test-convenience failed (macOS finished first; the failure is
platform-independent) because check_escaped() calls the internal
serializer::dump_escaped() directly and then reads the output stream.
Since dump_escaped() now writes into the serializer's internal write
buffer, the bytes were still buffered and the stream was empty.
Expose flush() under JSON_PRIVATE_UNLESS_TESTED (same visibility as
dump_escaped) and flush in check_escaped() before inspecting the output.
Per-string flushing inside dump_escaped() was rejected on purpose: it
would defeat the buffering that makes object/array-heavy dumps faster.
Library behavior is unchanged (flush()'s body is identical; only its
access label moved).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XAYM1qhSA2FDaDcGfPW3fG
Signed-off-by: Niels Lohmann <mail@nlohmann.me>
Two further serialization speedups on top of the ensure_ascii=false bulk
copy, both reusing the SWAR primitives in detail/input/string_scan.hpp.
1. Internal write buffer (devirtualization). Every structural character
('{', '"', ',', ...) previously went straight to the output adapter
through a virtual call. Route all writes through put_char/put_chars
into a 1 KiB buffer that flushes in bulk; the public dump() flushes
once the top-level value is done (the recursive worker is split out as
dump_internal). Runs larger than the buffer are written straight
through, so large payloads are not copied twice. This is the dominant
cost for object/array-heavy values.
2. ensure_ascii fast path. dump_escaped previously ran the UTF-8 DFA over
every byte when escaping non-ASCII. Add find_ascii_copyable_run() (a
SWAR scan stopping at '"', '\\', < 0x20, 0x7F, and >= 0x80) so runs of
printable ASCII are bulk-copied, with the byte path handling each
escape/non-ASCII byte exactly as before.
Behavior is unchanged: dump output is byte-for-byte identical to the
previous implementation across ~20k randomized byte strings plus curated
edge cases (all escapes, control chars, 0x7F, valid multibyte,
surrogates, overlong, truncated), for object/array/pretty output, both
ensure_ascii settings, and all three error handlers, in C++11/17/20 at
-O2/-O3. New unit tests cover the buffer flush boundaries, the escape and
0x7F handling, multibyte under both settings, and invalid-UTF-8 handling.
Throughput (g++ -O3, vs the ensure_ascii=false-only baseline):
long ASCII, ensure_ascii=0 4.2x
long ASCII, ensure_ascii=1 4.1x
twitter-like objects 2.7x
dense CJK 1.8x
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XAYM1qhSA2FDaDcGfPW3fG
Signed-off-by: Niels Lohmann <mail@nlohmann.me>
When ensure_ascii is false, dump_escaped previously ran every byte of
every string and object key through the UTF-8 DFA decoder, even for the
common case of ordinary text with nothing to escape. This mirrors the
per-byte cost the parser had before the contiguous fast paths.
At a character boundary, bulk-copy the longest run of bytes that need no
escaping using string_bulk_run() - the same SWAR scanner and UTF-8 bulk
validator the lexer's contiguous path uses - and only fall back to the
byte-at-a-time DFA loop for the first byte that needs individual handling
(a quote, backslash, control character, or ill-formed/truncated UTF-8).
Because every "hard" or invalid byte is still processed by the unchanged
byte path, escaping output and error handling (including strict-mode
error 316 position and message) are byte-identical to before.
The ensure_ascii=true path is unchanged: it must escape non-ASCII and
0x7F, which string_bulk_run does not stop on, so a separate predicate
would be needed for it.
Verified byte-for-byte identical dump output against the pre-change
implementation across ~20k randomized byte strings plus curated edge
cases (all escapes, control chars, valid multibyte, surrogates,
overlong, truncated sequences) for both ensure_ascii settings and all
three error handlers, in C++11/17/20 at -O2/-O3.
Throughput (g++ -O3, ensure_ascii=false, vs pre-change):
long ASCII strings 4.2x
twitter-like objects 2.3x
dense CJK 1.4x (further headroom with JSON_USE_SIMDUTF)
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XAYM1qhSA2FDaDcGfPW3fG
Signed-off-by: Niels Lohmann <mail@nlohmann.me>
* Replace snprintf with a branch-free writer for \uXXXX escapes
dump_escaped called std::snprintf(..., "\u%04x", ...) once per escaped
code point in the string serialization hot path. snprintf re-parses
the format string and pulls in locale/printf machinery on every call,
which is far heavier than the fixed 6-/12-byte output warrants. This
is hot for any string containing control characters, and for all
non-ASCII text when ensure_ascii is set.
Replace it with write_u_escape, a small helper that writes the escape
directly into string_buffer via a nibble-to-hex lookup table, mirroring
the existing hand-rolled dump_integer fast path in the same file.
Signed-off-by: Niels Lohmann <mail@nlohmann.me>
* Fix clang-tidy avoid-c-arrays warning in write_u_escape
Use a const char* rather than a char[] lookup table, matching the
existing hex_bytes helper in the same file.
Signed-off-by: Niels Lohmann <mail@nlohmann.me>
* ♻️ adjust write_u_escape signature
Signed-off-by: Niels Lohmann <mail@nlohmann.me>
---------
Signed-off-by: Niels Lohmann <mail@nlohmann.me>
* Add versioned inline namespace
Add a versioned inline namespace to prevent ABI issues when linking code
using multiple library versions.
* Add namespace macros
* Encode ABI information in inline namespace
Add _diag suffix to inline namespace if JSON_DIAGNOSTICS is enabled, and
_ldvcmp suffix if JSON_USE_LEGACY_DISCARDED_VALUE_COMPARISON is enabled.
* Move ABI-affecting macros into abi_macros.hpp
* Move std_fs namespace definition into std_fs.hpp
* Remove std_fs namespace from unit test
* Format more files in tests directory
* Add unit tests
* Update documentation
* Fix GDB pretty printer
* fixup! Add namespace macros
* Derive ABI prefix from NLOHMANN_JSON_VERSION_*
* Add benchmark for cbor serialization and cbor binary data serialization
* Use std::vector::insert method instead of std::copy in output_vector_adapter::write_characters
This change increases a lot the performance when writing lots of binary data.
* support UBJSON-derived Binary JData (BJData) format
* fix Codacy warning
* partially fix VS compilation errors
* fix additional VS errors
* fix more VS compilation errors
* fix additional warnings and errors for clang and msvc
* add more tests to cover the new bjdata types
* add tests for optimized ndarray, improve coverage, fix clang/gcc warnings
* gcc warn useless conversion but msvc gives an error
* fix ci_test errors
* complete test coverage, fix ci_test errors
* add half precision error test
* fix No newline at end of file error by clang
* simplify endian condition, format unit-bjdata
* remove broken test due to alloc limit
* full coverage, I hope
* move bjdata new markers from default to the same level as ubjson markers
* fix ci errors, add tests for new bjdata switch structure
* make is_bjdata const after using initializer list
* remove the unwanted assert
* move is_bjdata to an optional param to write_ubjson
* pass use_bjdata via output adapter
* revert order to avoid msvc 2015 unreferenced formal param error
* update BJData Spect V1 Draft-2 URL after spec release
* amalgamate code
* code polishing following @gregmarr's feedback
* make use_bjdata a non-default parameter
* fix ci error, remove unwanted param comment
* encode and decode bjdata ndarray in jdata annotations, enable roundtrip tests
* partially fix ci errors, add tests to improve coverage
* polish patch to remove ci errors
* fix a ndarray dim vector condition
* fix clang tidy error
* add sax test cases for ndarray
* add additional sax event tests
* adjust sax event numbering
* fix sax tests
* ndarray can only be used with array containers, discard if used in object
* complete test coverage
* disable [{SHTFNZ in optimized type due to security risks in #2793 and hampered readability
* fix ci error
* move OutputIsLittleEndian from tparam to param to replace use_bjdata
* fix ci clang gcc error
* fix ci static analysis error
* update json_test_data to 3.1.0, enable file-based bjdata unit tests
* fix stack overflow error on msvc 2019 and 2022
* use https link, update sax_parse_error after rebase
* make input_format const and use initializer
* return bool for write_bjdata_ndarray
* test write_bjdata_ndarray return value as boolean
* fix ci error
* Make exception context optional
Change exception context parameter to pointer and replace context with
nullptr where appropriate.
* Support escaping other string types
* Add string concatenation function
Add variadic concat() function for concatenating char *, char, and
string types.
* Replace string concatenations using + with concat()
* Template json_pointer on string type
Change json_pointer from being templated on basic_json to being
templated on string type.
* Add unit test for #3388Closes#3388.
* Fix regression test for #2958
* Add backwards compatibility with json_pointer<basic_json>
* Update json_pointer docs
* Allow comparing different json_pointers
* Update version numbers
Niels Lohmann
Kirill Lokotkov
Charles Cabergs
NmPassTHFan
Michael Valladolid
Nebojša Cvetković
Raphael Grimm
Florian Albrechtskirchinger
Romain Reignier
Qianqian Fang
Niels Lohmann
Alex Beregszaszi
Pierre Hallot
Ferry Huberts