
}
bool get_glyph_document_index (hb_codepoint_t glyph_id, unsigned *index) const
{
if (!has_data ())
return false;
return (this + svgDocEntries).bfind (glyph_id, index);
}
bool get_document_glyph_range (unsigned index,
hb_codepoint_t *start_glyph,
hb_codepoint_t *end_glyph) const
{
if (!has_data ())
return false;
const auto &entries = this + svgDocEntries;
if (index >= entries.len)
return false;
const auto &entry = entries.arrayZ[index];
if (start_glyph) *start_glyph = entry.get_start_glyph ();
if (end_glyph) *end_glyph = entry.get_end_glyph ();
return true;
}
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (likely (c->check_struct (this) &&
(this+svgDocEntries).sanitize_shallow (c)));
}
protected:
HBUINT16 version; /* Table version (starting at 0). */
Offset32To<SortedArray16Of<SVGDocumentIndexEntry>>
svgDocEntries; /* Offset (relative to the start of the SVG table) to the
* SVG Documents Index. Must be non-zero. */
/* Array of SVG Document Index Entries. */
HBUINT32 reserved; /* Set to 0. */
public:
DEFINE_SIZE_STATIC (10);
};
namespace _hb_svg_cache_impl {
struct glyph_entry_t
{
hb_codepoint_t glyph;
uint32_t start;
uint32_t end;
};
struct id_entry_t
{
SVG::svg_id_span_t id;
uint32_t start;
uint32_t end;
};
struct open_elem_t
{
unsigned start;
SVG::svg_id_span_t id;
bool in_defs_content;
bool is_defs;
};
static const unsigned MAX_DEPTH = 128;
static inline int
find_substr (const char *s,
unsigned n,
unsigned from,
const char *needle,
unsigned needle_len)
{
if (!needle_len || from >= n || needle_len > n)
return -1;
for (unsigned i = from; i + needle_len <= n; i++)
if (s[i] == needle[0] && !memcmp (s + i, needle, needle_len))
return (int) i;
return -1;
}
static inline bool
parse_id_in_start_tag (const char *svg,
unsigned tag_start,
unsigned tag_end,
SVG::svg_id_span_t *id)
{
unsigned p = tag_start;
while (p + 4 <= tag_end)
{
if (!memcmp (svg + p, "id=\"", 4))
{
unsigned b = p + 4;
unsigned e = b;
while (e < tag_end && svg[e] != '"') e++;
if (e <= tag_end && e > b)
{
*id = {svg + b, e - b};
return true;
}
}
if (!memcmp (svg + p, "id='", 4))
{
unsigned b = p + 4;
unsigned e = b;
while (e < tag_end && svg[e] != '\'') e++;
if (e <= tag_end && e > b)
{
*id = {svg + b, e - b};
return true;
}
}
p++;
}
return false;
}
static inline bool
parse_glyph_id_span (const SVG::svg_id_span_t &id,
hb_codepoint_t *glyph)
{
if (id.len <= 5 || memcmp (id.p, "glyph", 5))
return false;
hb_codepoint_t gid = 0;
for (unsigned i = 5; i < id.len; i++)
{
unsigned char c = (unsigned char) id.p[i];
if (c < '0' || c > '9')
return false;
hb_codepoint_t digit = (hb_codepoint_t) (c - '0');
if (unlikely (gid > HB_CODEPOINT_INVALID / 10 ||
(gid == HB_CODEPOINT_INVALID / 10 &&
digit > HB_CODEPOINT_INVALID % 10)))
return false;
gid = (hb_codepoint_t) (gid * 10 + digit);
}
*glyph = gid;
return true;
}
static inline bool
parse_cache_entries_linear (const char *svg,
unsigned len,
hb_vector_t<SVG::svg_defs_entry_t> *defs_entries,
hb_vector_t<glyph_entry_t> *glyph_spans,
hb_vector_t<id_entry_t> *id_entries)
{
open_elem_t stack[MAX_DEPTH] = {};
unsigned depth = 0;
if (unlikely (!defs_entries->alloc (256) ||
!glyph_spans->alloc (256) ||
!id_entries->alloc (256)))
return false;
unsigned defs_depth = 0;
unsigned i = 0;
while (i < len)
{
if (svg[i] != '<')
{
i++;
continue;
}
if (i + 4 <= len && !memcmp (svg + i, "<!--", 4))
{
int cend = find_substr (svg, len, i + 4, "-->", 3);
if (cend < 0) return false;
i = (unsigned) cend + 3;
continue;
}
if (i + 9 <= len && !memcmp (svg + i, "<![CDATA[", 9))
{
int cend = find_substr (svg, len, i + 9, "]]>", 3);
if (cend < 0) return false;
i = (unsigned) cend + 3;
continue;
}
bool closing = (i + 1 < len && svg[i + 1] == '/');
bool special = (i + 1 < len && (svg[i + 1] == '!' || svg[i + 1] == '?'));
unsigned gt = i + 1;
char quote = 0;
while (gt < len)
{
char c = svg[gt];
if (quote)
{
if (c == quote) quote = 0;
}
else
{
if (c == '"' || c == '\'')
quote = c;
else if (c == '>')
break;
}
gt++;
}
if (gt >= len)
return false;
if (special)
{
i = gt + 1;
continue;
}
unsigned p = i + (closing ? 2 : 1);
while (p < gt && isspace ((unsigned char) svg[p])) p++;
const char *name = svg + p;
unsigned name_len = 0;
while (p + name_len < gt)
{
unsigned char c = (unsigned char) name[name_len];
if (!(isalnum (c) || c == '_' || c == '-' || c == ':'))
break;
name_len++;
}
bool is_defs = (name_len == 4 && !memcmp (name, "defs", 4));
if (closing)
{
if (!depth)
{
i = gt + 1;
continue;
}
open_elem_t e = stack[--depth];
unsigned end = gt + 1;
if (e.id.len)
{
if (unlikely (!id_entries->push_or_fail (id_entry_t {e.id, (uint32_t) e.start, (uint32_t) end})))
return false;
if (e.in_defs_content)
{
if (unlikely (!defs_entries->push_or_fail ()))
return false;
auto &slot = defs_entries->tail ();
slot.id = e.id;
slot.start = e.start;
slot.end = end;
}
hb_codepoint_t gid;
if (parse_glyph_id_span (e.id, &gid))
{
if (unlikely (!glyph_spans->push_or_fail (glyph_entry_t {gid, (uint32_t) e.start, (uint32_t) end})))
return false;
}
}
if (e.is_defs && defs_depth)
defs_depth--;
i = end;
continue;
}
SVG::svg_id_span_t id = {};
parse_id_in_start_tag (svg, i, gt, &id);
unsigned r = gt;
while (r > i && isspace ((unsigned char) svg[r - 1])) r--;
bool self_closing = (r > i && svg[r - 1] == '/');
open_elem_t e = {};
e.start = i;
e.id = id;
e.in_defs_content = defs_depth > 0;
e.is_defs = is_defs;
if (self_closing)
{
unsigned end = gt + 1;
if (e.id.len)
{
if (unlikely (!id_entries->push_or_fail (id_entry_t {e.id, (uint32_t) e.start, (uint32_t) end})))
return false;
if (e.in_defs_content)
{
if (unlikely (!defs_entries->push_or_fail ()))
return false;
auto &slot = defs_entries->tail ();
slot.id = e.id;
slot.start = e.start;
slot.end = end;
}
hb_codepoint_t gid;
if (parse_glyph_id_span (e.id, &gid))
{
if (unlikely (!glyph_spans->push_or_fail (glyph_entry_t {gid, (uint32_t) e.start, (uint32_t) end})))
return false;
}
}
}
else
{
if (unlikely (depth >= MAX_DEPTH))
return false;
stack[depth++] = e;
if (is_defs)
defs_depth++;
}
i = gt + 1;
}
return true;
}
} /* namespace _hb_svg_cache_impl */
inline
SVG::accelerator_t::accelerator_t (hb_face_t *face)
{
table = hb_sanitize_context_t ().reference_table<SVG> (face);
doc_caches.init ();
unsigned doc_count = table->get_document_count ();
if (doc_count && unlikely (!doc_caches.resize (doc_count)))
doc_caches.clear ();
for (unsigned i = 0; i < doc_caches.length; i++)
doc_caches.arrayZ[i].set_relaxed (nullptr);
}
inline
SVG::accelerator_t::~accelerator_t ()
{
for (unsigned i = 0; i < doc_caches.length; i++)
destroy_doc_cache (doc_caches.arrayZ[i].get_relaxed ());
doc_caches.fini ();
table.destroy ();
}
inline void
SVG::accelerator_t::destroy_doc_cache (svg_doc_cache_t *doc)
{
if (!doc)
return;
doc->glyph_spans.fini ();
doc->defs_entries.fini ();
doc->id_spans.fini ();
hb_blob_destroy (doc->blob);
hb_free (doc);
}
inline SVG::svg_doc_cache_t *
SVG::accelerator_t::make_doc_cache (hb_blob_t *image,
const char *svg,
unsigned len,
hb_codepoint_t start_glyph,
hb_codepoint_t end_glyph) const
{
static const uint32_t INVALID_SPAN = 0xFFFFFFFFu;
auto *doc = (svg_doc_cache_t *) hb_malloc (sizeof (svg_doc_cache_t));
if (!doc)
return nullptr;
doc->blob = nullptr;
doc->svg = nullptr;
doc->len = 0;
doc->defs_entries.init ();
doc->start_glyph = HB_CODEPOINT_INVALID;
doc->end_glyph = HB_CODEPOINT_INVALID;
doc->glyph_spans.init ();
doc->id_spans.init ();
doc->blob = hb_blob_reference (image);
doc->svg = svg;
doc->len = len;
doc->start_glyph = start_glyph;
doc->end_glyph = end_glyph;
if (unlikely (start_glyph == HB_CODEPOINT_INVALID || end_glyph < start_glyph))
{
destroy_doc_cache (doc);
return nullptr;
}
unsigned glyph_count = end_glyph - start_glyph + 1;
if (!doc->glyph_spans.resize ((int) glyph_count))
{
destroy_doc_cache (doc);
return nullptr;
}
for (unsigned i = 0; i < glyph_count; i++)
doc->glyph_spans.arrayZ[i] = hb_pair_t<uint32_t, uint32_t> (INVALID_SPAN, INVALID_SPAN);
hb_vector_t<_hb_svg_cache_impl::glyph_entry_t> glyph_spans;
glyph_spans.init ();
hb_vector_t<_hb_svg_cache_impl::id_entry_t> id_entries;
id_entries.init ();
if (!_hb_svg_cache_impl::parse_cache_entries_linear (svg, len,
&doc->defs_entries,
&glyph_spans,
&id_entries))
{
id_entries.fini ();
glyph_spans.fini ();
destroy_doc_cache (doc);
return nullptr;
}
for (unsigned i = 0; i < glyph_spans.length; i++)
{
const auto &span = glyph_spans.arrayZ[i];
if (unlikely (span.glyph < start_glyph || span.glyph > end_glyph))
continue;
doc->glyph_spans.arrayZ[span.glyph - start_glyph] = hb_pair_t<uint32_t, uint32_t> (span.start, span.end);
}
for (unsigned i = 0; i < id_entries.length; i++)
{
const auto &e = id_entries.arrayZ[i];
hb_pair_t<uint32_t, uint32_t> *out = nullptr;
if (doc->id_spans.has (e.id, &out))
continue;
if (unlikely (!doc->id_spans.set (e.id, hb_pair_t<uint32_t, uint32_t> (e.start, e.end))))
{
id_entries.fini ();
glyph_spans.fini ();
destroy_doc_cache (doc);
return nullptr;
}
}
id_entries.fini ();
glyph_spans.fini ();
return doc;
}
inline const SVG::svg_doc_cache_t *
SVG::accelerator_t::get_or_create_doc_cache (hb_blob_t *image,
const char *svg,
unsigned len,
unsigned doc_index,
hb_codepoint_t start_glyph,
hb_codepoint_t end_glyph) const
{
if (doc_index >= doc_caches.length)
return nullptr;
auto &slot = doc_caches.arrayZ[doc_index];
auto *doc = slot.get_acquire ();
if (doc)
return doc;
auto *fresh = make_doc_cache (image, svg, len, start_glyph, end_glyph);
if (!fresh)
return nullptr;
auto *expected = (svg_doc_cache_t *) nullptr;
if (slot.cmpexch (expected, fresh))
return fresh;
destroy_doc_cache (fresh);
return expected;
}
inline const char *
SVG::accelerator_t::doc_cache_get_svg (const svg_doc_cache_t *doc,
unsigned *len) const
{
if (!doc)
{
if (len) *len = 0;
return nullptr;
}
if (len) *len = doc->len;
return doc->svg;
}
inline const hb_vector_t<SVG::svg_defs_entry_t> *
SVG::accelerator_t::doc_cache_get_defs_entries (const svg_doc_cache_t *doc) const
{
return doc ? &doc->defs_entries : nullptr;
}
inline bool
SVG::accelerator_t::doc_cache_get_glyph_span (const svg_doc_cache_t *doc,
hb_codepoint_t glyph,
unsigned *start,
unsigned *end) const
{
static const uint32_t INVALID_SPAN = 0xFFFFFFFFu;
if (!doc || doc->start_glyph == HB_CODEPOINT_INVALID ||
glyph < doc->start_glyph || glyph > doc->end_glyph)
return false;
const auto &span = doc->glyph_spans.arrayZ[glyph - doc->start_glyph];
if (span.first == INVALID_SPAN)
return false;
if (unlikely (span.first > span.second || span.second > doc->len))
return false;
if (start) *start = span.first;
if (end) *end = span.second;
return true;
}
inline bool
SVG::accelerator_t::doc_cache_find_id_span (const svg_doc_cache_t *doc,
svg_id_span_t id,
unsigned *start,
unsigned *end) const
{
if (!doc || !id.p || !id.len)
return false;
hb_pair_t<uint32_t, uint32_t> *span = nullptr;
if (!doc->id_spans.has (id, &span))
return false;
if (unlikely (span->first > span->second || span->second > doc->len))
return false;
if (start) *start = span->first;
if (end) *end = span->second;
return true;
}
inline bool
SVG::accelerator_t::doc_cache_find_id_cstr (const svg_doc_cache_t *doc,
const char *id,
unsigned *start,
unsigned *end) const
{
if (!id) return false;
svg_id_span_t key = {id, (unsigned) strlen (id)};
return doc_cache_find_id_span (doc, key, start, end);
}
struct SVG_accelerator_t : SVG::accelerator_t {
SVG_accelerator_t (hb_face_t *face) : SVG::accelerator_t (face) {}
};
} /* namespace OT */
#endif /* OT_COLOR_SVG_SVG_HH */
-394
View File
@@ -1,394 +0,0 @@
/*
* Copyright © 2007,2008,2009 Red Hat, Inc.
* Copyright © 2010,2012 Google, Inc.
*
* This is part of HarfBuzz, a text shaping library.
*
* Permission is hereby granted, without written agreement and without
* license or royalty fees, to use, copy, modify, and distribute this
* software and its documentation for any purpose, provided that the
* above copyright notice and the following two paragraphs appear in
* all copies of this software.
*
* IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
* DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
* ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
* IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
* DAMAGE.
*
* THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
* BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
* FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
* ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
* PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
*
* Red Hat Author(s): Behdad Esfahbod
* Google Author(s): Behdad Esfahbod, Garret Rieger
*/
#ifndef OT_LAYOUT_COMMON_COVERAGE_HH
#define OT_LAYOUT_COMMON_COVERAGE_HH
#include "../types.hh"
#include "CoverageFormat1.hh"
#include "CoverageFormat2.hh"
namespace OT {
namespace Layout {
namespace Common {
template<typename Iterator>
static inline void Coverage_serialize (hb_serialize_context_t *c,
Iterator it);
struct Coverage
{
protected:
union {
struct { HBUINT16 v; } format; /* Format identifier */
CoverageFormat1_3<SmallTypes> format1;
CoverageFormat2_4<SmallTypes> format2;
#ifndef HB_NO_BEYOND_64K
CoverageFormat1_3<MediumTypes>format3;
CoverageFormat2_4<MediumTypes>format4;
#endif
} u;
public:
DEFINE_SIZE_UNION (2, format.v);
#ifndef HB_OPTIMIZE_SIZE
HB_ALWAYS_INLINE
#endif
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
if (!u.format.v.sanitize (c)) return_trace (false);
hb_barrier ();
switch (u.format.v)
{
case 1: return_trace (u.format1.sanitize (c));
case 2: return_trace (u.format2.sanitize (c));
#ifndef HB_NO_BEYOND_64K
case 3: return_trace (u.format3.sanitize (c));
case 4: return_trace (u.format4.sanitize (c));
#endif
default:return_trace (true);
}
}
/* Has interface. */
unsigned operator [] (hb_codepoint_t k) const { return get (k); }
bool has (hb_codepoint_t k) const { return (*this)[k] != NOT_COVERED; }
/* Predicate. */
bool operator () (hb_codepoint_t k) const { return has (k); }
unsigned int get (hb_codepoint_t k) const { return get_coverage (k); }
unsigned int get_coverage (hb_codepoint_t glyph_id) const
{
switch (u.format.v) {
case 1: return u.format1.get_coverage (glyph_id);
case 2: return u.format2.get_coverage (glyph_id);
#ifndef HB_NO_BEYOND_64K
case 3: return u.format3.get_coverage (glyph_id);
case 4: return u.format4.get_coverage (glyph_id);
#endif
default:return NOT_COVERED;
}
}
unsigned int get_coverage (hb_codepoint_t glyph_id,
hb_ot_layout_mapping_cache_t *cache) const
{
unsigned coverage;
if (cache && cache->get (glyph_id, &coverage)) return coverage < cache->MAX_VALUE ? coverage : NOT_COVERED;
coverage = get_coverage (glyph_id);
if (cache) {
if (coverage == NOT_COVERED)
cache->set_unchecked (glyph_id, cache->MAX_VALUE);
else if (likely (coverage < cache->MAX_VALUE))
cache->set_unchecked (glyph_id, coverage);
}
return coverage;
}
unsigned int get_coverage_binary (hb_codepoint_t glyph_id,
hb_ot_layout_binary_cache_t *cache) const
{
unsigned coverage;
if (cache && cache->get (glyph_id, &coverage)) return coverage < cache->MAX_VALUE ? coverage : NOT_COVERED;
coverage = get_coverage (glyph_id);
if (cache) {
if (coverage == NOT_COVERED)
cache->set_unchecked (glyph_id, cache->MAX_VALUE);
else
cache->set_unchecked (glyph_id, 0);
}
return coverage;
}
unsigned get_population () const
{
switch (u.format.v) {
case 1: return u.format1.get_population ();
case 2: return u.format2.get_population ();
#ifndef HB_NO_BEYOND_64K
case 3: return u.format3.get_population ();
case 4: return u.format4.get_population ();
#endif
default:return NOT_COVERED;
}
}
template <typename Iterator,
hb_requires (hb_is_sorted_source_of (Iterator, hb_codepoint_t))>
bool serialize (hb_serialize_context_t *c, Iterator glyphs)
{
TRACE_SERIALIZE (this);
if (unlikely (!c->extend_min (this))) return_trace (false);
unsigned count = hb_len (glyphs);
unsigned num_ranges = 0;
hb_codepoint_t last = (hb_codepoint_t) -2;
hb_codepoint_t max = 0;
bool unsorted = false;
for (auto g: glyphs)
{
if (last != (hb_codepoint_t) -2 && g < last)
unsorted = true;
if (last + 1 != g)
num_ranges++;
last = g;
if (g > max) max = g;
}
u.format.v = !unsorted && count <= num_ranges * 3 ? 1 : 2;
#ifndef HB_NO_BEYOND_64K
if (max > 0xFFFFu)
u.format.v += 2;
if (unlikely (max > 0xFFFFFFu))
#else
if (unlikely (max > 0xFFFFu))
#endif
{
c->check_success (false, HB_SERIALIZE_ERROR_INT_OVERFLOW);
return_trace (false);
}
switch (u.format.v)
{
case 1: return_trace (u.format1.serialize (c, glyphs));
case 2: return_trace (u.format2.serialize (c, glyphs));
#ifndef HB_NO_BEYOND_64K
case 3: return_trace (u.format3.serialize (c, glyphs));
case 4: return_trace (u.format4.serialize (c, glyphs));
#endif
default:return_trace (false);
}
}
bool subset (hb_subset_context_t *c) const
{
TRACE_SUBSET (this);
auto it =
+ iter ()
| hb_take (c->plan->source->get_num_glyphs ())
| hb_map_retains_sorting (c->plan->glyph_map_gsub)
| hb_filter ([] (hb_codepoint_t glyph) { return glyph != HB_MAP_VALUE_INVALID; })
;
// Cache the iterator result as it will be iterated multiple times
// by the serialize code below.
hb_sorted_vector_t<hb_codepoint_t> glyphs (it);
Coverage_serialize (c->serializer, glyphs.iter ());
return_trace (bool (glyphs));
}
bool intersects (const hb_set_t *glyphs) const
{
switch (u.format.v)
{
case 1: return u.format1.intersects (glyphs);
case 2: return u.format2.intersects (glyphs);
#ifndef HB_NO_BEYOND_64K
case 3: return u.format3.intersects (glyphs);
case 4: return u.format4.intersects (glyphs);
#endif
default:return false;
}
}
bool intersects_coverage (const hb_set_t *glyphs, unsigned int index) const
{
switch (u.format.v)
{
case 1: return u.format1.intersects_coverage (glyphs, index);
case 2: return u.format2.intersects_coverage (glyphs, index);
#ifndef HB_NO_BEYOND_64K
case 3: return u.format3.intersects_coverage (glyphs, index);
case 4: return u.format4.intersects_coverage (glyphs, index);
#endif
default:return false;
}
}
unsigned cost () const
{
switch (u.format.v) {
case 1: hb_barrier (); return u.format1.cost ();
case 2: hb_barrier (); return u.format2.cost ();
#ifndef HB_NO_BEYOND_64K
case 3: hb_barrier (); return u.format3.cost ();
case 4: hb_barrier (); return u.format4.cost ();
#endif
default:return 0u;
}
}
/* Might return false if array looks unsorted.
* Used for faster rejection of corrupt data. */
template <typename set_t>
bool collect_coverage (set_t *glyphs) const
{
switch (u.format.v)
{
case 1: return u.format1.collect_coverage (glyphs);
case 2: return u.format2.collect_coverage (glyphs);
#ifndef HB_NO_BEYOND_64K
case 3: return u.format3.collect_coverage (glyphs);
case 4: return u.format4.collect_coverage (glyphs);
#endif
default:return false;
}
}
template <typename IterableOut,
hb_requires (hb_is_sink_of (IterableOut, hb_codepoint_t))>
void intersect_set (const hb_set_t &glyphs, IterableOut&& intersect_glyphs) const
{
switch (u.format.v)
{
case 1: return u.format1.intersect_set (glyphs, intersect_glyphs);
case 2: return u.format2.intersect_set (glyphs, intersect_glyphs);
#ifndef HB_NO_BEYOND_64K
case 3: return u.format3.intersect_set (glyphs, intersect_glyphs);
case 4: return u.format4.intersect_set (glyphs, intersect_glyphs);
#endif
default:return ;
}
}
struct iter_t : hb_iter_with_fallback_t<iter_t, hb_codepoint_t>
{
static constexpr bool is_sorted_iterator = true;
iter_t (const Coverage &c_ = Null (Coverage))
{
hb_memset (this, 0, sizeof (*this));
format = c_.u.format.v;
switch (format)
{
case 1: u.format1.init (c_.u.format1); return;
case 2: u.format2.init (c_.u.format2); return;
#ifndef HB_NO_BEYOND_64K
case 3: u.format3.init (c_.u.format3); return;
case 4: u.format4.init (c_.u.format4); return;
#endif
default: return;
}
}
bool __more__ () const
{
switch (format)
{
case 1: return u.format1.__more__ ();
case 2: return u.format2.__more__ ();
#ifndef HB_NO_BEYOND_64K
case 3: return u.format3.__more__ ();
case 4: return u.format4.__more__ ();
#endif
default:return false;
}
}
void __next__ ()
{
switch (format)
{
case 1: u.format1.__next__ (); break;
case 2: u.format2.__next__ (); break;
#ifndef HB_NO_BEYOND_64K
case 3: u.format3.__next__ (); break;
case 4: u.format4.__next__ (); break;
#endif
default: break;
}
}
typedef hb_codepoint_t __item_t__;
__item_t__ __item__ () const { return get_glyph (); }
hb_codepoint_t get_glyph () const
{
switch (format)
{
case 1: return u.format1.get_glyph ();
case 2: return u.format2.get_glyph ();
#ifndef HB_NO_BEYOND_64K
case 3: return u.format3.get_glyph ();
case 4: return u.format4.get_glyph ();
#endif
default:return 0;
}
}
bool operator != (const iter_t& o) const
{
if (unlikely (format != o.format)) return true;
switch (format)
{
case 1: return u.format1 != o.u.format1;
case 2: return u.format2 != o.u.format2;
#ifndef HB_NO_BEYOND_64K
case 3: return u.format3 != o.u.format3;
case 4: return u.format4 != o.u.format4;
#endif
default:return false;
}
}
iter_t __end__ () const
{
iter_t it;
it.format = format;
switch (format)
{
case 1: it.u.format1 = u.format1.__end__ (); break;
case 2: it.u.format2 = u.format2.__end__ (); break;
#ifndef HB_NO_BEYOND_64K
case 3: it.u.format3 = u.format3.__end__ (); break;
case 4: it.u.format4 = u.format4.__end__ (); break;
#endif
default: break;
}
return it;
}
private:
unsigned int format;
union {
#ifndef HB_NO_BEYOND_64K
CoverageFormat2_4<MediumTypes>::iter_t format4; /* Put this one first since it's larger; helps shut up compiler. */
CoverageFormat1_3<MediumTypes>::iter_t format3;
#endif
CoverageFormat2_4<SmallTypes>::iter_t format2; /* Put this one first since it's larger; helps shut up compiler. */
CoverageFormat1_3<SmallTypes>::iter_t format1;
} u;
};
iter_t iter () const { return iter_t (*this); }
};
template<typename Iterator>
static inline void
Coverage_serialize (hb_serialize_context_t *c,
Iterator it)
{ c->start_embed<Coverage> ()->serialize (c, it); }
}
}
}
#endif // #ifndef OT_LAYOUT_COMMON_COVERAGE_HH
@@ -1,135 +0,0 @@
/*
* Copyright © 2007,2008,2009 Red Hat, Inc.
* Copyright © 2010,2012 Google, Inc.
*
* This is part of HarfBuzz, a text shaping library.
*
* Permission is hereby granted, without written agreement and without
* license or royalty fees, to use, copy, modify, and distribute this
* software and its documentation for any purpose, provided that the
* above copyright notice and the following two paragraphs appear in
* all copies of this software.
*
* IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
* DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
* ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
* IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
* DAMAGE.
*
* THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
* BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
* FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
* ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
* PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
*
* Red Hat Author(s): Behdad Esfahbod
* Google Author(s): Behdad Esfahbod, Garret Rieger
*/
#ifndef OT_LAYOUT_COMMON_COVERAGEFORMAT1_HH
#define OT_LAYOUT_COMMON_COVERAGEFORMAT1_HH
namespace OT {
namespace Layout {
namespace Common {
#define NOT_COVERED ((unsigned int) -1)
template <typename Types>
struct CoverageFormat1_3
{
friend struct Coverage;
public:
HBUINT16 coverageFormat; /* Format identifier--format = 1 */
SortedArray16Of<typename Types::HBGlyphID>
glyphArray; /* Array of GlyphIDs--in numerical order */
DEFINE_SIZE_ARRAY (4, glyphArray);
private:
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (glyphArray.sanitize (c));
}
unsigned int get_coverage (hb_codepoint_t glyph_id) const
{
unsigned int i;
glyphArray.bfind (glyph_id, &i, HB_NOT_FOUND_STORE, NOT_COVERED);
return i;
}
unsigned get_population () const
{
return glyphArray.len;
}
template <typename Iterator,
hb_requires (hb_is_sorted_source_of (Iterator, hb_codepoint_t))>
bool serialize (hb_serialize_context_t *c, Iterator glyphs)
{
TRACE_SERIALIZE (this);
return_trace (glyphArray.serialize (c, glyphs));
}
bool intersects (const hb_set_t *glyphs) const
{
if (glyphArray.len > glyphs->get_population () * hb_bit_storage ((unsigned) glyphArray.len))
{
for (auto g : *glyphs)
if (get_coverage (g) != NOT_COVERED)
return true;
return false;
}
for (const auto& g : glyphArray.as_array ())
if (glyphs->has (g))
return true;
return false;
}
bool intersects_coverage (const hb_set_t *glyphs, unsigned int index) const
{ return glyphs->has (glyphArray[index]); }
template <typename IterableOut,
hb_requires (hb_is_sink_of (IterableOut, hb_codepoint_t))>
void intersect_set (const hb_set_t &glyphs, IterableOut&& intersect_glyphs) const
{
unsigned count = glyphArray.len;
for (unsigned i = 0; i < count; i++)
if (glyphs.has (glyphArray[i]))
intersect_glyphs << glyphArray[i];
}
unsigned cost () const { return hb_bit_storage ((unsigned) glyphArray.len); /* bsearch cost */ }
template <typename set_t>
bool collect_coverage (set_t *glyphs) const
{ return glyphs->add_sorted_array (glyphArray.as_array ()); }
public:
/* Older compilers need this to be public. */
struct iter_t
{
void init (const struct CoverageFormat1_3 &c_) { c = &c_; i = 0; }
bool __more__ () const { return i < c->glyphArray.len; }
void __next__ () { i++; }
hb_codepoint_t get_glyph () const { return c->glyphArray[i]; }
bool operator != (const iter_t& o) const
{ return i != o.i; }
iter_t __end__ () const { iter_t it; it.init (*c); it.i = c->glyphArray.len; return it; }
private:
const struct CoverageFormat1_3 *c;
unsigned int i;
};
private:
};
}
}
}
#endif // #ifndef OT_LAYOUT_COMMON_COVERAGEFORMAT1_HH
@@ -1,241 +0,0 @@
/*
* Copyright © 2007,2008,2009 Red Hat, Inc.
* Copyright © 2010,2012 Google, Inc.
*
* This is part of HarfBuzz, a text shaping library.
*
* Permission is hereby granted, without written agreement and without
* license or royalty fees, to use, copy, modify, and distribute this
* software and its documentation for any purpose, provided that the
* above copyright notice and the following two paragraphs appear in
* all copies of this software.
*
* IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
* DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
* ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
* IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
* DAMAGE.
*
* THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
* BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
* FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
* ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
* PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
*
* Red Hat Author(s): Behdad Esfahbod
* Google Author(s): Behdad Esfahbod, Garret Rieger
*/
#ifndef OT_LAYOUT_COMMON_COVERAGEFORMAT2_HH
#define OT_LAYOUT_COMMON_COVERAGEFORMAT2_HH
#include "RangeRecord.hh"
namespace OT {
namespace Layout {
namespace Common {
template <typename Types>
struct CoverageFormat2_4
{
friend struct Coverage;
public:
HBUINT16 coverageFormat; /* Format identifier--format = 2 */
SortedArray16Of<RangeRecord<Types>>
rangeRecord; /* Array of glyph ranges--ordered by
* Start GlyphID. rangeCount entries
* long */
public:
DEFINE_SIZE_ARRAY (4, rangeRecord);
private:
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (rangeRecord.sanitize (c));
}
unsigned int get_coverage (hb_codepoint_t glyph_id) const
{
const RangeRecord<Types> &range = rangeRecord.bsearch (glyph_id);
return likely (range.first <= range.last)
? (unsigned int) range.value + (glyph_id - range.first)
: NOT_COVERED;
}
unsigned get_population () const
{
typename Types::large_int ret = 0;
for (const auto &r : rangeRecord)
ret += r.get_population ();
return ret > UINT_MAX ? UINT_MAX : (unsigned) ret;
}
template <typename Iterator,
hb_requires (hb_is_sorted_source_of (Iterator, hb_codepoint_t))>
bool serialize (hb_serialize_context_t *c, Iterator glyphs)
{
TRACE_SERIALIZE (this);
if (unlikely (!c->extend_min (this))) return_trace (false);
unsigned num_ranges = 0;
hb_codepoint_t last = (hb_codepoint_t) -2;
for (auto g: glyphs)
{
if (last + 1 != g)
num_ranges++;
last = g;
}
if (unlikely (!rangeRecord.serialize (c, num_ranges))) return_trace (false);
if (!num_ranges) return_trace (true);
unsigned count = 0;
unsigned range = (unsigned) -1;
last = (hb_codepoint_t) -2;
unsigned unsorted = false;
for (auto g: glyphs)
{
if (last + 1 != g)
{
if (unlikely (last != (hb_codepoint_t) -2 && last + 1 > g))
unsorted = true;
range++;
rangeRecord.arrayZ[range].first = g;
rangeRecord.arrayZ[range].value = count;
}
rangeRecord.arrayZ[range].last = g;
last = g;
count++;
}
if (unlikely (unsorted))
rangeRecord.as_array ().qsort (RangeRecord<Types>::cmp_range);
return_trace (true);
}
bool intersects (const hb_set_t *glyphs) const
{
if (rangeRecord.len > glyphs->get_population () * hb_bit_storage ((unsigned) rangeRecord.len))
{
for (auto g : *glyphs)
if (get_coverage (g) != NOT_COVERED)
return true;
return false;
}
return hb_any (+ hb_iter (rangeRecord)
| hb_map ([glyphs] (const RangeRecord<Types> &range) { return range.intersects (*glyphs); }));
}
bool intersects_coverage (const hb_set_t *glyphs, unsigned int index) const
{
auto *range = rangeRecord.as_array ().bsearch (index);
if (range)
return range->intersects (*glyphs);
return false;
}
template <typename IterableOut,
hb_requires (hb_is_sink_of (IterableOut, hb_codepoint_t))>
void intersect_set (const hb_set_t &glyphs, IterableOut&& intersect_glyphs) const
{
/* Break out of loop for overlapping, broken, tables,
* to avoid fuzzer timouts. */
hb_codepoint_t last = 0;
for (const auto& range : rangeRecord)
{
if (unlikely (range.first < last))
break;
last = range.last;
for (hb_codepoint_t g = range.first - 1;
glyphs.next (&g) && g <= last;)
intersect_glyphs << g;
}
}
unsigned cost () const { return hb_bit_storage ((unsigned) rangeRecord.len); /* bsearch cost */ }
template <typename set_t>
bool collect_coverage (set_t *glyphs) const
{
for (const auto& range: rangeRecord)
if (unlikely (!range.collect_coverage (glyphs)))
return false;
return true;
}
public:
/* Older compilers need this to be public. */
struct iter_t
{
void init (const CoverageFormat2_4 &c_)
{
c = &c_;
coverage = 0;
i = 0;
j = c->rangeRecord.len ? c->rangeRecord[0].first : 0;
if (unlikely (c->rangeRecord[0].first > c->rangeRecord[0].last))
{
/* Broken table. Skip. */
i = c->rangeRecord.len;
j = 0;
}
}
bool __more__ () const { return i < c->rangeRecord.len; }
void __next__ ()
{
if (j >= c->rangeRecord[i].last)
{
i++;
if (__more__ ())
{
unsigned int old = coverage;
j = c->rangeRecord.arrayZ[i].first;
coverage = c->rangeRecord.arrayZ[i].value;
if (unlikely (coverage != old + 1))
{
/* Broken table. Skip. Important to avoid DoS.
* Also, our callers depend on coverage being
* consecutive and monotonically increasing,
* ie. iota(). */
i = c->rangeRecord.len;
j = 0;
return;
}
}
else
j = 0;
return;
}
coverage++;
j++;
}
hb_codepoint_t get_glyph () const { return j; }
bool operator != (const iter_t& o) const
{ return i != o.i || j != o.j; }
iter_t __end__ () const
{
iter_t it;
it.init (*c);
it.i = c->rangeRecord.len;
it.j = 0;
return it;
}
private:
const struct CoverageFormat2_4 *c;
unsigned int i, coverage;
hb_codepoint_t j;
};
private:
};
}
}
}
#endif // #ifndef OT_LAYOUT_COMMON_COVERAGEFORMAT2_HH
-97
View File
@@ -1,97 +0,0 @@
/*
* Copyright © 2007,2008,2009 Red Hat, Inc.
* Copyright © 2010,2012 Google, Inc.
*
* This is part of HarfBuzz, a text shaping library.
*
* Permission is hereby granted, without written agreement and without
* license or royalty fees, to use, copy, modify, and distribute this
* software and its documentation for any purpose, provided that the
* above copyright notice and the following two paragraphs appear in
* all copies of this software.
*
* IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
* DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
* ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
* IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
* DAMAGE.
*
* THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
* BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
* FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
* ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
* PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
*
* Red Hat Author(s): Behdad Esfahbod
* Google Author(s): Behdad Esfahbod, Garret Rieger
*/
#ifndef OT_LAYOUT_COMMON_RANGERECORD_HH
#define OT_LAYOUT_COMMON_RANGERECORD_HH
namespace OT {
namespace Layout {
namespace Common {
template <typename Types>
struct RangeRecord
{
typename Types::HBGlyphID first; /* First GlyphID in the range */
typename Types::HBGlyphID last; /* Last GlyphID in the range */
HBUINT16 value; /* Value */
DEFINE_SIZE_STATIC (2 + 2 * Types::size);
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (c->check_struct (this));
}
int cmp (hb_codepoint_t g) const
{ return g < first ? -1 : g <= last ? 0 : +1; }
HB_INTERNAL static int cmp_range (const void *pa, const void *pb) {
const RangeRecord *a = (const RangeRecord *) pa;
const RangeRecord *b = (const RangeRecord *) pb;
if (a->first < b->first) return -1;
if (a->first > b->first) return +1;
if (a->last < b->last) return -1;
if (a->last > b->last) return +1;
if (a->value < b->value) return -1;
if (a->value > b->value) return +1;
return 0;
}
unsigned get_population () const
{
if (unlikely (last < first)) return 0;
return (last - first + 1);
}
bool intersects (const hb_set_t &glyphs) const
{ return glyphs.intersects (first, last); }
template <typename set_t>
bool collect_coverage (set_t *glyphs) const
{ return glyphs->add_range (first, last); }
};
}
}
}
// TODO(garretrieger): This was previously implemented using
// DECLARE_NULL_NAMESPACE_BYTES_TEMPLATE1 (OT, RangeRecord, 9);
// but that only works when there is only a single namespace level.
// The macro should probably be fixed so it can work in this situation.
extern HB_INTERNAL const unsigned char _hb_Null_OT_RangeRecord[9];
template <typename Spec>
struct Null<OT::Layout::Common::RangeRecord<Spec>> {
static OT::Layout::Common::RangeRecord<Spec> const & get_null () {
return *reinterpret_cast<const OT::Layout::Common::RangeRecord<Spec> *> (_hb_Null_OT_RangeRecord);
}
};
#endif // #ifndef OT_LAYOUT_COMMON_RANGERECORD_HH
File diff suppressed because it is too large Load Diff
-84
View File
@@ -1,84 +0,0 @@
#ifndef OT_LAYOUT_GPOS_ANCHOR_HH
#define OT_LAYOUT_GPOS_ANCHOR_HH
#include "AnchorFormat1.hh"
#include "AnchorFormat2.hh"
#include "AnchorFormat3.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct Anchor
{
protected:
union {
struct { HBUINT16 v; } format; /* Format identifier */
AnchorFormat1 format1;
AnchorFormat2 format2;
AnchorFormat3 format3;
} u;
public:
DEFINE_SIZE_UNION (2, format.v);
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
if (!u.format.v.sanitize (c)) return_trace (false);
hb_barrier ();
switch (u.format.v) {
case 1: return_trace (u.format1.sanitize (c));
case 2: return_trace (u.format2.sanitize (c));
case 3: return_trace (u.format3.sanitize (c));
default:return_trace (true);
}
}
void get_anchor (hb_ot_apply_context_t *c, hb_codepoint_t glyph_id,
float *x, float *y) const
{
*x = *y = 0;
switch (u.format.v) {
case 1: u.format1.get_anchor (c, glyph_id, x, y); return;
case 2: u.format2.get_anchor (c, glyph_id, x, y); return;
case 3: u.format3.get_anchor (c, glyph_id, x, y); return;
default: return;
}
}
bool subset (hb_subset_context_t *c) const
{
TRACE_SUBSET (this);
switch (u.format.v) {
case 1: return_trace (bool (reinterpret_cast<Anchor *> (u.format1.copy (c->serializer))));
case 2:
if (c->plan->flags & HB_SUBSET_FLAGS_NO_HINTING)
{
// AnchorFormat 2 just containins extra hinting information, so
// if hints are being dropped convert to format 1.
return_trace (bool (reinterpret_cast<Anchor *> (u.format1.copy (c->serializer))));
}
return_trace (bool (reinterpret_cast<Anchor *> (u.format2.copy (c->serializer))));
case 3: return_trace (u.format3.subset (c));
default:return_trace (false);
}
}
void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
{
switch (u.format.v) {
case 1: case 2:
return;
case 3:
u.format3.collect_variation_indices (c);
return;
default: return;
}
}
};
}
}
}
#endif // OT_LAYOUT_GPOS_ANCHOR_HH
-46
View File
@@ -1,46 +0,0 @@
#ifndef OT_LAYOUT_GPOS_ANCHORFORMAT1_HH
#define OT_LAYOUT_GPOS_ANCHORFORMAT1_HH
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct AnchorFormat1
{
protected:
HBUINT16 format; /* Format identifier--format = 1 */
FWORD xCoordinate; /* Horizontal value--in design units */
FWORD yCoordinate; /* Vertical value--in design units */
public:
DEFINE_SIZE_STATIC (6);
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (c->check_struct (this));
}
void get_anchor (hb_ot_apply_context_t *c, hb_codepoint_t glyph_id HB_UNUSED,
float *x, float *y) const
{
hb_font_t *font = c->font;
*x = font->em_fscale_x (xCoordinate);
*y = font->em_fscale_y (yCoordinate);
}
AnchorFormat1* copy (hb_serialize_context_t *c) const
{
TRACE_SERIALIZE (this);
AnchorFormat1* out = c->embed<AnchorFormat1> (this);
if (!out) return_trace (out);
out->format = 1;
return_trace (out);
}
};
}
}
}
#endif // OT_LAYOUT_GPOS_ANCHORFORMAT1_HH
-58
View File
@@ -1,58 +0,0 @@
#ifndef OT_LAYOUT_GPOS_ANCHORFORMAT2_HH
#define OT_LAYOUT_GPOS_ANCHORFORMAT2_HH
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct AnchorFormat2
{
protected:
HBUINT16 format; /* Format identifier--format = 2 */
FWORD xCoordinate; /* Horizontal value--in design units */
FWORD yCoordinate; /* Vertical value--in design units */
HBUINT16 anchorPoint; /* Index to glyph contour point */
public:
DEFINE_SIZE_STATIC (8);
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (c->check_struct (this));
}
void get_anchor (hb_ot_apply_context_t *c, hb_codepoint_t glyph_id,
float *x, float *y) const
{
hb_font_t *font = c->font;
#ifdef HB_NO_HINTING
*x = font->em_fscale_x (xCoordinate);
*y = font->em_fscale_y (yCoordinate);
return;
#endif
unsigned int x_ppem = font->x_ppem;
unsigned int y_ppem = font->y_ppem;
hb_position_t cx = 0, cy = 0;
bool ret;
ret = (x_ppem || y_ppem) &&
font->get_glyph_contour_point_for_origin (glyph_id, anchorPoint, HB_DIRECTION_LTR, &cx, &cy);
*x = ret && x_ppem ? cx : font->em_fscale_x (xCoordinate);
*y = ret && y_ppem ? cy : font->em_fscale_y (yCoordinate);
}
AnchorFormat2* copy (hb_serialize_context_t *c) const
{
TRACE_SERIALIZE (this);
return_trace (c->embed<AnchorFormat2> (this));
}
};
}
}
}
#endif // OT_LAYOUT_GPOS_ANCHORFORMAT2_HH
-123
View File
@@ -1,123 +0,0 @@
#ifndef OT_LAYOUT_GPOS_ANCHORFORMAT3_HH
#define OT_LAYOUT_GPOS_ANCHORFORMAT3_HH
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct AnchorFormat3
{
protected:
HBUINT16 format; /* Format identifier--format = 3 */
FWORD xCoordinate; /* Horizontal value--in design units */
FWORD yCoordinate; /* Vertical value--in design units */
Offset16To<Device>
xDeviceTable; /* Offset to Device table for X
* coordinate-- from beginning of
* Anchor table (may be NULL) */
Offset16To<Device>
yDeviceTable; /* Offset to Device table for Y
* coordinate-- from beginning of
* Anchor table (may be NULL) */
public:
DEFINE_SIZE_STATIC (10);
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
if (unlikely (!c->check_struct (this))) return_trace (false);
return_trace (xDeviceTable.sanitize (c, this) && yDeviceTable.sanitize (c, this));
}
void get_anchor (hb_ot_apply_context_t *c, hb_codepoint_t glyph_id HB_UNUSED,
float *x, float *y) const
{
hb_font_t *font = c->font;
*x = font->em_fscale_x (xCoordinate);
*y = font->em_fscale_y (yCoordinate);
if ((font->x_ppem || font->has_nonzero_coords) && xDeviceTable.sanitize (&c->sanitizer, this))
{
hb_barrier ();
*x += (this+xDeviceTable).get_x_delta (font, c->var_store, c->var_store_cache);
}
if ((font->y_ppem || font->has_nonzero_coords) && yDeviceTable.sanitize (&c->sanitizer, this))
{
hb_barrier ();
*y += (this+yDeviceTable).get_y_delta (font, c->var_store, c->var_store_cache);
}
}
bool subset (hb_subset_context_t *c) const
{
TRACE_SUBSET (this);
auto *out = c->serializer->start_embed (*this);
if (unlikely (!c->serializer->embed (format))) return_trace (false);
if (unlikely (!c->serializer->embed (xCoordinate))) return_trace (false);
if (unlikely (!c->serializer->embed (yCoordinate))) return_trace (false);
unsigned x_varidx = xDeviceTable ? (this+xDeviceTable).get_variation_index () : HB_OT_LAYOUT_NO_VARIATIONS_INDEX;
if (x_varidx != HB_OT_LAYOUT_NO_VARIATIONS_INDEX)
{
hb_pair_t<unsigned, int> *new_varidx_delta;
if (!c->plan->layout_variation_idx_delta_map.has (x_varidx, &new_varidx_delta))
return_trace (false);
x_varidx = hb_first (*new_varidx_delta);
int delta = hb_second (*new_varidx_delta);
if (delta != 0)
{
if (!c->serializer->check_assign (out->xCoordinate, xCoordinate + delta,
HB_SERIALIZE_ERROR_INT_OVERFLOW))
return_trace (false);
}
}
unsigned y_varidx = yDeviceTable ? (this+yDeviceTable).get_variation_index () : HB_OT_LAYOUT_NO_VARIATIONS_INDEX;
if (y_varidx != HB_OT_LAYOUT_NO_VARIATIONS_INDEX)
{
hb_pair_t<unsigned, int> *new_varidx_delta;
if (!c->plan->layout_variation_idx_delta_map.has (y_varidx, &new_varidx_delta))
return_trace (false);
y_varidx = hb_first (*new_varidx_delta);
int delta = hb_second (*new_varidx_delta);
if (delta != 0)
{
if (!c->serializer->check_assign (out->yCoordinate, yCoordinate + delta,
HB_SERIALIZE_ERROR_INT_OVERFLOW))
return_trace (false);
}
}
bool no_downgrade = (!xDeviceTable.is_null () && !(this+xDeviceTable).is_variation_device ()) ||
x_varidx != HB_OT_LAYOUT_NO_VARIATIONS_INDEX ||
y_varidx != HB_OT_LAYOUT_NO_VARIATIONS_INDEX ||
(!yDeviceTable.is_null () && !(this+yDeviceTable).is_variation_device ());
if (!no_downgrade)
return_trace (c->serializer->check_assign (out->format, 1, HB_SERIALIZE_ERROR_INT_OVERFLOW));
if (!c->serializer->embed (xDeviceTable)) return_trace (false);
if (!c->serializer->embed (yDeviceTable)) return_trace (false);
out->xDeviceTable.serialize_copy (c->serializer, xDeviceTable, this, 0, hb_serialize_context_t::Head, &c->plan->layout_variation_idx_delta_map);
out->yDeviceTable.serialize_copy (c->serializer, yDeviceTable, this, 0, hb_serialize_context_t::Head, &c->plan->layout_variation_idx_delta_map);
return_trace (out);
}
void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
{
(this+xDeviceTable).collect_variation_indices (c);
(this+yDeviceTable).collect_variation_indices (c);
}
};
}
}
}
#endif // OT_LAYOUT_GPOS_ANCHORFORMAT3_HH
-94
View File
@@ -1,94 +0,0 @@
#ifndef OT_LAYOUT_GPOS_ANCHORMATRIX_HH
#define OT_LAYOUT_GPOS_ANCHORMATRIX_HH
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct AnchorMatrix
{
HBUINT16 rows; /* Number of rows */
UnsizedArrayOf<Offset16To<Anchor, AnchorMatrix>>
matrixZ; /* Matrix of offsets to Anchor tables--
* from beginning of AnchorMatrix table */
public:
DEFINE_SIZE_ARRAY (2, matrixZ);
bool sanitize (hb_sanitize_context_t *c, unsigned int cols) const
{
TRACE_SANITIZE (this);
if (!c->check_struct (this)) return_trace (false);
hb_barrier ();
if (unlikely (hb_unsigned_mul_overflows (rows, cols))) return_trace (false);
unsigned int count = rows * cols;
if (!c->check_array (matrixZ.arrayZ, count)) return_trace (false);
if (c->lazy_some_gpos)
return_trace (true);
hb_barrier ();
for (unsigned int i = 0; i < count; i++)
if (!matrixZ[i].sanitize (c, this)) return_trace (false);
return_trace (true);
}
const Anchor& get_anchor (hb_ot_apply_context_t *c,
unsigned int row, unsigned int col,
unsigned int cols, bool *found) const
{
*found = false;
if (unlikely (row >= rows || col >= cols)) return Null (Anchor);
auto &offset = matrixZ[row * cols + col];
if (unlikely (!offset.sanitize (&c->sanitizer, this))) return Null (Anchor);
hb_barrier ();
*found = !offset.is_null ();
return this+offset;
}
template <typename Iterator,
hb_requires (hb_is_iterator (Iterator))>
void collect_variation_indices (hb_collect_variation_indices_context_t *c,
Iterator index_iter) const
{
for (unsigned i : index_iter)
(this+matrixZ[i]).collect_variation_indices (c);
}
template <typename Iterator,
hb_requires (hb_is_iterator (Iterator))>
bool subset (hb_subset_context_t *c,
unsigned num_rows,
Iterator index_iter) const
{
TRACE_SUBSET (this);
auto *out = c->serializer->start_embed (this);
if (!index_iter) return_trace (false);
if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
out->rows = num_rows;
for (const unsigned i : index_iter)
{
auto *offset = c->serializer->embed (matrixZ[i]);
if (!offset) return_trace (false);
offset->serialize_subset (c, matrixZ[i], this);
}
return_trace (true);
}
bool offset_is_null (unsigned row, unsigned col, unsigned num_cols) const
{
if (unlikely (row >= rows || col >= num_cols)) return true;
auto &offset = matrixZ[row * num_cols + col];
return offset.is_null ();
}
};
}
}
}
#endif /* OT_LAYOUT_GPOS_ANCHORMATRIX_HH */
-14
View File
@@ -1,14 +0,0 @@
#ifndef OT_LAYOUT_GPOS_CHAINCONTEXTPOS_HH
#define OT_LAYOUT_GPOS_CHAINCONTEXTPOS_HH
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct ChainContextPos : ChainContext {};
}
}
}
#endif /* OT_LAYOUT_GPOS_CHAINCONTEXTPOS_HH */
-33
View File
@@ -1,33 +0,0 @@
#ifndef OT_LAYOUT_GPOS_COMMON_HH
#define OT_LAYOUT_GPOS_COMMON_HH
namespace OT {
namespace Layout {
namespace GPOS_impl {
enum attach_type_t {
ATTACH_TYPE_NONE = 0X00,
/* Each attachment should be either a mark or a cursive; can't be both. */
ATTACH_TYPE_MARK = 0X01,
ATTACH_TYPE_CURSIVE = 0X02,
};
/* buffer **position** var allocations */
#define attach_chain() var.i16[0] /* glyph to which this attaches to, relative to current glyphs; negative for going back, positive for forward. */
#define attach_type() var.u8[2] /* attachment type */
/* Note! if attach_chain() is zero, the value of attach_type() is irrelevant. */
template<typename Iterator, typename SrcLookup>
static void SinglePos_serialize (hb_serialize_context_t *c,
const SrcLookup *src,
Iterator it,
const hb_hashmap_t<unsigned, hb_pair_t<unsigned, int>> *layout_variation_idx_delta_map,
unsigned new_format);
}
}
}
#endif // OT_LAYOUT_GPOS_COMMON_HH
-14
View File
@@ -1,14 +0,0 @@
#ifndef OT_LAYOUT_GPOS_CONTEXTPOS_HH
#define OT_LAYOUT_GPOS_CONTEXTPOS_HH
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct ContextPos : Context {};
}
}
}
#endif /* OT_LAYOUT_GPOS_CONTEXTPOS_HH */
-35
View File
@@ -1,35 +0,0 @@
#ifndef OT_LAYOUT_GPOS_CURSIVEPOS_HH
#define OT_LAYOUT_GPOS_CURSIVEPOS_HH
#include "CursivePosFormat1.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct CursivePos
{
protected:
union {
struct { HBUINT16 v; } format; /* Format identifier */
CursivePosFormat1 format1;
} u;
public:
template <typename context_t, typename ...Ts>
typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
{
if (unlikely (!c->may_dispatch (this, &u.format.v))) return c->no_dispatch_return_value ();
TRACE_DISPATCH (this, u.format.v);
switch (u.format.v) {
case 1: return_trace (c->dispatch (u.format1, std::forward<Ts> (ds)...));
default:return_trace (c->default_return_value ());
}
}
};
}
}
}
#endif /* OT_LAYOUT_GPOS_CURSIVEPOS_HH */
@@ -1,332 +0,0 @@
#ifndef OT_LAYOUT_GPOS_CURSIVEPOSFORMAT1_HH
#define OT_LAYOUT_GPOS_CURSIVEPOSFORMAT1_HH
#include "Anchor.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct EntryExitRecord
{
friend struct CursivePosFormat1;
bool sanitize (hb_sanitize_context_t *c, const struct CursivePosFormat1 *base) const
{
TRACE_SANITIZE (this);
return_trace (entryAnchor.sanitize (c, base) && exitAnchor.sanitize (c, base));
}
void collect_variation_indices (hb_collect_variation_indices_context_t *c,
const struct CursivePosFormat1 *src_base) const
{
(src_base+entryAnchor).collect_variation_indices (c);
(src_base+exitAnchor).collect_variation_indices (c);
}
bool subset (hb_subset_context_t *c,
const struct CursivePosFormat1 *src_base) const
{
TRACE_SERIALIZE (this);
auto *out = c->serializer->embed (this);
if (unlikely (!out)) return_trace (false);
bool ret = false;
ret |= out->entryAnchor.serialize_subset (c, entryAnchor, src_base);
ret |= out->exitAnchor.serialize_subset (c, exitAnchor, src_base);
return_trace (ret);
}
protected:
Offset16To<Anchor, struct CursivePosFormat1>
entryAnchor; /* Offset to EntryAnchor table--from
* beginning of CursivePos
* subtable--may be NULL */
Offset16To<Anchor, struct CursivePosFormat1>
exitAnchor; /* Offset to ExitAnchor table--from
* beginning of CursivePos
* subtable--may be NULL */
public:
DEFINE_SIZE_STATIC (4);
};
static inline void
reverse_cursive_minor_offset (hb_glyph_position_t *pos,
unsigned int len,
unsigned int i,
hb_direction_t direction,
unsigned int new_parent)
{
int chain = pos[i].attach_chain(), type = pos[i].attach_type();
if (likely (!chain || 0 == (type & ATTACH_TYPE_CURSIVE)))
return;
pos[i].attach_chain() = 0;
unsigned int j = (int) i + chain;
if (unlikely (j >= len))
return;
/* Stop if we see new parent in the chain. */
if (j == new_parent)
return;
int16_t reversed_chain = -chain;
/* The old edge was cleared above; if the reversed distance truncates,
* keep it detached instead of storing a poisoned chain.
*/
if (unlikely (reversed_chain != -chain))
return;
reverse_cursive_minor_offset (pos, len, j, direction, new_parent);
if (HB_DIRECTION_IS_HORIZONTAL (direction))
pos[j].y_offset = -pos[i].y_offset;
else
pos[j].x_offset = -pos[i].x_offset;
pos[j].attach_chain() = reversed_chain;
pos[j].attach_type() = type;
}
struct CursivePosFormat1
{
protected:
HBUINT16 format; /* Format identifier--format = 1 */
Offset16To<Coverage>
coverage; /* Offset to Coverage table--from
* beginning of subtable */
Array16Of<EntryExitRecord>
entryExitRecord; /* Array of EntryExit records--in
* Coverage Index order */
public:
DEFINE_SIZE_ARRAY (6, entryExitRecord);
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
if (unlikely (!coverage.sanitize (c, this)))
return_trace (false);
if (c->lazy_some_gpos)
return_trace (entryExitRecord.sanitize_shallow (c));
else
return_trace (entryExitRecord.sanitize (c, this));
}
bool intersects (const hb_set_t *glyphs) const
{ return (this+coverage).intersects (glyphs); }
void closure_lookups (hb_closure_lookups_context_t *c) const {}
void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
{
+ hb_zip (this+coverage, entryExitRecord)
| hb_filter (c->glyph_set, hb_first)
| hb_map (hb_second)
| hb_apply ([&] (const EntryExitRecord& record) { record.collect_variation_indices (c, this); })
;
}
void collect_glyphs (hb_collect_glyphs_context_t *c) const
{ if (unlikely (!(this+coverage).collect_coverage (c->input))) return; }
const Coverage &get_coverage () const { return this+coverage; }
bool apply (hb_ot_apply_context_t *c) const
{
TRACE_APPLY (this);
hb_buffer_t *buffer = c->buffer;
const EntryExitRecord &this_record = entryExitRecord[(this+coverage).get_coverage (buffer->cur().codepoint)];
if (!this_record.entryAnchor ||
unlikely (!this_record.entryAnchor.sanitize (&c->sanitizer, this))) return_trace (false);
hb_barrier ();
auto &skippy_iter = c->iter_input;
skippy_iter.reset_fast (buffer->idx);
unsigned unsafe_from;
if (unlikely (!skippy_iter.prev (&unsafe_from)))
{
buffer->unsafe_to_concat_from_outbuffer (unsafe_from, buffer->idx + 1);
return_trace (false);
}
const EntryExitRecord &prev_record = entryExitRecord[(this+coverage).get_coverage (buffer->info[skippy_iter.idx].codepoint)];
if (!prev_record.exitAnchor ||
unlikely (!prev_record.exitAnchor.sanitize (&c->sanitizer, this)))
{
buffer->unsafe_to_concat_from_outbuffer (skippy_iter.idx, buffer->idx + 1);
return_trace (false);
}
hb_barrier ();
unsigned int i = skippy_iter.idx;
unsigned int j = buffer->idx;
if (HB_BUFFER_MESSAGE_MORE && c->buffer->messaging ())
{
c->buffer->message (c->font,
"cursive attaching glyph at %u to glyph at %u",
i, j);
}
buffer->unsafe_to_break (i, j + 1);
float entry_x, entry_y, exit_x, exit_y;
(this+prev_record.exitAnchor).get_anchor (c, buffer->info[i].codepoint, &exit_x, &exit_y);
(this+this_record.entryAnchor).get_anchor (c, buffer->info[j].codepoint, &entry_x, &entry_y);
hb_glyph_position_t *pos = buffer->pos;
hb_position_t d;
/* Main-direction adjustment */
switch (c->direction) {
case HB_DIRECTION_LTR:
pos[i].x_advance = roundf (exit_x) + pos[i].x_offset;
d = roundf (entry_x) + pos[j].x_offset;
pos[j].x_advance -= d;
pos[j].x_offset -= d;
break;
case HB_DIRECTION_RTL:
d = roundf (exit_x) + pos[i].x_offset;
pos[i].x_advance -= d;
pos[i].x_offset -= d;
pos[j].x_advance = roundf (entry_x) + pos[j].x_offset;
break;
case HB_DIRECTION_TTB:
pos[i].y_advance = roundf (exit_y) + pos[i].y_offset;
d = roundf (entry_y) + pos[j].y_offset;
pos[j].y_advance -= d;
pos[j].y_offset -= d;
break;
case HB_DIRECTION_BTT:
d = roundf (exit_y) + pos[i].y_offset;
pos[i].y_advance -= d;
pos[i].y_offset -= d;
pos[j].y_advance = roundf (entry_y);
break;
case HB_DIRECTION_INVALID:
default:
break;
}
/* Cross-direction adjustment */
/* We attach child to parent (think graph theory and rooted trees whereas
* the root stays on baseline and each node aligns itself against its
* parent.
*
* Optimize things for the case of RightToLeft, as that's most common in
* Arabic. */
unsigned int child = i;
unsigned int parent = j;
hb_position_t x_offset = roundf (entry_x - exit_x);
hb_position_t y_offset = roundf (entry_y - exit_y);
if (!(c->lookup_props & LookupFlag::RightToLeft))
{
unsigned int k = child;
child = parent;
parent = k;
x_offset = -x_offset;
y_offset = -y_offset;
}
/* If child was already connected to someone else, walk through its old
* chain and reverse the link direction, such that the whole tree of its
* previous connection now attaches to new parent. Watch out for case
* where new parent is on the path from old chain...
*/
reverse_cursive_minor_offset (pos, buffer->len, child, c->direction, parent);
pos[child].attach_chain() = (int) parent - (int) child;
if (pos[child].attach_chain() != (int) parent - (int) child)
{
pos[child].attach_chain() = 0;
goto overflow;
}
pos[child].attach_type() = ATTACH_TYPE_CURSIVE;
buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_GPOS_ATTACHMENT;
if (likely (HB_DIRECTION_IS_HORIZONTAL (c->direction)))
pos[child].y_offset = y_offset;
else
pos[child].x_offset = x_offset;
/* If parent was attached to child, separate them.
* https://github.com/harfbuzz/harfbuzz/issues/2469
*/
if (unlikely (pos[parent].attach_chain() == -pos[child].attach_chain()))
{
pos[parent].attach_chain() = 0;
if (likely (HB_DIRECTION_IS_HORIZONTAL (c->direction)))
pos[parent].y_offset = 0;
else
pos[parent].x_offset = 0;
}
if (HB_BUFFER_MESSAGE_MORE && c->buffer->messaging ())
{
c->buffer->message (c->font,
"cursive attached glyph at %u to glyph at %u",
i, j);
}
overflow:
buffer->idx++;
return_trace (true);
}
template <typename Iterator,
hb_requires (hb_is_iterator (Iterator))>
void serialize (hb_subset_context_t *c,
Iterator it,
const struct CursivePosFormat1 *src_base)
{
if (unlikely (!c->serializer->extend_min ((*this)))) return;
this->format = 1;
this->entryExitRecord.len = it.len ();
for (const EntryExitRecord& entry_record : + it
| hb_map (hb_second))
entry_record.subset (c, src_base);
auto glyphs =
+ it
| hb_map_retains_sorting (hb_first)
;
coverage.serialize_serialize (c->serializer, glyphs);
}
bool subset (hb_subset_context_t *c) const
{
TRACE_SUBSET (this);
const hb_set_t &glyphset = *c->plan->glyphset_gsub ();
const hb_map_t &glyph_map = *c->plan->glyph_map;
auto *out = c->serializer->start_embed (*this);
auto it =
+ hb_zip (this+coverage, entryExitRecord)
| hb_filter (glyphset, hb_first)
| hb_map_retains_sorting ([&] (hb_pair_t<hb_codepoint_t, const EntryExitRecord&> p) -> hb_pair_t<hb_codepoint_t, const EntryExitRecord&>
{ return hb_pair (glyph_map[p.first], p.second);})
;
bool ret = bool (it);
out->serialize (c, it, this);
return_trace (ret);
}
};
}
}
}
#endif /* OT_LAYOUT_GPOS_CURSIVEPOSFORMAT1_HH */
-17
View File
@@ -1,17 +0,0 @@
#ifndef OT_LAYOUT_GPOS_EXTENSIONPOS_HH
#define OT_LAYOUT_GPOS_EXTENSIONPOS_HH
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct ExtensionPos : Extension<ExtensionPos>
{
typedef struct PosLookupSubTable SubTable;
};
}
}
}
#endif /* OT_LAYOUT_GPOS_EXTENSIONPOS_HH */
-206
View File
@@ -1,206 +0,0 @@
#ifndef OT_LAYOUT_GPOS_GPOS_HH
#define OT_LAYOUT_GPOS_GPOS_HH
#include "../../../hb-ot-layout-common.hh"
#include "../../../hb-ot-layout-gsubgpos.hh"
#include "Common.hh"
#include "PosLookup.hh"
namespace OT {
using Layout::GPOS_impl::PosLookup;
namespace Layout {
static void
propagate_attachment_offsets (hb_glyph_position_t *pos,
unsigned int len,
unsigned int i,
hb_direction_t direction,
unsigned nesting_level = HB_MAX_NESTING_LEVEL);
/*
* GPOS -- Glyph Positioning
* https://docs.microsoft.com/en-us/typography/opentype/spec/gpos
*/
struct GPOS : GSUBGPOS
{
static constexpr hb_tag_t tableTag = HB_OT_TAG_GPOS;
using Lookup = PosLookup;
const PosLookup& get_lookup (unsigned int i) const
{ return static_cast<const PosLookup &> (GSUBGPOS::get_lookup (i)); }
static inline void position_start (hb_font_t *font, hb_buffer_t *buffer);
static inline void position_finish_advances (hb_font_t *font, hb_buffer_t *buffer);
static inline void position_finish_offsets (hb_font_t *font, hb_buffer_t *buffer);
bool subset (hb_subset_context_t *c) const
{
hb_subset_layout_context_t l (c, tableTag);
return GSUBGPOS::subset<PosLookup> (&l);
}
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (GSUBGPOS::sanitize<PosLookup> (c));
}
HB_INTERNAL bool is_blocklisted (hb_blob_t *blob,
hb_face_t *face) const;
void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
{
for (unsigned i = 0; i < GSUBGPOS::get_lookup_count (); i++)
{
if (!c->gpos_lookups->has (i)) continue;
const PosLookup &l = get_lookup (i);
l.dispatch (c);
}
}
void closure_lookups (hb_face_t *face,
const hb_set_t *glyphs,
hb_set_t *lookup_indexes /* IN/OUT */) const
{ GSUBGPOS::closure_lookups<PosLookup> (face, glyphs, lookup_indexes); }
typedef GSUBGPOS::accelerator_t<GPOS> accelerator_t;
};
static void
propagate_attachment_offsets (hb_glyph_position_t *pos,
unsigned int len,
unsigned int i,
hb_direction_t direction,
unsigned nesting_level)
{
/* Adjusts offsets of attached glyphs (both cursive and mark) to accumulate
* offset of glyph they are attached to. */
int chain = pos[i].attach_chain();
int type = pos[i].attach_type();
pos[i].attach_chain() = 0;
unsigned int j = (int) i + chain;
if (unlikely (j >= len))
return;
if (unlikely (!nesting_level))
return;
if (pos[j].attach_chain())
propagate_attachment_offsets (pos, len, j, direction, nesting_level - 1);
assert (!!(type & GPOS_impl::ATTACH_TYPE_MARK) ^ !!(type & GPOS_impl::ATTACH_TYPE_CURSIVE));
if (type & GPOS_impl::ATTACH_TYPE_CURSIVE)
{
if (HB_DIRECTION_IS_HORIZONTAL (direction))
pos[i].y_offset += pos[j].y_offset;
else
pos[i].x_offset += pos[j].x_offset;
}
else /*if (type & GPOS_impl::ATTACH_TYPE_MARK)*/
{
pos[i].x_offset += pos[j].x_offset;
pos[i].y_offset += pos[j].y_offset;
// i is the position of the mark; j is the base.
if (j < i)
{
/* This is the common case: mark follows base.
* And currently the only way in OpenType. */
if (HB_DIRECTION_IS_FORWARD (direction))
for (unsigned int k = j; k < i; k++) {
pos[i].x_offset -= pos[k].x_advance;
pos[i].y_offset -= pos[k].y_advance;
}
else
for (unsigned int k = j + 1; k < i + 1; k++) {
pos[i].x_offset += pos[k].x_advance;
pos[i].y_offset += pos[k].y_advance;
}
}
else // j > i
{
/* This can happen with `kerx`: a mark attaching
* to a base after it in the logical order. */
if (HB_DIRECTION_IS_FORWARD (direction))
for (unsigned int k = i; k < j; k++) {
pos[i].x_offset += pos[k].x_advance;
pos[i].y_offset += pos[k].y_advance;
}
else
for (unsigned int k = i + 1; k < j + 1; k++) {
pos[i].x_offset -= pos[k].x_advance;
pos[i].y_offset -= pos[k].y_advance;
}
}
}
}
void
GPOS::position_start (hb_font_t *font HB_UNUSED, hb_buffer_t *buffer)
{
unsigned int count = buffer->len;
for (unsigned int i = 0; i < count; i++)
buffer->pos[i].attach_chain() = buffer->pos[i].attach_type() = 0;
}
void
GPOS::position_finish_advances (hb_font_t *font HB_UNUSED, hb_buffer_t *buffer HB_UNUSED)
{
//_hb_buffer_assert_gsubgpos_vars (buffer);
}
void
GPOS::position_finish_offsets (hb_font_t *font, hb_buffer_t *buffer)
{
_hb_buffer_assert_gsubgpos_vars (buffer);
unsigned int len;
hb_glyph_position_t *pos = hb_buffer_get_glyph_positions (buffer, &len);
hb_direction_t direction = buffer->props.direction;
/* Handle attachments */
if (buffer->scratch_flags & HB_BUFFER_SCRATCH_FLAG_HAS_GPOS_ATTACHMENT)
{
auto *pos = buffer->pos;
// https://github.com/harfbuzz/harfbuzz/issues/5514
if (HB_DIRECTION_IS_FORWARD (direction))
{
for (unsigned i = 0; i < len; i++)
if (pos[i].attach_chain())
propagate_attachment_offsets (pos, len, i, direction);
} else {
for (unsigned i = len; i-- > 0; )
if (pos[i].attach_chain())
propagate_attachment_offsets (pos, len, i, direction);
}
}
if (unlikely (font->slant_xy) &&
HB_DIRECTION_IS_HORIZONTAL (direction))
{
/* Slanting shaping results is only supported for horizontal text,
* as it gets weird otherwise. */
for (unsigned i = 0; i < len; i++)
if (unlikely (pos[i].y_offset))
pos[i].x_offset += roundf (font->slant_xy * pos[i].y_offset);
}
}
}
struct GPOS_accelerator_t : Layout::GPOS::accelerator_t {
GPOS_accelerator_t (hb_face_t *face) : Layout::GPOS::accelerator_t (face) {}
};
}
#endif /* OT_LAYOUT_GPOS_GPOS_HH */
-68
View File
@@ -1,68 +0,0 @@
#ifndef OT_LAYOUT_GPOS_LIGATUREARRAY_HH
#define OT_LAYOUT_GPOS_LIGATUREARRAY_HH
namespace OT {
namespace Layout {
namespace GPOS_impl {
typedef AnchorMatrix LigatureAttach; /* component-major--
* in order of writing direction--,
* mark-minor--
* ordered by class--zero-based. */
/* Array of LigatureAttach tables ordered by LigatureCoverage Index */
struct LigatureArray : List16OfOffset16To<LigatureAttach>
{
template <typename Iterator,
hb_requires (hb_is_iterator (Iterator))>
bool subset (hb_subset_context_t *c,
Iterator coverage,
unsigned class_count,
const hb_map_t *klass_mapping,
hb_sorted_vector_t<hb_codepoint_t> &new_coverage /* OUT */) const
{
TRACE_SUBSET (this);
const hb_map_t &glyph_map = c->plan->glyph_map_gsub;
auto *out = c->serializer->start_embed (this);
if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
bool ret = false;
for (const auto _ : + hb_zip (coverage, *this)
| hb_filter (glyph_map, hb_first))
{
const LigatureAttach& src = (this + _.second);
bool non_empty = + hb_range (src.rows * class_count)
| hb_filter ([=] (unsigned index) { return klass_mapping->has (index % class_count); })
| hb_map ([&] (const unsigned index) { return !src.offset_is_null (index / class_count, index % class_count, class_count); })
| hb_any;
if (!non_empty) continue;
auto *matrix = out->serialize_append (c->serializer);
if (unlikely (!matrix)) return_trace (false);
auto indexes =
+ hb_range (src.rows * class_count)
| hb_filter ([=] (unsigned index) { return klass_mapping->has (index % class_count); })
;
ret |= matrix->serialize_subset (c,
_.second,
this,
src.rows,
indexes);
hb_codepoint_t new_gid = glyph_map.get (_.first);
new_coverage.push (new_gid);
}
return_trace (ret);
}
};
}
}
}
#endif /* OT_LAYOUT_GPOS_LIGATUREARRAY_HH */
-134
View File
@@ -1,134 +0,0 @@
#ifndef OT_LAYOUT_GPOS_MARKARRAY_HH
#define OT_LAYOUT_GPOS_MARKARRAY_HH
#include "AnchorMatrix.hh"
#include "MarkRecord.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct MarkArray : Array16Of<MarkRecord> /* Array of MarkRecords--in Coverage order */
{
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (Array16Of<MarkRecord>::sanitize (c, this));
}
bool apply (hb_ot_apply_context_t *c,
unsigned int mark_index, unsigned int glyph_index,
const AnchorMatrix &anchors, unsigned int class_count,
unsigned int glyph_pos) const
{
TRACE_APPLY (this);
hb_buffer_t *buffer = c->buffer;
const MarkRecord &record = Array16Of<MarkRecord>::operator[](mark_index);
unsigned int mark_class = record.klass;
const Anchor& mark_anchor = this + record.markAnchor;
bool found;
const Anchor& glyph_anchor = anchors.get_anchor (c, glyph_index, mark_class, class_count, &found);
/* If this subtable doesn't have an anchor for this base and this class,
* return false such that the subsequent subtables have a chance at it. */
if (unlikely (!found)) return_trace (false);
float mark_x, mark_y, base_x, base_y;
buffer->unsafe_to_break (glyph_pos, buffer->idx + 1);
mark_anchor.get_anchor (c, buffer->cur().codepoint, &mark_x, &mark_y);
glyph_anchor.get_anchor (c, buffer->info[glyph_pos].codepoint, &base_x, &base_y);
if (HB_BUFFER_MESSAGE_MORE && c->buffer->messaging ())
{
c->buffer->message (c->font,
"attaching mark glyph at %u to glyph at %u",
c->buffer->idx, glyph_pos);
}
hb_glyph_position_t &o = buffer->cur_pos();
o.attach_chain() = (int) glyph_pos - (int) buffer->idx;
if (o.attach_chain() != (int) glyph_pos - (int) buffer->idx)
{
o.attach_chain() = 0;
goto overflow;
}
o.attach_type() = ATTACH_TYPE_MARK;
o.x_offset = roundf (base_x - mark_x);
o.y_offset = roundf (base_y - mark_y);
buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_GPOS_ATTACHMENT;
if (HB_BUFFER_MESSAGE_MORE && c->buffer->messaging ())
{
c->buffer->message (c->font,
"attached mark glyph at %u to glyph at %u",
c->buffer->idx, glyph_pos);
}
overflow:
buffer->idx++;
return_trace (true);
}
template <typename Iterator,
hb_requires (hb_is_iterator (Iterator))>
bool subset (hb_subset_context_t *c,
Iterator coverage,
const hb_map_t *klass_mapping) const
{
TRACE_SUBSET (this);
const hb_set_t &glyphset = *c->plan->glyphset_gsub ();
auto* out = c->serializer->start_embed (this);
if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
auto mark_iter =
+ hb_zip (coverage, this->iter ())
| hb_filter (glyphset, hb_first)
| hb_map (hb_second)
;
bool ret = false;
unsigned new_length = 0;
for (const auto& mark_record : mark_iter) {
ret |= mark_record.subset (c, this, klass_mapping);
new_length++;
}
if (unlikely (!c->serializer->check_assign (out->len, new_length,
HB_SERIALIZE_ERROR_ARRAY_OVERFLOW)))
return_trace (false);
return_trace (ret);
}
};
HB_INTERNAL inline
void Markclass_closure_and_remap_indexes (const Coverage &mark_coverage,
const MarkArray &mark_array,
const hb_set_t &glyphset,
hb_map_t* klass_mapping /* INOUT */)
{
hb_set_t orig_classes;
+ hb_zip (mark_coverage, mark_array)
| hb_filter (glyphset, hb_first)
| hb_map (hb_second)
| hb_map (&MarkRecord::get_class)
| hb_sink (orig_classes)
;
unsigned idx = 0;
for (auto klass : orig_classes.iter ())
{
if (klass_mapping->has (klass)) continue;
klass_mapping->set (klass, idx);
idx++;
}
}
}
}
}
#endif /* OT_LAYOUT_GPOS_MARKARRAY_HH */
-41
View File
@@ -1,41 +0,0 @@
#ifndef OT_LAYOUT_GPOS_MARKBASEPOS_HH
#define OT_LAYOUT_GPOS_MARKBASEPOS_HH
#include "MarkBasePosFormat1.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct MarkBasePos
{
protected:
union {
struct { HBUINT16 v; } format; /* Format identifier */
MarkBasePosFormat1_2<SmallTypes> format1;
#ifndef HB_NO_BEYOND_64K
MarkBasePosFormat1_2<MediumTypes> format2;
#endif
} u;
public:
template <typename context_t, typename ...Ts>
typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
{
if (unlikely (!c->may_dispatch (this, &u.format.v))) return c->no_dispatch_return_value ();
TRACE_DISPATCH (this, u.format.v);
switch (u.format.v) {
case 1: return_trace (c->dispatch (u.format1, std::forward<Ts> (ds)...));
#ifndef HB_NO_BEYOND_64K
case 2: return_trace (c->dispatch (u.format2, std::forward<Ts> (ds)...));
#endif
default:return_trace (c->default_return_value ());
}
}
};
}
}
}
#endif /* OT_LAYOUT_GPOS_MARKBASEPOS_HH */
@@ -1,250 +0,0 @@
#ifndef OT_LAYOUT_GPOS_MARKBASEPOSFORMAT1_HH
#define OT_LAYOUT_GPOS_MARKBASEPOSFORMAT1_HH
#include "MarkArray.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
typedef AnchorMatrix BaseArray; /* base-major--
* in order of BaseCoverage Index--,
* mark-minor--
* ordered by class--zero-based. */
template <typename Types>
struct MarkBasePosFormat1_2
{
protected:
HBUINT16 format; /* Format identifier--format = 1 */
typename Types::template OffsetTo<Coverage>
markCoverage; /* Offset to MarkCoverage table--from
* beginning of MarkBasePos subtable */
typename Types::template OffsetTo<Coverage>
baseCoverage; /* Offset to BaseCoverage table--from
* beginning of MarkBasePos subtable */
HBUINT16 classCount; /* Number of classes defined for marks */
typename Types::template OffsetTo<MarkArray>
markArray; /* Offset to MarkArray table--from
* beginning of MarkBasePos subtable */
typename Types::template OffsetTo<BaseArray>
baseArray; /* Offset to BaseArray table--from
* beginning of MarkBasePos subtable */
public:
DEFINE_SIZE_STATIC (4 + 4 * Types::size);
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (c->check_struct (this) &&
markCoverage.sanitize (c, this) &&
baseCoverage.sanitize (c, this) &&
markArray.sanitize (c, this) &&
baseArray.sanitize (c, this, (unsigned int) classCount));
}
bool intersects (const hb_set_t *glyphs) const
{
return (this+markCoverage).intersects (glyphs) &&
(this+baseCoverage).intersects (glyphs);
}
void closure_lookups (hb_closure_lookups_context_t *c) const {}
void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
{
+ hb_zip (this+markCoverage, this+markArray)
| hb_filter (c->glyph_set, hb_first)
| hb_map (hb_second)
| hb_apply ([&] (const MarkRecord& record) { record.collect_variation_indices (c, &(this+markArray)); })
;
hb_map_t klass_mapping;
Markclass_closure_and_remap_indexes (this+markCoverage, this+markArray, *c->glyph_set, &klass_mapping);
unsigned basecount = (this+baseArray).rows;
auto base_iter =
+ hb_zip (this+baseCoverage, hb_range (basecount))
| hb_filter (c->glyph_set, hb_first)
| hb_map (hb_second)
;
hb_sorted_vector_t<unsigned> base_indexes;
for (const unsigned row : base_iter)
{
+ hb_range ((unsigned) classCount)
| hb_filter (klass_mapping)
| hb_map ([&] (const unsigned col) { return row * (unsigned) classCount + col; })
| hb_sink (base_indexes)
;
}
(this+baseArray).collect_variation_indices (c, base_indexes.iter ());
}
void collect_glyphs (hb_collect_glyphs_context_t *c) const
{
if (unlikely (!(this+markCoverage).collect_coverage (c->input))) return;
if (unlikely (!(this+baseCoverage).collect_coverage (c->input))) return;
}
const Coverage &get_coverage () const { return this+markCoverage; }
static inline bool accept (hb_buffer_t *buffer, unsigned idx)
{
/* We only want to attach to the first of a MultipleSubst sequence.
* https://github.com/harfbuzz/harfbuzz/issues/740
* Reject others...
* ...but stop if we find a mark in the MultipleSubst sequence:
* https://github.com/harfbuzz/harfbuzz/issues/1020 */
return !_hb_glyph_info_multiplied (&buffer->info[idx]) ||
0 == _hb_glyph_info_get_lig_comp (&buffer->info[idx]) ||
(idx == 0 ||
_hb_glyph_info_is_mark (&buffer->info[idx - 1]) ||
!_hb_glyph_info_multiplied (&buffer->info[idx - 1]) ||
_hb_glyph_info_get_lig_id (&buffer->info[idx]) !=
_hb_glyph_info_get_lig_id (&buffer->info[idx - 1]) ||
_hb_glyph_info_get_lig_comp (&buffer->info[idx]) !=
_hb_glyph_info_get_lig_comp (&buffer->info[idx - 1]) + 1
);
}
bool apply (hb_ot_apply_context_t *c) const
{
TRACE_APPLY (this);
hb_buffer_t *buffer = c->buffer;
unsigned int mark_index = (this+markCoverage).get_coverage (buffer->cur().codepoint);
if (likely (mark_index == NOT_COVERED)) return_trace (false);
/* Now we search backwards for a non-mark glyph.
* We don't use skippy_iter.prev() to avoid O(n^2) behavior. */
auto &skippy_iter = c->iter_input;
skippy_iter.set_lookup_props (LookupFlag::IgnoreMarks);
if (c->last_base_until > buffer->idx)
{
c->last_base_until = 0;
c->last_base = -1;
}
unsigned j;
for (j = buffer->idx; j > c->last_base_until; j--)
{
auto match = skippy_iter.match (buffer->info[j - 1]);
if (match == skippy_iter.MATCH)
{
// https://github.com/harfbuzz/harfbuzz/issues/4124
if (!accept (buffer, j - 1) &&
NOT_COVERED == (this+baseCoverage).get_coverage (buffer->info[j - 1].codepoint))
match = skippy_iter.SKIP;
}
if (match == skippy_iter.MATCH)
{
c->last_base = (signed) j - 1;
break;
}
}
c->last_base_until = buffer->idx;
if (c->last_base == -1)
{
buffer->unsafe_to_concat_from_outbuffer (0, buffer->idx + 1);
return_trace (false);
}
unsigned idx = (unsigned) c->last_base;
/* Checking that matched glyph is actually a base glyph by GDEF is too strong; disabled */
//if (!_hb_glyph_info_is_base_glyph (&buffer->info[idx])) { return_trace (false); }
unsigned int base_index = (this+baseCoverage).get_coverage (buffer->info[idx].codepoint);
if (base_index == NOT_COVERED)
{
buffer->unsafe_to_concat_from_outbuffer (idx, buffer->idx + 1);
return_trace (false);
}
return_trace ((this+markArray).apply (c, mark_index, base_index, this+baseArray, classCount, idx));
}
bool subset (hb_subset_context_t *c) const
{
TRACE_SUBSET (this);
const hb_set_t &glyphset = *c->plan->glyphset_gsub ();
const hb_map_t &glyph_map = *c->plan->glyph_map;
auto *out = c->serializer->start_embed (*this);
if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
out->format = format;
hb_map_t klass_mapping;
Markclass_closure_and_remap_indexes (this+markCoverage, this+markArray, glyphset, &klass_mapping);
if (!klass_mapping.get_population ()) return_trace (false);
out->classCount = klass_mapping.get_population ();
auto mark_iter =
+ hb_zip (this+markCoverage, this+markArray)
| hb_filter (glyphset, hb_first)
;
hb_sorted_vector_t<hb_codepoint_t> new_coverage;
+ mark_iter
| hb_map (hb_first)
| hb_map (glyph_map)
| hb_sink (new_coverage)
;
if (!out->markCoverage.serialize_serialize (c->serializer, new_coverage.iter ()))
return_trace (false);
if (unlikely (!out->markArray.serialize_subset (c, markArray, this,
(this+markCoverage).iter (),
&klass_mapping)))
return_trace (false);
unsigned basecount = (this+baseArray).rows;
auto base_iter =
+ hb_zip (this+baseCoverage, hb_range (basecount))
| hb_filter (glyphset, hb_first)
;
new_coverage.reset ();
hb_sorted_vector_t<unsigned> base_indexes;
auto &base_array = (this+baseArray);
for (const auto _ : + base_iter)
{
unsigned row = _.second;
bool non_empty = + hb_range ((unsigned) classCount)
| hb_filter (klass_mapping)
| hb_map ([&] (const unsigned col) { return !base_array.offset_is_null (row, col, (unsigned) classCount); })
| hb_any
;
if (!non_empty) continue;
hb_codepoint_t new_g = glyph_map.get ( _.first);
new_coverage.push (new_g);
+ hb_range ((unsigned) classCount)
| hb_filter (klass_mapping)
| hb_map ([&] (const unsigned col) { return row * (unsigned) classCount + col; })
| hb_sink (base_indexes)
;
}
if (!new_coverage) return_trace (false);
if (!out->baseCoverage.serialize_serialize (c->serializer, new_coverage.iter ()))
return_trace (false);
return_trace (out->baseArray.serialize_subset (c, baseArray, this,
new_coverage.length,
base_indexes.iter ()));
}
};
}
}
}
#endif /* OT_LAYOUT_GPOS_MARKBASEPOSFORMAT1_HH */
-41
View File
@@ -1,41 +0,0 @@
#ifndef OT_LAYOUT_GPOS_MARKLIGPOS_HH
#define OT_LAYOUT_GPOS_MARKLIGPOS_HH
#include "MarkLigPosFormat1.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct MarkLigPos
{
protected:
union {
struct { HBUINT16 v; } format; /* Format identifier */
MarkLigPosFormat1_2<SmallTypes> format1;
#ifndef HB_NO_BEYOND_64K
MarkLigPosFormat1_2<MediumTypes> format2;
#endif
} u;
public:
template <typename context_t, typename ...Ts>
typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
{
if (unlikely (!c->may_dispatch (this, &u.format.v))) return c->no_dispatch_return_value ();
TRACE_DISPATCH (this, u.format.v);
switch (u.format.v) {
case 1: return_trace (c->dispatch (u.format1, std::forward<Ts> (ds)...));
#ifndef HB_NO_BEYOND_64K
case 2: return_trace (c->dispatch (u.format2, std::forward<Ts> (ds)...));
#endif
default:return_trace (c->default_return_value ());
}
}
};
}
}
}
#endif /* OT_LAYOUT_GPOS_MARKLIGPOS_HH */
@@ -1,218 +0,0 @@
#ifndef OT_LAYOUT_GPOS_MARKLIGPOSFORMAT1_HH
#define OT_LAYOUT_GPOS_MARKLIGPOSFORMAT1_HH
#include "LigatureArray.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
template <typename Types>
struct MarkLigPosFormat1_2
{
protected:
HBUINT16 format; /* Format identifier--format = 1 */
typename Types::template OffsetTo<Coverage>
markCoverage; /* Offset to Mark Coverage table--from
* beginning of MarkLigPos subtable */
typename Types::template OffsetTo<Coverage>
ligatureCoverage; /* Offset to Ligature Coverage
* table--from beginning of MarkLigPos
* subtable */
HBUINT16 classCount; /* Number of defined mark classes */
typename Types::template OffsetTo<MarkArray>
markArray; /* Offset to MarkArray table--from
* beginning of MarkLigPos subtable */
typename Types::template OffsetTo<LigatureArray>
ligatureArray; /* Offset to LigatureArray table--from
* beginning of MarkLigPos subtable */
public:
DEFINE_SIZE_STATIC (4 + 4 * Types::size);
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (c->check_struct (this) &&
markCoverage.sanitize (c, this) &&
ligatureCoverage.sanitize (c, this) &&
markArray.sanitize (c, this) &&
ligatureArray.sanitize (c, this, (unsigned int) classCount));
}
bool intersects (const hb_set_t *glyphs) const
{
return (this+markCoverage).intersects (glyphs) &&
(this+ligatureCoverage).intersects (glyphs);
}
void closure_lookups (hb_closure_lookups_context_t *c) const {}
void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
{
+ hb_zip (this+markCoverage, this+markArray)
| hb_filter (c->glyph_set, hb_first)
| hb_map (hb_second)
| hb_apply ([&] (const MarkRecord& record) { record.collect_variation_indices (c, &(this+markArray)); })
;
hb_map_t klass_mapping;
Markclass_closure_and_remap_indexes (this+markCoverage, this+markArray, *c->glyph_set, &klass_mapping);
unsigned ligcount = (this+ligatureArray).len;
auto lig_iter =
+ hb_zip (this+ligatureCoverage, hb_range (ligcount))
| hb_filter (c->glyph_set, hb_first)
| hb_map (hb_second)
;
const LigatureArray& lig_array = this+ligatureArray;
for (const unsigned i : lig_iter)
{
hb_sorted_vector_t<unsigned> lig_indexes;
unsigned row_count = lig_array[i].rows;
for (unsigned row : + hb_range (row_count))
{
+ hb_range ((unsigned) classCount)
| hb_filter (klass_mapping)
| hb_map ([&] (const unsigned col) { return row * (unsigned) classCount + col; })
| hb_sink (lig_indexes)
;
}
lig_array[i].collect_variation_indices (c, lig_indexes.iter ());
}
}
void collect_glyphs (hb_collect_glyphs_context_t *c) const
{
if (unlikely (!(this+markCoverage).collect_coverage (c->input))) return;
if (unlikely (!(this+ligatureCoverage).collect_coverage (c->input))) return;
}
const Coverage &get_coverage () const { return this+markCoverage; }
bool apply (hb_ot_apply_context_t *c) const
{
TRACE_APPLY (this);
hb_buffer_t *buffer = c->buffer;
unsigned int mark_index = (this+markCoverage).get_coverage (buffer->cur().codepoint);
if (likely (mark_index == NOT_COVERED)) return_trace (false);
/* Now we search backwards for a non-mark glyph */
auto &skippy_iter = c->iter_input;
skippy_iter.set_lookup_props (LookupFlag::IgnoreMarks);
if (c->last_base_until > buffer->idx)
{
c->last_base_until = 0;
c->last_base = -1;
}
unsigned j;
for (j = buffer->idx; j > c->last_base_until; j--)
{
auto match = skippy_iter.match (buffer->info[j - 1]);
if (match == skippy_iter.MATCH)
{
c->last_base = (signed) j - 1;
break;
}
}
c->last_base_until = buffer->idx;
if (c->last_base == -1)
{
buffer->unsafe_to_concat_from_outbuffer (0, buffer->idx + 1);
return_trace (false);
}
unsigned idx = (unsigned) c->last_base;
/* Checking that matched glyph is actually a ligature by GDEF is too strong; disabled */
//if (!_hb_glyph_info_is_ligature (&buffer->info[idx])) { return_trace (false); }
unsigned int lig_index = (this+ligatureCoverage).get_coverage (buffer->info[idx].codepoint);
if (lig_index == NOT_COVERED)
{
buffer->unsafe_to_concat_from_outbuffer (idx, buffer->idx + 1);
return_trace (false);
}
const LigatureArray& lig_array = this+ligatureArray;
const LigatureAttach& lig_attach = lig_array[lig_index];
/* Find component to attach to */
unsigned int comp_count = lig_attach.rows;
if (unlikely (!comp_count))
{
buffer->unsafe_to_concat_from_outbuffer (idx, buffer->idx + 1);
return_trace (false);
}
/* We must now check whether the ligature ID of the current mark glyph
* is identical to the ligature ID of the found ligature. If yes, we
* can directly use the component index. If not, we attach the mark
* glyph to the last component of the ligature. */
unsigned int comp_index;
unsigned int lig_id = _hb_glyph_info_get_lig_id (&buffer->info[idx]);
unsigned int mark_id = _hb_glyph_info_get_lig_id (&buffer->cur());
unsigned int mark_comp = _hb_glyph_info_get_lig_comp (&buffer->cur());
if (lig_id && lig_id == mark_id && mark_comp > 0)
comp_index = hb_min (comp_count, _hb_glyph_info_get_lig_comp (&buffer->cur())) - 1;
else
comp_index = comp_count - 1;
return_trace ((this+markArray).apply (c, mark_index, comp_index, lig_attach, classCount, idx));
}
bool subset (hb_subset_context_t *c) const
{
TRACE_SUBSET (this);
const hb_set_t &glyphset = *c->plan->glyphset_gsub ();
const hb_map_t &glyph_map = c->plan->glyph_map_gsub;
auto *out = c->serializer->start_embed (*this);
if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
out->format = format;
hb_map_t klass_mapping;
Markclass_closure_and_remap_indexes (this+markCoverage, this+markArray, glyphset, &klass_mapping);
if (!klass_mapping.get_population ()) return_trace (false);
out->classCount = klass_mapping.get_population ();
auto mark_iter =
+ hb_zip (this+markCoverage, this+markArray)
| hb_filter (glyphset, hb_first)
;
auto new_mark_coverage =
+ mark_iter
| hb_map_retains_sorting (hb_first)
| hb_map_retains_sorting (glyph_map)
;
if (!out->markCoverage.serialize_serialize (c->serializer, new_mark_coverage))
return_trace (false);
if (unlikely (!out->markArray.serialize_subset (c, markArray, this,
(this+markCoverage).iter (),
&klass_mapping)))
return_trace (false);
hb_sorted_vector_t<hb_codepoint_t> new_lig_coverage;
if (!out->ligatureArray.serialize_subset (c, ligatureArray, this,
hb_iter (this+ligatureCoverage),
classCount, &klass_mapping, new_lig_coverage))
return_trace (false);
return_trace (out->ligatureCoverage.serialize_serialize (c->serializer, new_lig_coverage.iter ()));
}
};
}
}
}
#endif /* OT_LAYOUT_GPOS_MARKLIGPOSFORMAT1_HH */
-42
View File
@@ -1,42 +0,0 @@
#ifndef OT_LAYOUT_GPOS_MARKMARKPOS_HH
#define OT_LAYOUT_GPOS_MARKMARKPOS_HH
#include "MarkMarkPosFormat1.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct MarkMarkPos
{
protected:
union {
struct { HBUINT16 v; } format; /* Format identifier */
MarkMarkPosFormat1_2<SmallTypes> format1;
#ifndef HB_NO_BEYOND_64K
MarkMarkPosFormat1_2<MediumTypes> format2;
#endif
} u;
public:
template <typename context_t, typename ...Ts>
typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
{
if (unlikely (!c->may_dispatch (this, &u.format.v))) return c->no_dispatch_return_value ();
TRACE_DISPATCH (this, u.format.v);
switch (u.format.v) {
case 1: return_trace (c->dispatch (u.format1, std::forward<Ts> (ds)...));
#ifndef HB_NO_BEYOND_64K
case 2: return_trace (c->dispatch (u.format2, std::forward<Ts> (ds)...));
#endif
default:return_trace (c->default_return_value ());
}
}
};
}
}
}
#endif /* OT_LAYOUT_GPOS_MARKMARKPOS_HH */
@@ -1,239 +0,0 @@
#ifndef OT_LAYOUT_GPOS_MARKMARKPOSFORMAT1_HH
#define OT_LAYOUT_GPOS_MARKMARKPOSFORMAT1_HH
#include "MarkMarkPosFormat1.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
typedef AnchorMatrix Mark2Array; /* mark2-major--
* in order of Mark2Coverage Index--,
* mark1-minor--
* ordered by class--zero-based. */
template <typename Types>
struct MarkMarkPosFormat1_2
{
protected:
HBUINT16 format; /* Format identifier--format = 1 */
typename Types::template OffsetTo<Coverage>
mark1Coverage; /* Offset to Combining Mark1 Coverage
* table--from beginning of MarkMarkPos
* subtable */
typename Types::template OffsetTo<Coverage>
mark2Coverage; /* Offset to Combining Mark2 Coverage
* table--from beginning of MarkMarkPos
* subtable */
HBUINT16 classCount; /* Number of defined mark classes */
typename Types::template OffsetTo<MarkArray>
mark1Array; /* Offset to Mark1Array table--from
* beginning of MarkMarkPos subtable */
typename Types::template OffsetTo<Mark2Array>
mark2Array; /* Offset to Mark2Array table--from
* beginning of MarkMarkPos subtable */
public:
DEFINE_SIZE_STATIC (4 + 4 * Types::size);
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (c->check_struct (this) &&
mark1Coverage.sanitize (c, this) &&
mark2Coverage.sanitize (c, this) &&
mark1Array.sanitize (c, this) &&
hb_barrier () &&
mark2Array.sanitize (c, this, (unsigned int) classCount));
}
bool intersects (const hb_set_t *glyphs) const
{
return (this+mark1Coverage).intersects (glyphs) &&
(this+mark2Coverage).intersects (glyphs);
}
void closure_lookups (hb_closure_lookups_context_t *c) const {}
void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
{
+ hb_zip (this+mark1Coverage, this+mark1Array)
| hb_filter (c->glyph_set, hb_first)
| hb_map (hb_second)
| hb_apply ([&] (const MarkRecord& record) { record.collect_variation_indices (c, &(this+mark1Array)); })
;
hb_map_t klass_mapping;
Markclass_closure_and_remap_indexes (this+mark1Coverage, this+mark1Array, *c->glyph_set, &klass_mapping);
unsigned mark2_count = (this+mark2Array).rows;
auto mark2_iter =
+ hb_zip (this+mark2Coverage, hb_range (mark2_count))
| hb_filter (c->glyph_set, hb_first)
| hb_map (hb_second)
;
hb_sorted_vector_t<unsigned> mark2_indexes;
for (const unsigned row : mark2_iter)
{
+ hb_range ((unsigned) classCount)
| hb_filter (klass_mapping)
| hb_map ([&] (const unsigned col) { return row * (unsigned) classCount + col; })
| hb_sink (mark2_indexes)
;
}
(this+mark2Array).collect_variation_indices (c, mark2_indexes.iter ());
}
void collect_glyphs (hb_collect_glyphs_context_t *c) const
{
if (unlikely (!(this+mark1Coverage).collect_coverage (c->input))) return;
if (unlikely (!(this+mark2Coverage).collect_coverage (c->input))) return;
}
const Coverage &get_coverage () const { return this+mark1Coverage; }
bool apply (hb_ot_apply_context_t *c) const
{
TRACE_APPLY (this);
hb_buffer_t *buffer = c->buffer;
unsigned int mark1_index = (this+mark1Coverage).get_coverage (buffer->cur().codepoint);
if (likely (mark1_index == NOT_COVERED)) return_trace (false);
/* now we search backwards for a suitable mark glyph until a non-mark glyph */
auto &skippy_iter = c->iter_input;
skippy_iter.reset_fast (buffer->idx);
skippy_iter.set_lookup_props (c->lookup_props & ~(uint32_t)LookupFlag::IgnoreFlags);
unsigned unsafe_from;
if (unlikely (!skippy_iter.prev (&unsafe_from)))
{
buffer->unsafe_to_concat_from_outbuffer (unsafe_from, buffer->idx + 1);
return_trace (false);
}
if (likely (!_hb_glyph_info_is_mark (&buffer->info[skippy_iter.idx])))
{
buffer->unsafe_to_concat_from_outbuffer (skippy_iter.idx, buffer->idx + 1);
return_trace (false);
}
unsigned int j = skippy_iter.idx;
unsigned int id1 = _hb_glyph_info_get_lig_id (&buffer->cur());
unsigned int id2 = _hb_glyph_info_get_lig_id (&buffer->info[j]);
unsigned int comp1 = _hb_glyph_info_get_lig_comp (&buffer->cur());
unsigned int comp2 = _hb_glyph_info_get_lig_comp (&buffer->info[j]);
if (likely (id1 == id2))
{
if (id1 == 0) /* Marks belonging to the same base. */
goto good;
else if (comp1 == comp2) /* Marks belonging to the same ligature component. */
goto good;
}
else
{
/* If ligature ids don't match, it may be the case that one of the marks
* itself is a ligature. In which case match. */
if ((id1 > 0 && !comp1) || (id2 > 0 && !comp2))
goto good;
}
/* Didn't match. */
buffer->unsafe_to_concat_from_outbuffer (skippy_iter.idx, buffer->idx + 1);
return_trace (false);
good:
unsigned int mark2_index = (this+mark2Coverage).get_coverage (buffer->info[j].codepoint);
if (mark2_index == NOT_COVERED)
{
buffer->unsafe_to_concat_from_outbuffer (skippy_iter.idx, buffer->idx + 1);
return_trace (false);
}
return_trace ((this+mark1Array).apply (c, mark1_index, mark2_index, this+mark2Array, classCount, j));
}
bool subset (hb_subset_context_t *c) const
{
TRACE_SUBSET (this);
const hb_set_t &glyphset = *c->plan->glyphset_gsub ();
const hb_map_t &glyph_map = *c->plan->glyph_map;
auto *out = c->serializer->start_embed (*this);
if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
out->format = format;
hb_map_t klass_mapping;
Markclass_closure_and_remap_indexes (this+mark1Coverage, this+mark1Array, glyphset, &klass_mapping);
if (!klass_mapping.get_population ()) return_trace (false);
out->classCount = klass_mapping.get_population ();
auto mark1_iter =
+ hb_zip (this+mark1Coverage, this+mark1Array)
| hb_filter (glyphset, hb_first)
;
hb_sorted_vector_t<hb_codepoint_t> new_coverage;
+ mark1_iter
| hb_map (hb_first)
| hb_map (glyph_map)
| hb_sink (new_coverage)
;
if (!out->mark1Coverage.serialize_serialize (c->serializer, new_coverage.iter ()))
return_trace (false);
if (unlikely (!out->mark1Array.serialize_subset (c, mark1Array, this,
(this+mark1Coverage).iter (),
&klass_mapping)))
return_trace (false);
unsigned mark2count = (this+mark2Array).rows;
auto mark2_iter =
+ hb_zip (this+mark2Coverage, hb_range (mark2count))
| hb_filter (glyphset, hb_first)
;
new_coverage.reset ();
hb_sorted_vector_t<unsigned> mark2_indexes;
auto &mark2_array = (this+mark2Array);
for (const auto _ : + mark2_iter)
{
unsigned row = _.second;
bool non_empty = + hb_range ((unsigned) classCount)
| hb_filter (klass_mapping)
| hb_map ([&] (const unsigned col) { return !mark2_array.offset_is_null (row, col, (unsigned) classCount); })
| hb_any
;
if (!non_empty) continue;
hb_codepoint_t new_g = glyph_map.get ( _.first);
new_coverage.push (new_g);
+ hb_range ((unsigned) classCount)
| hb_filter (klass_mapping)
| hb_map ([&] (const unsigned col) { return row * (unsigned) classCount + col; })
| hb_sink (mark2_indexes)
;
}
if (!new_coverage) return_trace (false);
if (!out->mark2Coverage.serialize_serialize (c->serializer, new_coverage.iter ()))
return_trace (false);
return_trace (out->mark2Array.serialize_subset (c, mark2Array, this,
mark2_iter.len (),
mark2_indexes.iter ()));
}
};
}
}
}
#endif /* OT_LAYOUT_GPOS_MARKMARKPOSFORMAT1_HH */
-51
View File
@@ -1,51 +0,0 @@
#ifndef OT_LAYOUT_GPOS_MARKRECORD_HH
#define OT_LAYOUT_GPOS_MARKRECORD_HH
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct MarkRecord
{
friend struct MarkArray;
public:
HBUINT16 klass; /* Class defined for this mark */
Offset16To<Anchor>
markAnchor; /* Offset to Anchor table--from
* beginning of MarkArray table */
public:
DEFINE_SIZE_STATIC (4);
unsigned get_class () const { return (unsigned) klass; }
bool sanitize (hb_sanitize_context_t *c, const void *base) const
{
TRACE_SANITIZE (this);
return_trace (c->check_struct (this) && markAnchor.sanitize (c, base));
}
bool subset (hb_subset_context_t *c,
const void *src_base,
const hb_map_t *klass_mapping) const
{
TRACE_SUBSET (this);
auto *out = c->serializer->embed (this);
if (unlikely (!out)) return_trace (false);
out->klass = klass_mapping->get (klass);
return_trace (out->markAnchor.serialize_subset (c, markAnchor, src_base));
}
void collect_variation_indices (hb_collect_variation_indices_context_t *c,
const void *src_base) const
{
(src_base+markAnchor).collect_variation_indices (c);
}
};
}
}
}
#endif /* OT_LAYOUT_GPOS_MARKRECORD_HH */
-46
View File
@@ -1,46 +0,0 @@
#ifndef OT_LAYOUT_GPOS_PAIRPOS_HH
#define OT_LAYOUT_GPOS_PAIRPOS_HH
#include "PairPosFormat1.hh"
#include "PairPosFormat2.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct PairPos
{
protected:
union {
struct { HBUINT16 v; } format; /* Format identifier */
PairPosFormat1_3<SmallTypes> format1;
PairPosFormat2_4<SmallTypes> format2;
#ifndef HB_NO_BEYOND_64K
PairPosFormat1_3<MediumTypes> format3;
PairPosFormat2_4<MediumTypes> format4;
#endif
} u;
public:
template <typename context_t, typename ...Ts>
typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
{
if (unlikely (!c->may_dispatch (this, &u.format.v))) return c->no_dispatch_return_value ();
TRACE_DISPATCH (this, u.format.v);
switch (u.format.v) {
case 1: return_trace (c->dispatch (u.format1, std::forward<Ts> (ds)...));
case 2: return_trace (c->dispatch (u.format2, std::forward<Ts> (ds)...));
#ifndef HB_NO_BEYOND_64K
case 3: return_trace (c->dispatch (u.format3, std::forward<Ts> (ds)...));
case 4: return_trace (c->dispatch (u.format4, std::forward<Ts> (ds)...));
#endif
default:return_trace (c->default_return_value ());
}
}
};
}
}
}
#endif // OT_LAYOUT_GPOS_PAIRPOS_HH
-254
View File
@@ -1,254 +0,0 @@
#ifndef OT_LAYOUT_GPOS_PAIRPOSFORMAT1_HH
#define OT_LAYOUT_GPOS_PAIRPOSFORMAT1_HH
#include "PairSet.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
template <typename Types>
struct PairPosFormat1_3
{
using PairSet = GPOS_impl::PairSet<Types>;
using PairValueRecord = GPOS_impl::PairValueRecord<Types>;
protected:
HBUINT16 format; /* Format identifier--format = 1 */
typename Types::template OffsetTo<Coverage>
coverage; /* Offset to Coverage table--from
* beginning of subtable */
ValueFormat valueFormat[2]; /* [0] Defines the types of data in
* ValueRecord1--for the first glyph
* in the pair--may be zero (0) */
/* [1] Defines the types of data in
* ValueRecord2--for the second glyph
* in the pair--may be zero (0) */
Array16Of<typename Types::template OffsetTo<PairSet>>
pairSet; /* Array of PairSet tables
* ordered by Coverage Index */
public:
DEFINE_SIZE_ARRAY (8 + Types::size, pairSet);
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
if (!c->check_struct (this)) return_trace (false);
hb_barrier ();
unsigned int len1 = valueFormat[0].get_len ();
unsigned int len2 = valueFormat[1].get_len ();
typename PairSet::sanitize_closure_t closure =
{
valueFormat,
len1,
PairSet::get_size (len1, len2)
};
return_trace (coverage.sanitize (c, this) && pairSet.sanitize (c, this, &closure));
}
bool intersects (const hb_set_t *glyphs) const
{
auto &cov = this+coverage;
if (pairSet.len > glyphs->get_population () * hb_bit_storage ((unsigned) pairSet.len))
{
for (hb_codepoint_t g : glyphs->iter())
{
unsigned i = cov.get_coverage (g);
if ((this+pairSet[i]).intersects (glyphs, valueFormat))
return true;
}
return false;
}
return
+ hb_zip (cov, pairSet)
| hb_filter (*glyphs, hb_first)
| hb_map (hb_second)
| hb_map ([glyphs, this] (const typename Types::template OffsetTo<PairSet> &_)
{ return (this+_).intersects (glyphs, valueFormat); })
| hb_any
;
}
void closure_lookups (hb_closure_lookups_context_t *c) const {}
void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
{
if ((!valueFormat[0].has_device ()) && (!valueFormat[1].has_device ())) return;
auto it =
+ hb_zip (this+coverage, pairSet)
| hb_filter (c->glyph_set, hb_first)
| hb_map (hb_second)
;
if (!it) return;
+ it
| hb_map (hb_add (this))
| hb_apply ([&] (const PairSet& _) { _.collect_variation_indices (c, valueFormat); })
;
}
void collect_glyphs (hb_collect_glyphs_context_t *c) const
{
if (unlikely (!(this+coverage).collect_coverage (c->input))) return;
unsigned int count = pairSet.len;
for (unsigned int i = 0; i < count; i++)
(this+pairSet[i]).collect_glyphs (c, valueFormat);
}
const Coverage &get_coverage () const { return this+coverage; }
struct external_cache_t
{
hb_ot_layout_mapping_cache_t coverage;
};
void *external_cache_create () const
{
external_cache_t *cache = (external_cache_t *) hb_malloc (sizeof (external_cache_t));
if (likely (cache))
{
cache->coverage.clear ();
}
return cache;
}
bool apply (hb_ot_apply_context_t *c, void *external_cache) const
{
TRACE_APPLY (this);
hb_buffer_t *buffer = c->buffer;
#ifndef HB_NO_OT_LAYOUT_LOOKUP_CACHE
external_cache_t *cache = (external_cache_t *) external_cache;
unsigned int index = (this+coverage).get_coverage (buffer->cur().codepoint, cache ? &cache->coverage : nullptr);
#else
unsigned int index = (this+coverage).get_coverage (buffer->cur().codepoint);
#endif
if (index == NOT_COVERED) return_trace (false);
auto &skippy_iter = c->iter_input;
skippy_iter.reset_fast (buffer->idx);
unsigned unsafe_to;
if (unlikely (!skippy_iter.next (&unsafe_to)))
{
buffer->unsafe_to_concat (buffer->idx, unsafe_to);
return_trace (false);
}
return_trace ((this+pairSet[index]).apply (c, valueFormat, skippy_iter.idx));
}
bool subset (hb_subset_context_t *c) const
{
TRACE_SUBSET (this);
const hb_set_t &glyphset = *c->plan->glyphset_gsub ();
const hb_map_t &glyph_map = *c->plan->glyph_map;
auto *out = c->serializer->start_embed (*this);
if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
out->format = format;
hb_pair_t<unsigned, unsigned> newFormats = hb_pair (valueFormat[0], valueFormat[1]);
if (c->plan->normalized_coords)
{
/* all device flags will be dropped when full instancing, no need to strip
* hints, also do not strip emtpy cause we don't compute the new default
* value during stripping */
newFormats = compute_effective_value_formats (glyphset, false, false, &c->plan->layout_variation_idx_delta_map);
}
/* do not strip hints for VF */
else if (c->plan->flags & HB_SUBSET_FLAGS_NO_HINTING)
{
hb_blob_t* blob = hb_face_reference_table (c->plan->source, HB_TAG ('f','v','a','r'));
bool has_fvar = (blob != hb_blob_get_empty ());
hb_blob_destroy (blob);
bool strip = !has_fvar;
/* special case: strip hints when a VF has no GDEF varstore after
* subsetting*/
if (has_fvar && !c->plan->has_gdef_varstore)
strip = true;
newFormats = compute_effective_value_formats (glyphset, strip, true);
}
out->valueFormat[0] = newFormats.first;
out->valueFormat[1] = newFormats.second;
hb_sorted_vector_t<hb_codepoint_t> new_coverage;
+ hb_zip (this+coverage, pairSet)
| hb_filter (glyphset, hb_first)
| hb_filter ([this, c, out] (const typename Types::template OffsetTo<PairSet>& _)
{
auto snap = c->serializer->snapshot ();
auto *o = out->pairSet.serialize_append (c->serializer);
if (unlikely (!o)) return false;
bool ret = o->serialize_subset (c, _, this, valueFormat, out->valueFormat);
if (!ret)
{
out->pairSet.pop ();
c->serializer->revert (snap);
}
return ret;
},
hb_second)
| hb_map (hb_first)
| hb_map (glyph_map)
| hb_sink (new_coverage)
;
out->coverage.serialize_serialize (c->serializer, new_coverage.iter ());
return_trace (bool (new_coverage));
}
hb_pair_t<unsigned, unsigned> compute_effective_value_formats (const hb_set_t& glyphset,
bool strip_hints, bool strip_empty,
const hb_hashmap_t<unsigned, hb_pair_t<unsigned, int>> *varidx_delta_map = nullptr) const
{
unsigned record_size = PairSet::get_size (valueFormat);
unsigned format1 = 0;
unsigned format2 = 0;
for (const auto & _ :
+ hb_zip (this+coverage, pairSet)
| hb_filter (glyphset, hb_first)
| hb_map (hb_second)
)
{
const PairSet& set = (this + _);
const PairValueRecord *record = &set.firstPairValueRecord;
unsigned count = set.len;
for (unsigned i = 0; i < count; i++)
{
if (record->intersects (glyphset))
{
format1 = format1 | valueFormat[0].get_effective_format (record->get_values_1 (), strip_hints, strip_empty, &set, varidx_delta_map);
format2 = format2 | valueFormat[1].get_effective_format (record->get_values_2 (valueFormat[0]), strip_hints, strip_empty, &set, varidx_delta_map);
}
record = &StructAtOffset<const PairValueRecord> (record, record_size);
}
if (format1 == valueFormat[0] && format2 == valueFormat[1])
break;
}
return hb_pair (format1, format2);
}
};
}
}
}
#endif // OT_LAYOUT_GPOS_PAIRPOSFORMAT1_HH
-389
View File
@@ -1,389 +0,0 @@
#ifndef OT_LAYOUT_GPOS_PAIRPOSFORMAT2_HH
#define OT_LAYOUT_GPOS_PAIRPOSFORMAT2_HH
#include "ValueFormat.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
template <typename Types>
struct PairPosFormat2_4 : ValueBase
{
protected:
HBUINT16 format; /* Format identifier--format = 2 */
typename Types::template OffsetTo<Coverage>
coverage; /* Offset to Coverage table--from
* beginning of subtable */
ValueFormat valueFormat1; /* ValueRecord definition--for the
* first glyph of the pair--may be zero
* (0) */
ValueFormat valueFormat2; /* ValueRecord definition--for the
* second glyph of the pair--may be
* zero (0) */
typename Types::template OffsetTo<ClassDef>
classDef1; /* Offset to ClassDef table--from
* beginning of PairPos subtable--for
* the first glyph of the pair */
typename Types::template OffsetTo<ClassDef>
classDef2; /* Offset to ClassDef table--from
* beginning of PairPos subtable--for
* the second glyph of the pair */
HBUINT16 class1Count; /* Number of classes in ClassDef1
* table--includes Class0 */
HBUINT16 class2Count; /* Number of classes in ClassDef2
* table--includes Class0 */
ValueRecord values; /* Matrix of value pairs:
* class1-major, class2-minor,
* Each entry has value1 and value2 */
public:
DEFINE_SIZE_ARRAY (10 + 3 * Types::size, values);
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
if (!(c->check_struct (this)
&& coverage.sanitize (c, this)
&& classDef1.sanitize (c, this)
&& classDef2.sanitize (c, this))) return_trace (false);
unsigned int len1 = valueFormat1.get_len ();
unsigned int len2 = valueFormat2.get_len ();
unsigned int stride = HBUINT16::static_size * (len1 + len2);
unsigned int count = (unsigned int) class1Count * (unsigned int) class2Count;
return_trace (c->check_range ((const void *) values,
count,
stride) &&
(c->lazy_some_gpos ||
(valueFormat1.sanitize_values_stride_unsafe (c, this, &values[0], count, stride) &&
valueFormat2.sanitize_values_stride_unsafe (c, this, &values[len1], count, stride))));
}
bool intersects (const hb_set_t *glyphs) const
{
return (this+coverage).intersects (glyphs) &&
(this+classDef2).intersects (glyphs);
}
void closure_lookups (hb_closure_lookups_context_t *c) const {}
void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
{
if (!intersects (c->glyph_set)) return;
if ((!valueFormat1.has_device ()) && (!valueFormat2.has_device ())) return;
hb_set_t klass1_glyphs, klass2_glyphs;
if (!(this+classDef1).collect_coverage (&klass1_glyphs)) return;
if (!(this+classDef2).collect_coverage (&klass2_glyphs)) return;
hb_set_t class1_set, class2_set;
for (const unsigned cp : + c->glyph_set->iter () | hb_filter (this + coverage))
{
if (!klass1_glyphs.has (cp)) class1_set.add (0);
else
{
unsigned klass1 = (this+classDef1).get (cp);
class1_set.add (klass1);
}
}
class2_set.add (0);
for (const unsigned cp : + c->glyph_set->iter () | hb_filter (klass2_glyphs))
{
unsigned klass2 = (this+classDef2).get (cp);
class2_set.add (klass2);
}
if (class1_set.is_empty ()
|| class2_set.is_empty ()
|| (class2_set.get_population() == 1 && class2_set.has(0)))
return;
unsigned len1 = valueFormat1.get_len ();
unsigned len2 = valueFormat2.get_len ();
const hb_array_t<const Value> values_array = values.as_array ((unsigned)class1Count * (unsigned) class2Count * (len1 + len2));
for (const unsigned class1_idx : class1_set.iter ())
{
for (const unsigned class2_idx : class2_set.iter ())
{
unsigned start_offset = (class1_idx * (unsigned) class2Count + class2_idx) * (len1 + len2);
if (valueFormat1.has_device ())
valueFormat1.collect_variation_indices (c, this, values_array.sub_array (start_offset, len1));
if (valueFormat2.has_device ())
valueFormat2.collect_variation_indices (c, this, values_array.sub_array (start_offset+len1, len2));
}
}
}
void collect_glyphs (hb_collect_glyphs_context_t *c) const
{
if (unlikely (!(this+coverage).collect_coverage (c->input))) return;
if (unlikely (!(this+classDef2).collect_coverage (c->input))) return;
}
const Coverage &get_coverage () const { return this+coverage; }
struct external_cache_t
{
hb_ot_layout_mapping_cache_t coverage;
hb_ot_layout_mapping_cache_t first;
hb_ot_layout_mapping_cache_t second;
};
void *external_cache_create () const
{
external_cache_t *cache = (external_cache_t *) hb_malloc (sizeof (external_cache_t));
if (likely (cache))
{
cache->coverage.clear ();
cache->first.clear ();
cache->second.clear ();
}
return cache;
}
bool apply (hb_ot_apply_context_t *c, void *external_cache) const
{
TRACE_APPLY (this);
hb_buffer_t *buffer = c->buffer;
#ifndef HB_NO_OT_LAYOUT_LOOKUP_CACHE
external_cache_t *cache = (external_cache_t *) external_cache;
unsigned int index = (this+coverage).get_coverage (buffer->cur().codepoint, cache ? &cache->coverage : nullptr);
#else
unsigned int index = (this+coverage).get_coverage (buffer->cur().codepoint);
#endif
if (index == NOT_COVERED) return_trace (false);
auto &skippy_iter = c->iter_input;
skippy_iter.reset_fast (buffer->idx);
unsigned unsafe_to;
if (unlikely (!skippy_iter.next (&unsafe_to)))
{
buffer->unsafe_to_concat (buffer->idx, unsafe_to);
return_trace (false);
}
#ifndef HB_NO_OT_LAYOUT_LOOKUP_CACHE
unsigned int klass1 = (this+classDef1).get_class (buffer->cur().codepoint, cache ? &cache->first : nullptr);
unsigned int klass2 = (this+classDef2).get_class (buffer->info[skippy_iter.idx].codepoint, cache ? &cache->second : nullptr);
#else
unsigned int klass1 = (this+classDef1).get_class (buffer->cur().codepoint);
unsigned int klass2 = (this+classDef2).get_class (buffer->info[skippy_iter.idx].codepoint);
#endif
if (unlikely (klass1 >= class1Count || klass2 >= class2Count))
{
buffer->unsafe_to_concat (buffer->idx, skippy_iter.idx + 1);
return_trace (false);
}
unsigned int len1 = valueFormat1.get_len ();
unsigned int len2 = valueFormat2.get_len ();
unsigned int record_len = len1 + len2;
const Value *v = &values[record_len * (klass1 * class2Count + klass2)];
bool applied_first = false, applied_second = false;
/* Isolate simple kerning and apply it half to each side.
* Results in better cursor positioning / underline drawing.
*
* Disabled, because causes issues... :-(
* https://github.com/harfbuzz/harfbuzz/issues/3408
* https://github.com/harfbuzz/harfbuzz/pull/3235#issuecomment-1029814978
*/
#ifndef HB_SPLIT_KERN
if (false)
#endif
{
if (!len2)
{
const hb_direction_t dir = buffer->props.direction;
const bool horizontal = HB_DIRECTION_IS_HORIZONTAL (dir);
const bool backward = HB_DIRECTION_IS_BACKWARD (dir);
unsigned mask = horizontal ? ValueFormat::xAdvance : ValueFormat::yAdvance;
if (backward)
mask |= mask >> 2; /* Add eg. xPlacement in RTL. */
/* Add Devices. */
mask |= mask << 4;
if (valueFormat1 & ~mask)
goto bail;
/* Is simple kern. Apply value on an empty position slot,
* then split it between sides. */
hb_glyph_position_t pos{};
if (valueFormat1.apply_value (c, this, v, pos))
{
hb_position_t *src = &pos.x_advance;
hb_position_t *dst1 = &buffer->cur_pos().x_advance;
hb_position_t *dst2 = &buffer->pos[skippy_iter.idx].x_advance;
unsigned i = horizontal ? 0 : 1;
hb_position_t kern = src[i];
hb_position_t kern1 = kern >> 1;
hb_position_t kern2 = kern - kern1;
if (!backward)
{
dst1[i] += kern1;
dst2[i] += kern2;
dst2[i + 2] += kern2;
}
else
{
dst1[i] += kern1;
dst1[i + 2] += src[i + 2] - kern2;
dst2[i] += kern2;
}
applied_first = applied_second = kern != 0;
goto success;
}
goto boring;
}
}
bail:
if (HB_BUFFER_MESSAGE_MORE && c->buffer->messaging ())
{
c->buffer->message (c->font,
"try kerning glyphs at %u,%u",
c->buffer->idx, skippy_iter.idx);
}
applied_first = len1 && valueFormat1.apply_value (c, this, v, buffer->cur_pos());
applied_second = len2 && valueFormat2.apply_value (c, this, v + len1, buffer->pos[skippy_iter.idx]);
if (applied_first || applied_second)
if (HB_BUFFER_MESSAGE_MORE && c->buffer->messaging ())
{
c->buffer->message (c->font,
"kerned glyphs at %u,%u",
c->buffer->idx, skippy_iter.idx);
}
if (HB_BUFFER_MESSAGE_MORE && c->buffer->messaging ())
{
c->buffer->message (c->font,
"tried kerning glyphs at %u,%u",
c->buffer->idx, skippy_iter.idx);
}
success:
if (applied_first || applied_second)
buffer->unsafe_to_break (buffer->idx, skippy_iter.idx + 1);
else
boring:
buffer->unsafe_to_concat (buffer->idx, skippy_iter.idx + 1);
if (len2)
{
skippy_iter.idx++;
// https://github.com/harfbuzz/harfbuzz/issues/3824
// https://github.com/harfbuzz/harfbuzz/issues/3888#issuecomment-1326781116
buffer->unsafe_to_break (buffer->idx, skippy_iter.idx + 1);
}
buffer->idx = skippy_iter.idx;
return_trace (true);
}
bool subset (hb_subset_context_t *c) const
{
TRACE_SUBSET (this);
auto *out = c->serializer->start_embed (*this);
if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
out->format = format;
hb_map_t klass1_map;
out->classDef1.serialize_subset (c, classDef1, this, &klass1_map, true, true, &(this + coverage));
out->class1Count = klass1_map.get_population ();
hb_map_t klass2_map;
out->classDef2.serialize_subset (c, classDef2, this, &klass2_map, true, false);
out->class2Count = klass2_map.get_population ();
unsigned len1 = valueFormat1.get_len ();
unsigned len2 = valueFormat2.get_len ();
hb_pair_t<unsigned, unsigned> newFormats = hb_pair (valueFormat1, valueFormat2);
if (c->plan->normalized_coords)
{
/* in case of full instancing, all var device flags will be dropped so no
* need to strip hints here */
newFormats = compute_effective_value_formats (klass1_map, klass2_map, false, false, &c->plan->layout_variation_idx_delta_map);
}
/* do not strip hints for VF */
else if (c->plan->flags & HB_SUBSET_FLAGS_NO_HINTING)
{
hb_blob_t* blob = hb_face_reference_table (c->plan->source, HB_TAG ('f','v','a','r'));
bool has_fvar = (blob != hb_blob_get_empty ());
hb_blob_destroy (blob);
bool strip = !has_fvar;
/* special case: strip hints when a VF has no GDEF varstore after
* subsetting*/
if (has_fvar && !c->plan->has_gdef_varstore)
strip = true;
newFormats = compute_effective_value_formats (klass1_map, klass2_map, strip, true);
}
out->valueFormat1 = newFormats.first;
out->valueFormat2 = newFormats.second;
unsigned total_len = len1 + len2;
hb_vector_t<unsigned> class2_idxs (+ hb_range ((unsigned) class2Count) | hb_filter (klass2_map));
for (unsigned class1_idx : + hb_range ((unsigned) class1Count) | hb_filter (klass1_map))
{
for (unsigned class2_idx : class2_idxs)
{
unsigned idx = (class1_idx * (unsigned) class2Count + class2_idx) * total_len;
valueFormat1.copy_values (c->serializer, out->valueFormat1, this, &values[idx], &c->plan->layout_variation_idx_delta_map);
valueFormat2.copy_values (c->serializer, out->valueFormat2, this, &values[idx + len1], &c->plan->layout_variation_idx_delta_map);
}
}
bool ret = out->coverage.serialize_subset(c, coverage, this);
return_trace (out->class1Count && out->class2Count && ret);
}
hb_pair_t<unsigned, unsigned> compute_effective_value_formats (const hb_map_t& klass1_map,
const hb_map_t& klass2_map,
bool strip_hints, bool strip_empty,
const hb_hashmap_t<unsigned, hb_pair_t<unsigned, int>> *varidx_delta_map = nullptr) const
{
unsigned len1 = valueFormat1.get_len ();
unsigned len2 = valueFormat2.get_len ();
unsigned record_size = len1 + len2;
unsigned format1 = 0;
unsigned format2 = 0;
for (unsigned class1_idx : + hb_range ((unsigned) class1Count) | hb_filter (klass1_map))
{
for (unsigned class2_idx : + hb_range ((unsigned) class2Count) | hb_filter (klass2_map))
{
unsigned idx = (class1_idx * (unsigned) class2Count + class2_idx) * record_size;
format1 = format1 | valueFormat1.get_effective_format (&values[idx], strip_hints, strip_empty, this, varidx_delta_map);
format2 = format2 | valueFormat2.get_effective_format (&values[idx + len1], strip_hints, strip_empty, this, varidx_delta_map);
}
if (format1 == valueFormat1 && format2 == valueFormat2)
break;
}
return hb_pair (format1, format2);
}
};
}
}
}
#endif // OT_LAYOUT_GPOS_PAIRPOSFORMAT2_HH
-210
View File
@@ -1,210 +0,0 @@
#ifndef OT_LAYOUT_GPOS_PAIRSET_HH
#define OT_LAYOUT_GPOS_PAIRSET_HH
#include "PairValueRecord.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
template <typename Types>
struct PairSet : ValueBase
{
template <typename Types2>
friend struct PairPosFormat1_3;
using PairValueRecord = GPOS_impl::PairValueRecord<Types>;
protected:
HBUINT16 len; /* Number of PairValueRecords */
PairValueRecord firstPairValueRecord;
/* Array of PairValueRecords--ordered
* by GlyphID of the second glyph */
public:
DEFINE_SIZE_MIN (2);
static size_t get_size (unsigned len1, unsigned len2)
{
return Types::HBGlyphID::static_size + Value::static_size * (len1 + len2);
}
static size_t get_size (const ValueFormat valueFormats[2])
{
unsigned len1 = valueFormats[0].get_len ();
unsigned len2 = valueFormats[1].get_len ();
return get_size (len1, len2);
}
struct sanitize_closure_t
{
const ValueFormat *valueFormats;
unsigned int len1; /* valueFormats[0].get_len() */
size_t stride; /* bytes */
};
bool sanitize (hb_sanitize_context_t *c, const sanitize_closure_t *closure) const
{
TRACE_SANITIZE (this);
if (!(c->check_struct (this) &&
hb_barrier () &&
c->check_range (&firstPairValueRecord,
len,
closure->stride))) return_trace (false);
hb_barrier ();
unsigned int count = len;
const PairValueRecord *record = &firstPairValueRecord;
return_trace (c->lazy_some_gpos ||
(closure->valueFormats[0].sanitize_values_stride_unsafe (c, this, &record->values[0], count, closure->stride) &&
closure->valueFormats[1].sanitize_values_stride_unsafe (c, this, &record->values[closure->len1], count, closure->stride)));
}
bool intersects (const hb_set_t *glyphs,
const ValueFormat *valueFormats) const
{
unsigned record_size = get_size (valueFormats);
const PairValueRecord *record = &firstPairValueRecord;
unsigned int count = len;
for (unsigned int i = 0; i < count; i++)
{
if (glyphs->has (record->secondGlyph))
return true;
record = &StructAtOffset<const PairValueRecord> (record, record_size);
}
return false;
}
void collect_glyphs (hb_collect_glyphs_context_t *c,
const ValueFormat *valueFormats) const
{
unsigned record_size = get_size (valueFormats);
const PairValueRecord *record = &firstPairValueRecord;
c->input->add_array (&record->secondGlyph, len, record_size);
}
void collect_variation_indices (hb_collect_variation_indices_context_t *c,
const ValueFormat *valueFormats) const
{
unsigned record_size = get_size (valueFormats);
const PairValueRecord *record = &firstPairValueRecord;
unsigned count = len;
for (unsigned i = 0; i < count; i++)
{
if (c->glyph_set->has (record->secondGlyph))
{ record->collect_variation_indices (c, valueFormats, this); }
record = &StructAtOffset<const PairValueRecord> (record, record_size);
}
}
bool apply (hb_ot_apply_context_t *c,
const ValueFormat *valueFormats,
unsigned int pos) const
{
TRACE_APPLY (this);
hb_buffer_t *buffer = c->buffer;
unsigned int len1 = valueFormats[0].get_len ();
unsigned int len2 = valueFormats[1].get_len ();
unsigned record_size = get_size (len1, len2);
const PairValueRecord *record = hb_bsearch (buffer->info[pos].codepoint,
&firstPairValueRecord,
len,
record_size);
if (record)
{
if (HB_BUFFER_MESSAGE_MORE && c->buffer->messaging ())
{
c->buffer->message (c->font,
"try kerning glyphs at %u,%u",
c->buffer->idx, pos);
}
bool applied_first = len1 && valueFormats[0].apply_value (c, this, &record->values[0], buffer->cur_pos());
bool applied_second = len2 && valueFormats[1].apply_value (c, this, &record->values[len1], buffer->pos[pos]);
if (applied_first || applied_second)
if (HB_BUFFER_MESSAGE_MORE && c->buffer->messaging ())
{
c->buffer->message (c->font,
"kerned glyphs at %u,%u",
c->buffer->idx, pos);
}
if (HB_BUFFER_MESSAGE_MORE && c->buffer->messaging ())
{
c->buffer->message (c->font,
"tried kerning glyphs at %u,%u",
c->buffer->idx, pos);
}
if (applied_first || applied_second)
buffer->unsafe_to_break (buffer->idx, pos + 1);
if (len2)
{
pos++;
// https://github.com/harfbuzz/harfbuzz/issues/3824
// https://github.com/harfbuzz/harfbuzz/issues/3888#issuecomment-1326781116
buffer->unsafe_to_break (buffer->idx, pos + 1);
}
buffer->idx = pos;
return_trace (true);
}
buffer->unsafe_to_concat (buffer->idx, pos + 1);
return_trace (false);
}
bool subset (hb_subset_context_t *c,
const ValueFormat valueFormats[2],
const ValueFormat newFormats[2]) const
{
TRACE_SUBSET (this);
auto snap = c->serializer->snapshot ();
auto *out = c->serializer->start_embed (*this);
if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
out->len = 0;
const hb_set_t &glyphset = *c->plan->glyphset_gsub ();
const hb_map_t &glyph_map = *c->plan->glyph_map;
unsigned len1 = valueFormats[0].get_len ();
unsigned len2 = valueFormats[1].get_len ();
unsigned record_size = get_size (len1, len2);
typename PairValueRecord::context_t context =
{
this,
valueFormats,
newFormats,
len1,
&glyph_map,
&c->plan->layout_variation_idx_delta_map
};
const PairValueRecord *record = &firstPairValueRecord;
unsigned count = len, num = 0;
for (unsigned i = 0; i < count; i++)
{
if (glyphset.has (record->secondGlyph)
&& record->subset (c, &context)) num++;
record = &StructAtOffset<const PairValueRecord> (record, record_size);
}
out->len = num;
if (!num) c->serializer->revert (snap);
return_trace (num);
}
};
}
}
}
#endif // OT_LAYOUT_GPOS_PAIRSET_HH
-99
View File
@@ -1,99 +0,0 @@
#ifndef OT_LAYOUT_GPOS_PAIRVALUERECORD_HH
#define OT_LAYOUT_GPOS_PAIRVALUERECORD_HH
#include "ValueFormat.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
template <typename Types>
struct PairValueRecord
{
template <typename Types2>
friend struct PairSet;
protected:
typename Types::HBGlyphID
secondGlyph; /* GlyphID of second glyph in the
* pair--first glyph is listed in the
* Coverage table */
ValueRecord values; /* Positioning data for the first glyph
* followed by for second glyph */
public:
DEFINE_SIZE_ARRAY (Types::HBGlyphID::static_size, values);
int cmp (hb_codepoint_t k) const
{ return secondGlyph.cmp (k); }
struct context_t
{
const ValueBase *base;
const ValueFormat *valueFormats;
const ValueFormat *newFormats;
unsigned len1; /* valueFormats[0].get_len() */
const hb_map_t *glyph_map;
const hb_hashmap_t<unsigned, hb_pair_t<unsigned, int>> *layout_variation_idx_delta_map;
};
bool subset (hb_subset_context_t *c,
context_t *closure) const
{
TRACE_SERIALIZE (this);
auto *s = c->serializer;
auto *out = s->start_embed (*this);
if (unlikely (!s->extend_min (out))) return_trace (false);
out->secondGlyph = (*closure->glyph_map)[secondGlyph];
closure->valueFormats[0].copy_values (s,
closure->newFormats[0],
closure->base, &values[0],
closure->layout_variation_idx_delta_map);
closure->valueFormats[1].copy_values (s,
closure->newFormats[1],
closure->base,
&values[closure->len1],
closure->layout_variation_idx_delta_map);
return_trace (true);
}
void collect_variation_indices (hb_collect_variation_indices_context_t *c,
const ValueFormat *valueFormats,
const ValueBase *base) const
{
unsigned record1_len = valueFormats[0].get_len ();
unsigned record2_len = valueFormats[1].get_len ();
const hb_array_t<const Value> values_array = values.as_array (record1_len + record2_len);
if (valueFormats[0].has_device ())
valueFormats[0].collect_variation_indices (c, base, values_array.sub_array (0, record1_len));
if (valueFormats[1].has_device ())
valueFormats[1].collect_variation_indices (c, base, values_array.sub_array (record1_len, record2_len));
}
bool intersects (const hb_set_t& glyphset) const
{
return glyphset.has(secondGlyph);
}
const Value* get_values_1 () const
{
return &values[0];
}
const Value* get_values_2 (ValueFormat format1) const
{
return &values[format1.get_len ()];
}
};
}
}
}
#endif // OT_LAYOUT_GPOS_PAIRVALUERECORD_HH
-79
View File
@@ -1,79 +0,0 @@
#ifndef OT_LAYOUT_GPOS_POSLOOKUP_HH
#define OT_LAYOUT_GPOS_POSLOOKUP_HH
#include "PosLookupSubTable.hh"
#include "../../../hb-ot-layout-common.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct PosLookup : Lookup
{
using SubTable = PosLookupSubTable;
const SubTable& get_subtable (unsigned int i) const
{ return Lookup::get_subtable<SubTable> (i); }
bool is_reverse () const
{
return false;
}
bool apply (hb_ot_apply_context_t *c) const
{
TRACE_APPLY (this);
return_trace (dispatch (c));
}
bool intersects (const hb_set_t *glyphs) const
{
hb_intersects_context_t c (glyphs);
return dispatch (&c);
}
hb_collect_glyphs_context_t::return_t collect_glyphs (hb_collect_glyphs_context_t *c) const
{ return dispatch (c); }
hb_closure_lookups_context_t::return_t closure_lookups (hb_closure_lookups_context_t *c, unsigned this_index) const
{
if (c->is_lookup_visited (this_index))
return hb_closure_lookups_context_t::default_return_value ();
c->set_lookup_visited (this_index);
if (!intersects (c->glyphs))
{
c->set_lookup_inactive (this_index);
return hb_closure_lookups_context_t::default_return_value ();
}
hb_closure_lookups_context_t::return_t ret = dispatch (c);
return ret;
}
template <typename set_t>
void collect_coverage (set_t *glyphs) const
{
hb_collect_coverage_context_t<set_t> c (glyphs);
dispatch (&c);
}
template <typename context_t>
static typename context_t::return_t dispatch_recurse_func (context_t *c, unsigned int lookup_index);
template <typename context_t, typename ...Ts>
typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
{ return Lookup::dispatch<SubTable> (c, std::forward<Ts> (ds)...); }
bool subset (hb_subset_context_t *c) const
{ return Lookup::subset<SubTable> (c); }
bool sanitize (hb_sanitize_context_t *c) const
{ return Lookup::sanitize<SubTable> (c); }
};
}
}
}
#endif /* OT_LAYOUT_GPOS_POSLOOKUP_HH */
@@ -1,79 +0,0 @@
#ifndef OT_LAYOUT_GPOS_POSLOOKUPSUBTABLE_HH
#define OT_LAYOUT_GPOS_POSLOOKUPSUBTABLE_HH
#include "SinglePos.hh"
#include "PairPos.hh"
#include "CursivePos.hh"
#include "MarkBasePos.hh"
#include "MarkLigPos.hh"
#include "MarkMarkPos.hh"
#include "ContextPos.hh"
#include "ChainContextPos.hh"
#include "ExtensionPos.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct PosLookupSubTable
{
friend struct ::OT::Lookup;
friend struct PosLookup;
enum Type {
Single = 1,
Pair = 2,
Cursive = 3,
MarkBase = 4,
MarkLig = 5,
MarkMark = 6,
Context = 7,
ChainContext = 8,
Extension = 9
};
template <typename context_t, typename ...Ts>
typename context_t::return_t dispatch (context_t *c, unsigned int lookup_type, Ts&&... ds) const
{
TRACE_DISPATCH (this, lookup_type);
switch (lookup_type) {
case Single: return_trace (u.single.dispatch (c, std::forward<Ts> (ds)...));
case Pair: return_trace (u.pair.dispatch (c, std::forward<Ts> (ds)...));
case Cursive: return_trace (u.cursive.dispatch (c, std::forward<Ts> (ds)...));
case MarkBase: return_trace (u.markBase.dispatch (c, std::forward<Ts> (ds)...));
case MarkLig: return_trace (u.markLig.dispatch (c, std::forward<Ts> (ds)...));
case MarkMark: return_trace (u.markMark.dispatch (c, std::forward<Ts> (ds)...));
case Context: return_trace (u.context.dispatch (c, std::forward<Ts> (ds)...));
case ChainContext: return_trace (u.chainContext.dispatch (c, std::forward<Ts> (ds)...));
case Extension: return_trace (u.extension.dispatch (c, std::forward<Ts> (ds)...));
default: return_trace (c->default_return_value ());
}
}
bool intersects (const hb_set_t *glyphs, unsigned int lookup_type) const
{
hb_intersects_context_t c (glyphs);
return dispatch (&c, lookup_type);
}
protected:
union {
SinglePos single;
PairPos pair;
CursivePos cursive;
MarkBasePos markBase;
MarkLigPos markLig;
MarkMarkPos markMark;
ContextPos context;
ChainContextPos chainContext;
ExtensionPos extension;
} u;
public:
DEFINE_SIZE_MIN (0);
};
}
}
}
#endif /* HB_OT_LAYOUT_GPOS_POSLOOKUPSUBTABLE_HH */
-98
View File
@@ -1,98 +0,0 @@
#ifndef OT_LAYOUT_GPOS_SINGLEPOS_HH
#define OT_LAYOUT_GPOS_SINGLEPOS_HH
#include "SinglePosFormat1.hh"
#include "SinglePosFormat2.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct SinglePos
{
protected:
union {
struct { HBUINT16 v; } format; /* Format identifier */
SinglePosFormat1 format1;
SinglePosFormat2 format2;
} u;
public:
template<typename Iterator,
hb_requires (hb_is_iterator (Iterator))>
unsigned get_format (Iterator glyph_val_iter_pairs)
{
hb_array_t<const Value> first_val_iter = hb_second (*glyph_val_iter_pairs);
for (const auto iter : glyph_val_iter_pairs)
for (const auto _ : hb_zip (iter.second, first_val_iter))
if (_.first != _.second)
return 2;
return 1;
}
template<typename Iterator,
typename SrcLookup,
hb_requires (hb_is_iterator (Iterator))>
void serialize (hb_serialize_context_t *c,
const SrcLookup* src,
Iterator glyph_val_iter_pairs,
const hb_hashmap_t<unsigned, hb_pair_t<unsigned, int>> *layout_variation_idx_delta_map,
unsigned newFormat)
{
if (unlikely (!c->extend_min (u.format.v))) return;
unsigned format = 2;
ValueFormat new_format;
new_format = newFormat;
if (glyph_val_iter_pairs)
format = get_format (glyph_val_iter_pairs);
u.format.v = format;
switch (u.format.v) {
case 1: u.format1.serialize (c,
src,
glyph_val_iter_pairs,
new_format,
layout_variation_idx_delta_map);
return;
case 2: u.format2.serialize (c,
src,
glyph_val_iter_pairs,
new_format,
layout_variation_idx_delta_map);
return;
default:return;
}
}
template <typename context_t, typename ...Ts>
typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
{
if (unlikely (!c->may_dispatch (this, &u.format.v))) return c->no_dispatch_return_value ();
TRACE_DISPATCH (this, u.format.v);
switch (u.format.v) {
case 1: return_trace (c->dispatch (u.format1, std::forward<Ts> (ds)...));
case 2: return_trace (c->dispatch (u.format2, std::forward<Ts> (ds)...));
default:return_trace (c->default_return_value ());
}
}
};
template<typename Iterator, typename SrcLookup>
static void
SinglePos_serialize (hb_serialize_context_t *c,
const SrcLookup *src,
Iterator it,
const hb_hashmap_t<unsigned, hb_pair_t<unsigned, int>> *layout_variation_idx_delta_map,
unsigned new_format)
{ c->start_embed<SinglePos> ()->serialize (c, src, it, layout_variation_idx_delta_map, new_format); }
}
}
}
#endif /* OT_LAYOUT_GPOS_SINGLEPOS_HH */
-190
View File
@@ -1,190 +0,0 @@
#ifndef OT_LAYOUT_GPOS_SINGLEPOSFORMAT1_HH
#define OT_LAYOUT_GPOS_SINGLEPOSFORMAT1_HH
#include "Common.hh"
#include "ValueFormat.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct SinglePosFormat1 : ValueBase
{
protected:
HBUINT16 format; /* Format identifier--format = 1 */
Offset16To<Coverage>
coverage; /* Offset to Coverage table--from
* beginning of subtable */
ValueFormat valueFormat; /* Defines the types of data in the
* ValueRecord */
ValueRecord values; /* Defines positioning
* value(s)--applied to all glyphs in
* the Coverage table */
public:
DEFINE_SIZE_ARRAY (6, values);
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (c->check_struct (this) &&
coverage.sanitize (c, this) &&
hb_barrier () &&
/* The coverage table may use a range to represent a set
* of glyphs, which means a small number of bytes can
* generate a large glyph set. Manually modify the
* sanitizer max ops to take this into account.
*
* Note: This check *must* be right after coverage sanitize. */
c->check_ops ((this + coverage).get_population () >> 1) &&
valueFormat.sanitize_value (c, this, values));
}
bool intersects (const hb_set_t *glyphs) const
{ return (this+coverage).intersects (glyphs); }
void closure_lookups (hb_closure_lookups_context_t *c) const {}
void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
{
if (!valueFormat.has_device ()) return;
hb_set_t intersection;
(this+coverage).intersect_set (*c->glyph_set, intersection);
if (!intersection) return;
valueFormat.collect_variation_indices (c, this, values.as_array (valueFormat.get_len ()));
}
void collect_glyphs (hb_collect_glyphs_context_t *c) const
{ if (unlikely (!(this+coverage).collect_coverage (c->input))) return; }
const Coverage &get_coverage () const { return this+coverage; }
ValueFormat get_value_format () const { return valueFormat; }
bool apply (hb_ot_apply_context_t *c) const
{
TRACE_APPLY (this);
hb_buffer_t *buffer = c->buffer;
unsigned int index = (this+coverage).get_coverage (buffer->cur().codepoint);
if (index == NOT_COVERED) return_trace (false);
if (HB_BUFFER_MESSAGE_MORE && c->buffer->messaging ())
{
c->buffer->message (c->font,
"positioning glyph at %u",
c->buffer->idx);
}
valueFormat.apply_value (c, this, values, buffer->cur_pos());
if (HB_BUFFER_MESSAGE_MORE && c->buffer->messaging ())
{
c->buffer->message (c->font,
"positioned glyph at %u",
c->buffer->idx);
}
buffer->idx++;
return_trace (true);
}
bool
position_single (hb_font_t *font,
hb_blob_t *table_blob,
hb_direction_t direction,
hb_codepoint_t gid,
hb_glyph_position_t &pos) const
{
unsigned int index = (this+coverage).get_coverage (gid);
if (likely (index == NOT_COVERED)) return false;
/* This is ugly... */
hb_buffer_t buffer {};
buffer.props.direction = direction;
OT::hb_ot_apply_context_t c (1, font, &buffer, table_blob);
valueFormat.apply_value (&c, this, values, pos);
return true;
}
template<typename Iterator,
typename SrcLookup,
hb_requires (hb_is_iterator (Iterator))>
void serialize (hb_serialize_context_t *c,
const SrcLookup *src,
Iterator it,
ValueFormat newFormat,
const hb_hashmap_t<unsigned, hb_pair_t<unsigned, int>> *layout_variation_idx_delta_map)
{
if (unlikely (!c->extend_min (this))) return;
if (unlikely (!c->check_assign (valueFormat,
newFormat,
HB_SERIALIZE_ERROR_INT_OVERFLOW))) return;
for (const hb_array_t<const Value>& _ : + it | hb_map (hb_second))
{
src->get_value_format ().copy_values (c, newFormat, src, &_, layout_variation_idx_delta_map);
// Only serialize the first entry in the iterator, the rest are assumed to
// be the same.
break;
}
auto glyphs =
+ it
| hb_map_retains_sorting (hb_first)
;
coverage.serialize_serialize (c, glyphs);
}
bool subset (hb_subset_context_t *c) const
{
TRACE_SUBSET (this);
const hb_set_t &glyphset = *c->plan->glyphset_gsub ();
const hb_map_t &glyph_map = *c->plan->glyph_map;
hb_set_t intersection;
(this+coverage).intersect_set (glyphset, intersection);
unsigned new_format = valueFormat;
if (c->plan->normalized_coords)
{
new_format = valueFormat.get_effective_format (values.arrayZ, false, false, this, &c->plan->layout_variation_idx_delta_map);
}
/* do not strip hints for VF */
else if (c->plan->flags & HB_SUBSET_FLAGS_NO_HINTING)
{
hb_blob_t* blob = hb_face_reference_table (c->plan->source, HB_TAG ('f','v','a','r'));
bool has_fvar = (blob != hb_blob_get_empty ());
hb_blob_destroy (blob);
bool strip = !has_fvar;
/* special case: strip hints when a VF has no GDEF varstore after
* subsetting*/
if (has_fvar && !c->plan->has_gdef_varstore)
strip = true;
new_format = valueFormat.get_effective_format (values.arrayZ,
strip, /* strip hints */
true, /* strip empty */
this, nullptr);
}
auto it =
+ hb_iter (intersection)
| hb_map_retains_sorting (glyph_map)
| hb_zip (hb_repeat (values.as_array (valueFormat.get_len ())))
;
bool ret = bool (it);
SinglePos_serialize (c->serializer, this, it, &c->plan->layout_variation_idx_delta_map, new_format);
return_trace (ret);
}
};
}
}
}
#endif /* OT_LAYOUT_GPOS_SINGLEPOSFORMAT1_HH */
-213
View File
@@ -1,213 +0,0 @@
#ifndef OT_LAYOUT_GPOS_SINGLEPOSFORMAT2_HH
#define OT_LAYOUT_GPOS_SINGLEPOSFORMAT2_HH
#include "Common.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
struct SinglePosFormat2 : ValueBase
{
protected:
HBUINT16 format; /* Format identifier--format = 2 */
Offset16To<Coverage>
coverage; /* Offset to Coverage table--from
* beginning of subtable */
ValueFormat valueFormat; /* Defines the types of data in the
* ValueRecord */
HBUINT16 valueCount; /* Number of ValueRecords */
ValueRecord values; /* Array of ValueRecords--positioning
* values applied to glyphs */
public:
DEFINE_SIZE_ARRAY (8, values);
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (c->check_struct (this) &&
coverage.sanitize (c, this) &&
valueFormat.sanitize_values (c, this, values, valueCount));
}
bool intersects (const hb_set_t *glyphs) const
{ return (this+coverage).intersects (glyphs); }
void closure_lookups (hb_closure_lookups_context_t *c) const {}
void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
{
if (!valueFormat.has_device ()) return;
auto it =
+ hb_zip (this+coverage, hb_range ((unsigned) valueCount))
| hb_filter (c->glyph_set, hb_first)
;
if (!it) return;
unsigned sub_length = valueFormat.get_len ();
const hb_array_t<const Value> values_array = values.as_array (valueCount * sub_length);
for (unsigned i : + it
| hb_map (hb_second))
valueFormat.collect_variation_indices (c, this, values_array.sub_array (i * sub_length, sub_length));
}
void collect_glyphs (hb_collect_glyphs_context_t *c) const
{ if (unlikely (!(this+coverage).collect_coverage (c->input))) return; }
const Coverage &get_coverage () const { return this+coverage; }
ValueFormat get_value_format () const { return valueFormat; }
bool apply (hb_ot_apply_context_t *c) const
{
TRACE_APPLY (this);
hb_buffer_t *buffer = c->buffer;
unsigned int index = (this+coverage).get_coverage (buffer->cur().codepoint);
if (index == NOT_COVERED) return_trace (false);
if (unlikely (index >= valueCount)) return_trace (false);
if (HB_BUFFER_MESSAGE_MORE && c->buffer->messaging ())
{
c->buffer->message (c->font,
"positioning glyph at %u",
c->buffer->idx);
}
valueFormat.apply_value (c, this,
&values[index * valueFormat.get_len ()],
buffer->cur_pos());
if (HB_BUFFER_MESSAGE_MORE && c->buffer->messaging ())
{
c->buffer->message (c->font,
"positioned glyph at %u",
c->buffer->idx);
}
buffer->idx++;
return_trace (true);
}
bool
position_single (hb_font_t *font,
hb_blob_t *table_blob,
hb_direction_t direction,
hb_codepoint_t gid,
hb_glyph_position_t &pos) const
{
unsigned int index = (this+coverage).get_coverage (gid);
if (likely (index == NOT_COVERED)) return false;
if (unlikely (index >= valueCount)) return false;
/* This is ugly... */
hb_buffer_t buffer {};
buffer.props.direction = direction;
OT::hb_ot_apply_context_t c (1, font, &buffer, table_blob);
valueFormat.apply_value (&c, this,
&values[index * valueFormat.get_len ()],
pos);
return true;
}
template<typename Iterator,
typename SrcLookup,
hb_requires (hb_is_iterator (Iterator))>
void serialize (hb_serialize_context_t *c,
const SrcLookup *src,
Iterator it,
ValueFormat newFormat,
const hb_hashmap_t<unsigned, hb_pair_t<unsigned, int>> *layout_variation_idx_delta_map)
{
auto out = c->extend_min (this);
if (unlikely (!out)) return;
if (unlikely (!c->check_assign (valueFormat, newFormat, HB_SERIALIZE_ERROR_INT_OVERFLOW))) return;
if (unlikely (!c->check_assign (valueCount, it.len (), HB_SERIALIZE_ERROR_ARRAY_OVERFLOW))) return;
+ it
| hb_map (hb_second)
| hb_apply ([&] (hb_array_t<const Value> _)
{ src->get_value_format ().copy_values (c, newFormat, src, &_, layout_variation_idx_delta_map); })
;
auto glyphs =
+ it
| hb_map_retains_sorting (hb_first)
;
coverage.serialize_serialize (c, glyphs);
}
template<typename Iterator,
hb_requires (hb_is_iterator (Iterator))>
unsigned compute_effective_format (const hb_face_t *face,
Iterator it,
bool is_instancing, bool strip_hints,
bool has_gdef_varstore,
const hb_hashmap_t<unsigned, hb_pair_t<unsigned, int>> *varidx_delta_map) const
{
hb_blob_t* blob = hb_face_reference_table (face, HB_TAG ('f','v','a','r'));
bool has_fvar = (blob != hb_blob_get_empty ());
hb_blob_destroy (blob);
unsigned new_format = 0;
if (is_instancing)
{
new_format = new_format | valueFormat.get_effective_format (+ it | hb_map (hb_second), false, false, this, varidx_delta_map);
}
/* do not strip hints for VF */
else if (strip_hints)
{
bool strip = !has_fvar;
if (has_fvar && !has_gdef_varstore)
strip = true;
new_format = new_format | valueFormat.get_effective_format (+ it | hb_map (hb_second), strip, true, this, nullptr);
}
else
new_format = valueFormat;
return new_format;
}
bool subset (hb_subset_context_t *c) const
{
TRACE_SUBSET (this);
const hb_set_t &glyphset = *c->plan->glyphset_gsub ();
const hb_map_t &glyph_map = *c->plan->glyph_map;
unsigned sub_length = valueFormat.get_len ();
auto values_array = values.as_array (valueCount * sub_length);
auto it =
+ hb_zip (this+coverage, hb_range ((unsigned) valueCount))
| hb_filter (glyphset, hb_first)
| hb_map_retains_sorting ([&] (const hb_pair_t<hb_codepoint_t, unsigned>& _)
{
return hb_pair (glyph_map[_.first],
values_array.sub_array (_.second * sub_length,
sub_length));
})
;
unsigned new_format = compute_effective_format (c->plan->source, it,
bool (c->plan->normalized_coords),
bool (c->plan->flags & HB_SUBSET_FLAGS_NO_HINTING),
c->plan->has_gdef_varstore,
&c->plan->layout_variation_idx_delta_map);
bool ret = bool (it);
SinglePos_serialize (c->serializer, this, it, &c->plan->layout_variation_idx_delta_map, new_format);
return_trace (ret);
}
};
}
}
}
#endif /* OT_LAYOUT_GPOS_SINGLEPOSFORMAT2_HH */
-446
View File
@@ -1,446 +0,0 @@
#ifndef OT_LAYOUT_GPOS_VALUEFORMAT_HH
#define OT_LAYOUT_GPOS_VALUEFORMAT_HH
#include "../../../hb-ot-layout-gsubgpos.hh"
namespace OT {
namespace Layout {
namespace GPOS_impl {
typedef HBUINT16 Value;
struct ValueBase {}; // Dummy base class tag for OffsetTo<Value> bases.
typedef UnsizedArrayOf<Value> ValueRecord;
struct ValueFormat : HBUINT16
{
enum Flags {
xPlacement = 0x0001u, /* Includes horizontal adjustment for placement */
yPlacement = 0x0002u, /* Includes vertical adjustment for placement */
xAdvance = 0x0004u, /* Includes horizontal adjustment for advance */
yAdvance = 0x0008u, /* Includes vertical adjustment for advance */
xPlaDevice = 0x0010u, /* Includes horizontal Device table for placement */
yPlaDevice = 0x0020u, /* Includes vertical Device table for placement */
xAdvDevice = 0x0040u, /* Includes horizontal Device table for advance */
yAdvDevice = 0x0080u, /* Includes vertical Device table for advance */
ignored = 0x0F00u, /* Was used in TrueType Open for MM fonts */
reserved = 0xF000u, /* For future use */
devices = 0x00F0u /* Mask for having any Device table */
};
/* All fields are options. Only those available advance the value pointer. */
#if 0
HBINT16 xPlacement; /* Horizontal adjustment for
* placement--in design units */
HBINT16 yPlacement; /* Vertical adjustment for
* placement--in design units */
HBINT16 xAdvance; /* Horizontal adjustment for
* advance--in design units (only used
* for horizontal writing) */
HBINT16 yAdvance; /* Vertical adjustment for advance--in
* design units (only used for vertical
* writing) */
Offset16To<Device> xPlaDevice; /* Offset to Device table for
* horizontal placement--measured from
* beginning of PosTable (may be NULL) */
Offset16To<Device> yPlaDevice; /* Offset to Device table for vertical
* placement--measured from beginning
* of PosTable (may be NULL) */
Offset16To<Device> xAdvDevice; /* Offset to Device table for
* horizontal advance--measured from
* beginning of PosTable (may be NULL) */
Offset16To<Device> yAdvDevice; /* Offset to Device table for vertical
* advance--measured from beginning of
* PosTable (may be NULL) */
#endif
NumType& operator = (uint16_t i) { v = i; return *this; }
// Note: spec says skip 2 bytes per bit in the valueformat. But reports
// from Microsoft developers indicate that only the fields that are
// currently defined are counted. We don't expect any new fields to
// be added to ValueFormat. As such, we use the faster hb_popcount8
// that only processes the lowest 8 bits.
unsigned int get_len () const { return hb_popcount8 ((uint8_t) *this); }
size_t get_size () const { return get_len () * Value::static_size; }
hb_vector_t<unsigned> get_device_table_indices () const {
unsigned i = 0;
hb_vector_t<unsigned> result;
unsigned format = *this;
if (format & xPlacement) i++;
if (format & yPlacement) i++;
if (format & xAdvance) i++;
if (format & yAdvance) i++;
if (format & xPlaDevice) result.push (i++);
if (format & yPlaDevice) result.push (i++);
if (format & xAdvDevice) result.push (i++);
if (format & yAdvDevice) result.push (i++);
return result;
}
bool apply_value (hb_ot_apply_context_t *c,
const ValueBase *base,
const Value *values,
hb_glyph_position_t &glyph_pos) const
{
bool ret = false;
unsigned int format = *this;
if (!format) return ret;
hb_font_t *font = c->font;
bool horizontal =
#ifndef HB_NO_VERTICAL
HB_DIRECTION_IS_HORIZONTAL (c->direction)
#else
true
#endif
;
if (format & xPlacement) glyph_pos.x_offset += font->em_scale_x (get_short (values++, &ret));
if (format & yPlacement) glyph_pos.y_offset += font->em_scale_y (get_short (values++, &ret));
if (format & xAdvance) {
if (likely (horizontal)) glyph_pos.x_advance += font->em_scale_x (get_short (values, &ret));
values++;
}
/* y_advance values grow downward but font-space grows upward, hence negation */
if (format & yAdvance) {
if (unlikely (!horizontal)) glyph_pos.y_advance -= font->em_scale_y (get_short (values, &ret));
values++;
}
if (!has_device ()) return ret;
bool use_x_device = font->x_ppem || font->has_nonzero_coords;
bool use_y_device = font->y_ppem || font->has_nonzero_coords;
if (!use_x_device && !use_y_device) return ret;
const ItemVariationStore &store = c->var_store;
auto *cache = c->var_store_cache;
/* pixel -> fractional pixel */
if (format & xPlaDevice)
{
if (use_x_device) glyph_pos.x_offset += get_device (values, &ret, base, c->sanitizer).get_x_delta (font, store, cache);
values++;
}
if (format & yPlaDevice)
{
if (use_y_device) glyph_pos.y_offset += get_device (values, &ret, base, c->sanitizer).get_y_delta (font, store, cache);
values++;
}
if (format & xAdvDevice)
{
if (horizontal && use_x_device) glyph_pos.x_advance += get_device (values, &ret, base, c->sanitizer).get_x_delta (font, store, cache);
values++;
}
if (format & yAdvDevice)
{
/* y_advance values grow downward but font-space grows upward, hence negation */
if (!horizontal && use_y_device) glyph_pos.y_advance -= get_device (values, &ret, base, c->sanitizer).get_y_delta (font, store, cache);
values++;
}
return ret;
}
unsigned int get_effective_format (const Value *values, bool strip_hints, bool strip_empty, const ValueBase *base,
const hb_hashmap_t<unsigned, hb_pair_t<unsigned, int>> *varidx_delta_map) const
{
unsigned int format = *this;
for (unsigned flag = xPlacement; flag <= yAdvDevice; flag = flag << 1) {
if (format & flag)
{
if (strip_hints && flag >= xPlaDevice)
{
format = format & ~flag;
values++;
continue;
}
if (varidx_delta_map && flag >= xPlaDevice)
{
update_var_flag (values++, (Flags) flag, &format, base, varidx_delta_map);
continue;
}
/* do not strip empty when instancing, cause we don't know whether the new
* default value is 0 or not */
if (strip_empty) should_drop (*values, (Flags) flag, &format);
values++;
}
}
return format;
}
template<typename Iterator,
hb_requires (hb_is_iterator (Iterator))>
unsigned int get_effective_format (Iterator it, bool strip_hints, bool strip_empty, const ValueBase *base,
const hb_hashmap_t<unsigned, hb_pair_t<unsigned, int>> *varidx_delta_map) const {
unsigned int new_format = 0;
for (const hb_array_t<const Value>& values : it)
new_format = new_format | get_effective_format (&values, strip_hints, strip_empty, base, varidx_delta_map);
return new_format;
}
void copy_values (hb_serialize_context_t *c,
unsigned int new_format,
const ValueBase *base,
const Value *values,
const hb_hashmap_t<unsigned, hb_pair_t<unsigned, int>> *layout_variation_idx_delta_map) const
{
unsigned int format = *this;
if (!format) return;
HBINT16 *x_placement = nullptr, *y_placement = nullptr, *x_adv = nullptr, *y_adv = nullptr;
if (format & xPlacement) x_placement = copy_value (c, new_format, xPlacement, *values++);
if (format & yPlacement) y_placement = copy_value (c, new_format, yPlacement, *values++);
if (format & xAdvance) x_adv = copy_value (c, new_format, xAdvance, *values++);
if (format & yAdvance) y_adv = copy_value (c, new_format, yAdvance, *values++);
if (!has_device ())
return;
if (format & xPlaDevice)
{
add_delta_to_value (x_placement, base, values, layout_variation_idx_delta_map);
copy_device (c, base, values++, layout_variation_idx_delta_map, new_format, xPlaDevice);
}
if (format & yPlaDevice)
{
add_delta_to_value (y_placement, base, values, layout_variation_idx_delta_map);
copy_device (c, base, values++, layout_variation_idx_delta_map, new_format, yPlaDevice);
}
if (format & xAdvDevice)
{
add_delta_to_value (x_adv, base, values, layout_variation_idx_delta_map);
copy_device (c, base, values++, layout_variation_idx_delta_map, new_format, xAdvDevice);
}
if (format & yAdvDevice)
{
add_delta_to_value (y_adv, base, values, layout_variation_idx_delta_map);
copy_device (c, base, values++, layout_variation_idx_delta_map, new_format, yAdvDevice);
}
}
HBINT16* copy_value (hb_serialize_context_t *c,
unsigned int new_format,
Flags flag,
Value value) const
{
// Filter by new format.
if (!(new_format & flag)) return nullptr;
return reinterpret_cast<HBINT16 *> (c->copy (value));
}
void collect_variation_indices (hb_collect_variation_indices_context_t *c,
const ValueBase *base,
const hb_array_t<const Value>& values) const
{
unsigned format = *this;
unsigned i = 0;
if (format & xPlacement) i++;
if (format & yPlacement) i++;
if (format & xAdvance) i++;
if (format & yAdvance) i++;
if (format & xPlaDevice)
{
(base + get_device (&(values[i]))).collect_variation_indices (c);
i++;
}
if (format & ValueFormat::yPlaDevice)
{
(base + get_device (&(values[i]))).collect_variation_indices (c);
i++;
}
if (format & ValueFormat::xAdvDevice)
{
(base + get_device (&(values[i]))).collect_variation_indices (c);
i++;
}
if (format & ValueFormat::yAdvDevice)
{
(base + get_device (&(values[i]))).collect_variation_indices (c);
i++;
}
}
private:
bool sanitize_value_devices (hb_sanitize_context_t *c, const ValueBase *base, const Value *values) const
{
unsigned int format = *this;
if (format & xPlacement) values++;
if (format & yPlacement) values++;
if (format & xAdvance) values++;
if (format & yAdvance) values++;
if ((format & xPlaDevice) && !get_device (values++).sanitize (c, base)) return false;
if ((format & yPlaDevice) && !get_device (values++).sanitize (c, base)) return false;
if ((format & xAdvDevice) && !get_device (values++).sanitize (c, base)) return false;
if ((format & yAdvDevice) && !get_device (values++).sanitize (c, base)) return false;
return true;
}
static inline Offset16To<Device, ValueBase>& get_device (Value* value)
{
return *static_cast<Offset16To<Device, ValueBase> *> (value);
}
static inline const Offset16To<Device, ValueBase>& get_device (const Value* value)
{
return *static_cast<const Offset16To<Device, ValueBase> *> (value);
}
static inline const Device& get_device (const Value* value,
bool *worked,
const ValueBase *base,
hb_sanitize_context_t &c)
{
if (worked) *worked |= bool (*value);
auto &offset = *static_cast<const Offset16To<Device> *> (value);
if (unlikely (!offset.sanitize (&c, base)))
return Null(Device);
hb_barrier ();
return base + offset;
}
void add_delta_to_value (HBINT16 *value,
const ValueBase *base,
const Value *src_value,
const hb_hashmap_t<unsigned, hb_pair_t<unsigned, int>> *layout_variation_idx_delta_map) const
{
if (!value) return;
unsigned varidx = (base + get_device (src_value)).get_variation_index ();
hb_pair_t<unsigned, int> *varidx_delta;
if (!layout_variation_idx_delta_map->has (varidx, &varidx_delta)) return;
*value += hb_second (*varidx_delta);
}
bool copy_device (hb_serialize_context_t *c,
const ValueBase *base,
const Value *src_value,
const hb_hashmap_t<unsigned, hb_pair_t<unsigned, int>> *layout_variation_idx_delta_map,
unsigned int new_format, Flags flag) const
{
// Filter by new format.
if (!(new_format & flag)) return true;
Value *dst_value = c->copy (*src_value);
if (!dst_value) return false;
if (*dst_value == 0) return true;
*dst_value = 0;
c->push ();
if ((base + get_device (src_value)).copy (c, layout_variation_idx_delta_map))
{
c->add_link (*dst_value, c->pop_pack ());
return true;
}
else
{
c->pop_discard ();
return false;
}
}
static inline const HBINT16& get_short (const Value* value, bool *worked=nullptr)
{
if (worked) *worked |= bool (*value);
return *reinterpret_cast<const HBINT16 *> (value);
}
public:
bool has_device () const
{
unsigned int format = *this;
return (format & devices) != 0;
}
bool sanitize_value (hb_sanitize_context_t *c, const ValueBase *base, const Value *values) const
{
TRACE_SANITIZE (this);
if (unlikely (!c->check_range (values, get_size ()))) return_trace (false);
if (c->lazy_some_gpos)
return_trace (true);
return_trace (!has_device () || sanitize_value_devices (c, base, values));
}
bool sanitize_values (hb_sanitize_context_t *c, const ValueBase *base, const Value *values, unsigned int count) const
{
TRACE_SANITIZE (this);
unsigned size = get_size ();
if (!c->check_range (values, count, size)) return_trace (false);
if (c->lazy_some_gpos)
return_trace (true);
hb_barrier ();
return_trace (sanitize_values_stride_unsafe (c, base, values, count, size));
}
/* Just sanitize referenced Device tables. Doesn't check the values themselves. */
bool sanitize_values_stride_unsafe (hb_sanitize_context_t *c, const ValueBase *base, const Value *values, unsigned int count, unsigned int stride) const
{
TRACE_SANITIZE (this);
if (!has_device ()) return_trace (true);
for (unsigned int i = 0; i < count; i++) {
if (!sanitize_value_devices (c, base, values))
return_trace (false);
values = &StructAtOffset<const Value> (values, stride);
}
return_trace (true);
}
private:
void should_drop (Value value, Flags flag, unsigned int* format) const
{
if (value) return;
*format = *format & ~flag;
}
void update_var_flag (const Value* value, Flags flag,
unsigned int* format, const ValueBase *base,
const hb_hashmap_t<unsigned, hb_pair_t<unsigned, int>> *varidx_delta_map) const
{
if (*value)
{
unsigned varidx = (base + get_device (value)).get_variation_index ();
hb_pair_t<unsigned, int> *varidx_delta;
if (varidx_delta_map->has (varidx, &varidx_delta) &&
varidx_delta->first != HB_OT_LAYOUT_NO_VARIATIONS_INDEX)
return;
}
*format = *format & ~flag;
}
};
}
}
}
#endif // #ifndef OT_LAYOUT_GPOS_VALUEFORMAT_HH
-139
View File
@@ -1,139 +0,0 @@
#ifndef OT_LAYOUT_GSUB_ALTERNATESET_HH
#define OT_LAYOUT_GSUB_ALTERNATESET_HH
#include "Common.hh"
namespace OT {
namespace Layout {
namespace GSUB_impl {
template <typename Types>
struct AlternateSet
{
protected:
Array16Of<typename Types::HBGlyphID>
alternates; /* Array of alternate GlyphIDs--in
* arbitrary order */
public:
DEFINE_SIZE_ARRAY (2, alternates);
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (alternates.sanitize (c));
}
bool intersects (const hb_set_t *glyphs) const
{ return hb_any (alternates, glyphs); }
void closure (hb_closure_context_t *c) const
{ c->output->add_array (alternates.arrayZ, alternates.len); }
void collect_glyphs (hb_collect_glyphs_context_t *c) const
{ c->output->add_array (alternates.arrayZ, alternates.len); }
bool apply (hb_ot_apply_context_t *c) const
{
TRACE_APPLY (this);
unsigned int count = alternates.len;
if (unlikely (!count)) return_trace (false);
hb_mask_t glyph_mask = c->buffer->cur().mask;
hb_mask_t lookup_mask = c->lookup_mask;
/* Note: This breaks badly if two features enabled this lookup together. */
unsigned int shift = hb_ctz (lookup_mask);
unsigned int alt_index = ((lookup_mask & glyph_mask) >> shift);
/* If alt_index is MAX_VALUE, randomize feature if it is the rand feature. */
if (alt_index == HB_OT_MAP_MAX_VALUE && c->random)
{
/* Maybe we can do better than unsafe-to-break all; but since we are
* changing random state, it would be hard to track that. Good 'nough. */
c->buffer->unsafe_to_break (0, c->buffer->len);
alt_index = c->random_number () % count + 1;
}
if (unlikely (alt_index > count || alt_index == 0)) return_trace (false);
if (HB_BUFFER_MESSAGE_MORE && c->buffer->messaging ())
{
c->buffer->sync_so_far ();
c->buffer->message (c->font,
"replacing glyph at %u (alternate substitution)",
c->buffer->idx);
}
c->replace_glyph (alternates[alt_index - 1]);
if (HB_BUFFER_MESSAGE_MORE && c->buffer->messaging ())
{
c->buffer->message (c->font,
"replaced glyph at %u (alternate substitution)",
c->buffer->idx - 1u);
}
return_trace (true);
}
unsigned
get_alternates (unsigned start_offset,
unsigned *alternate_count /* IN/OUT. May be NULL. */,
hb_codepoint_t *alternate_glyphs /* OUT. May be NULL. */) const
{
if (alternates.len && alternate_count)
{
+ alternates.as_array ().sub_array (start_offset, alternate_count)
| hb_sink (hb_array (alternate_glyphs, *alternate_count))
;
}
return alternates.len;
}
void
collect_alternates (hb_codepoint_t gid,
hb_map_t *alternate_count /* IN/OUT */,
hb_map_t *alternate_glyphs /* IN/OUT */) const
{
+ hb_enumerate (alternates)
| hb_map ([gid] (hb_pair_t<unsigned, hb_codepoint_t> _) { return hb_pair (gid + (_.first << 24), _.second); })
| hb_apply ([&] (const hb_pair_t<hb_codepoint_t, hb_codepoint_t> &p) -> void
{ _hb_collect_glyph_alternates_add (p.first, p.second,
alternate_count, alternate_glyphs); })
;
}
template <typename Iterator,
hb_requires (hb_is_source_of (Iterator, hb_codepoint_t))>
bool serialize (hb_serialize_context_t *c,
Iterator alts)
{
TRACE_SERIALIZE (this);
return_trace (alternates.serialize (c, alts));
}
bool subset (hb_subset_context_t *c) const
{
TRACE_SUBSET (this);
const hb_set_t &glyphset = *c->plan->glyphset_gsub ();
const hb_map_t &glyph_map = *c->plan->glyph_map;
auto it =
+ hb_iter (alternates)
| hb_filter (glyphset)
| hb_map (glyph_map)
;
auto *out = c->serializer->start_embed (*this);
return_trace (out->serialize (c->serializer, it) &&
out->alternates);
}
};
}
}
}
#endif /* OT_LAYOUT_GSUB_ALTERNATESET_HH */
-62
View File
@@ -1,62 +0,0 @@
#ifndef OT_LAYOUT_GSUB_ALTERNATESUBST_HH
#define OT_LAYOUT_GSUB_ALTERNATESUBST_HH
#include "AlternateSubstFormat1.hh"
#include "Common.hh"
namespace OT {
namespace Layout {
namespace GSUB_impl {
struct AlternateSubst
{
protected:
union {
struct { HBUINT16 v; } format; /* Format identifier */
AlternateSubstFormat1_2<SmallTypes> format1;
#ifndef HB_NO_BEYOND_64K
AlternateSubstFormat1_2<MediumTypes> format2;
#endif
} u;
public:
template <typename context_t, typename ...Ts>
typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
{
if (unlikely (!c->may_dispatch (this, &u.format.v))) return c->no_dispatch_return_value ();
TRACE_DISPATCH (this, u.format.v);
switch (u.format.v) {
case 1: return_trace (c->dispatch (u.format1, std::forward<Ts> (ds)...));
#ifndef HB_NO_BEYOND_64K
case 2: return_trace (c->dispatch (u.format2, std::forward<Ts> (ds)...));
#endif
default:return_trace (c->default_return_value ());
}
}
/* TODO This function is unused and not updated to 24bit GIDs. Should be done by using
* iterators. While at it perhaps using iterator of arrays of hb_codepoint_t instead. */
bool serialize (hb_serialize_context_t *c,
hb_sorted_array_t<const HBGlyphID16> glyphs,
hb_array_t<const unsigned int> alternate_len_list,
hb_array_t<const HBGlyphID16> alternate_glyphs_list)
{
TRACE_SERIALIZE (this);
if (unlikely (!c->extend_min (u.format.v))) return_trace (false);
unsigned int format = 1;
u.format.v = format;
switch (u.format.v) {
case 1: return_trace (u.format1.serialize (c, glyphs, alternate_len_list, alternate_glyphs_list));
default:return_trace (false);
}
}
/* TODO subset() should choose format. */
};
}
}
}
#endif /* OT_LAYOUT_GSUB_ALTERNATESUBST_HH */
@@ -1,141 +0,0 @@
#ifndef OT_LAYOUT_GSUB_ALTERNATESUBSTFORMAT1_HH
#define OT_LAYOUT_GSUB_ALTERNATESUBSTFORMAT1_HH
#include "AlternateSet.hh"
#include "Common.hh"
namespace OT {
namespace Layout {
namespace GSUB_impl {
template <typename Types>
struct AlternateSubstFormat1_2
{
protected:
HBUINT16 format; /* Format identifier--format = 1 */
typename Types::template OffsetTo<Coverage>
coverage; /* Offset to Coverage table--from
* beginning of Substitution table */
Array16Of<typename Types::template OffsetTo<AlternateSet<Types>>>
alternateSet; /* Array of AlternateSet tables
* ordered by Coverage Index */
public:
DEFINE_SIZE_ARRAY (2 + 2 * Types::size, alternateSet);
bool sanitize (hb_sanitize_context_t *c) const
{
TRACE_SANITIZE (this);
return_trace (coverage.sanitize (c, this) && alternateSet.sanitize (c, this));
}
bool intersects (const hb_set_t *glyphs) const
{ return (this+coverage).intersects (glyphs); }
bool may_have_non_1to1 () const
{ return false; }
void closure (hb_closure_context_t *c) const
{
+ hb_zip (this+coverage, alternateSet)
| hb_filter (c->parent_active_glyphs (), hb_first)
| hb_map (hb_second)
| hb_map (hb_add (this))
| hb_apply ([c] (const AlternateSet<Types> &_) { _.closure (c); })
;
}
void closure_lookups (hb_closure_lookups_context_t *c) const {}
void collect_glyphs (hb_collect_glyphs_context_t *c) const
{
if (unlikely (!(this+coverage).collect_coverage (c->input))) return;
+ hb_zip (this+coverage, alternateSet)
| hb_map (hb_second)
| hb_map (hb_add (this))
| hb_apply ([c] (const AlternateSet<Types> &_) { _.collect_glyphs (c); })
;
}
const Coverage &get_coverage () const { return this+coverage; }
bool would_apply (hb_would_apply_context_t *c) const
{ return c->len == 1 && (this+coverage).get_coverage (c->glyphs[0]) != NOT_COVERED; }
unsigned
get_glyph_alternates (hb_codepoint_t gid,
unsigned start_offset,
unsigned *alternate_count /* IN/OUT. May be NULL. */,
hb_codepoint_t *alternate_glyphs /* OUT. May be NULL. */) const
{ return (this+alternateSet[(this+coverage).get_coverage (gid)])
.get_alternates (start_offset, alternate_count, alternate_glyphs); }
void
collect_glyph_alternates (hb_map_t *alternate_count /* IN/OUT */,
hb_map_t *alternate_glyphs /* IN/OUT */) const
{
+ hb_iter (alternateSet)
| hb_map (hb_add (this))
| hb_zip (this+coverage)
| hb_apply ([&] (const hb_pair_t<const AlternateSet<Types> &, hb_codepoint_t> _) {
_.first.collect_alternates (_.second, alternate_count, alternate_glyphs);
})
;
}
bool apply (hb_ot_apply_context_t *c) const
{
TRACE_APPLY (this);
unsigned int index = (this+coverage).get_coverage (c->buffer->cur().codepoint);
if (index == NOT_COVERED) return_trace (false);
return_trace ((this+alternateSet[index]).apply (c));
}
bool serialize (hb_serialize_context_t *c,
hb_sorted_array_t<const HBGlyphID16> glyphs,
hb_array_t<const unsigned int> alternate_len_list,
hb_array_t<const HBGlyphID16> alternate_glyphs_list)
{
TRACE_SERIALIZE (this);
if (unlikely (!c->extend_min (this))) return_trace (false);
if (unlikely (!alternateSet.serialize (c, glyphs.length))) return_trace (false);
for (unsigned int i = 0; i < glyphs.length; i++)
{
unsigned int alternate_len = alternate_len_list[i];
if (unlikely (!alternateSet[i]
.serialize_serialize (c, alternate_glyphs_list.sub_array (0, alternate_len))))
return_trace (false);
alternate_glyphs_list += alternate_len;
}
return_trace (coverage.serialize_serialize (c, glyphs));
}
bool subset (hb_subset_context_t *c) const
{
TRACE_SUBSET (this);
const hb_set_t &glyphset = *c->plan->glyphset_gsub ();
const hb_map_t &glyph_map = *c->plan->glyph_map;
auto *out = c->serializer->start_embed (*this);
if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
out->format = format;
hb_sorted_vector_t<hb_codepoint_t> new_coverage;
+ hb_zip (this+coverage, alternateSet)
| hb_filter (glyphset, hb_first)
| hb_filter (subset_offset_array (c, out->alternateSet, this), hb_second)
| hb_map (hb_first)
| hb_map (glyph_map)
| hb_sink (new_coverage)
;
out->coverage.serialize_serialize (c->serializer, new_coverage.iter ());
return_trace (bool (new_coverage));
}
};
}
}
}
#endif /* OT_LAYOUT_GSUB_ALTERNATESUBSTFORMAT1_HH */

Some files were not shown because too many files have changed in this diff Show More