Niels Lohmann 4fe670b5f4 json_view: decode chains of unicode escapes in a loop (#5969) ...

After the first \uXXXX escape of a string, slow_string() now decodes the
escapes that follow it directly, while the next six bytes are \u and four hex
digits that are not a surrogate, instead of going round its main loop for each
of them (that is how Python's json.dumps writes non-ASCII text by default).
A surrogate, a bad digit, or fewer than six bytes leave the position at the
backslash: the main loop handles them as before and reports the same error at
the same position.

twitterescaped.json (simdjson-data, 549 KB, 33,036 escapes), json_document::parse(),
A/B in one binary, 6 code layouts, time after / before:

- Apple M1 Max, Apple clang 21: 0.897
- x86-64 Haswell, GCC 16.2: 0.907
- x86-64 Haswell, Clang 23: 0.910

The other files of the benchmark set (twitter, citm_catalog, canada, gsoc-2018,
apache_builds, instruments, tree-pretty, github_events, random, update-center,
semanticscholar-corpus, twitter_timeline) stay within the noise of each machine.

New test: chains of every length from 1 to 40 of escapes that give one, two,
and three bytes of UTF-8, followed by the end of the string, text, escapes,
surrogates, bad digits, and cut in every place, in keys and in values, against
json::parse().

Signed-off-by: Niels Lohmann <mail@nlohmann.me>

2026-10-31 13:54:02 +01:00

..

bjdata_round_trip.hpp

Split the slowest Valgrind tests out of unit-large_json and unit-bjdata (#5862)

2026-10-17 08:28:28 +02:00

custom_key_test.hpp

Restore v3.12.0 support for custom object key types (#5795)

2026-10-11 08:24:31 +02:00

custom_object_key_type.hpp

Restore v3.12.0 support for custom object key types (#5795)

2026-10-11 08:24:31 +02:00

float_hard_cases.hpp

Speed up the lexer: own float parser, string scan, and \u table (#5738)

2026-10-14 17:38:36 +02:00

fuzzer_common.hpp

Fuzz the streaming input adapters against the contiguous one (#5827)

2026-10-15 20:04:57 +02:00

fuzzer-driver_afl.cpp

📄 adjust year (#5044)

2026-01-01 20:00:39 +01:00

fuzzer-json_view_image.cpp

Document clang-tidy exclusions, drop unneeded ones, and clean up doc examples (#5909)

2026-10-25 12:07:51 +01:00

fuzzer-parse_bjdata.cpp

Fuzz the streaming input adapters against the contiguous one (#5827)

2026-10-15 20:04:57 +02:00

fuzzer-parse_bon8.cpp

Fuzz the streaming input adapters against the contiguous one (#5827)

2026-10-15 20:04:57 +02:00

fuzzer-parse_bson.cpp

Fuzz the streaming input adapters against the contiguous one (#5827)

2026-10-15 20:04:57 +02:00

fuzzer-parse_cbor.cpp

Fuzz the streaming input adapters against the contiguous one (#5827)

2026-10-15 20:04:57 +02:00

fuzzer-parse_json_view.cpp

Document clang-tidy exclusions, drop unneeded ones, and clean up doc examples (#5909)

2026-10-25 12:07:51 +01:00

fuzzer-parse_json.cpp

Fuzz the streaming input adapters against the contiguous one (#5827)

2026-10-15 20:04:57 +02:00

fuzzer-parse_msgpack.cpp

Fuzz the streaming input adapters against the contiguous one (#5827)

2026-10-15 20:04:57 +02:00

fuzzer-parse_ubjson.cpp

Fuzz the streaming input adapters against the contiguous one (#5827)

2026-10-15 20:04:57 +02:00

json_view_test_helpers.hpp

Write json_view's dump() without a call or conversion per token (#5633)

2026-10-21 10:21:30 +02:00

make_test_data_available.hpp

📄 adjust year (#5044)

2026-01-01 20:00:39 +01:00

round_trip_corpus.hpp

Remove test debt: dead guards, discarded results, and unreferenced files (#5732)

2026-10-01 07:33:07 +02:00

sax_countdown.hpp

Remove test debt: dead guards, discarded results, and unreferenced files (#5732)

2026-10-01 07:33:07 +02:00

sax_event_loggers.hpp

Use and extend the detail helpers to remove duplicated code (#5783)

2026-10-11 08:21:13 +02:00

skip_library_version_check.cpp

Test suite: cover untested macro configs, std::formatter branches, patch_inplace, and fix a duplicate TEST_CASE name (#5492)

2026-09-16 20:15:46 +02:00

test_utils.hpp

Reduce test suite runtime and run the Unicode tests everywhere (#5605)

2026-10-06 22:53:03 +02:00

unit-32bit.cpp

Remove test debt: dead guards, discarded results, and unreferenced files (#5732)

2026-10-01 07:33:07 +02:00

unit-algorithms.cpp

Speed up JSON pointer parsing, formatting, comparison and lookup (#5836)

2026-10-15 20:01:20 +02:00

unit-allocator.cpp

Use the with_*_t aliases in tests, examples, and docs (#5787)

2026-10-08 16:11:16 +02:00

unit-alt-string.cpp

Strong exception guarantee for operator[](json_pointer) (#5940)

2026-10-27 17:15:55 +01:00

unit-assert_macro.cpp

Fix lint debt: enum-macro NOLINTs, doctest as SYSTEM, no-op analyzer (#5737)

2026-10-01 07:37:47 +02:00

unit-binary_formats.cpp

Write doubles with the shortest digits (Żmij), digits in registers (#5634)

2026-10-14 17:38:37 +02:00

unit-binary_utf8_error_handler.cpp

Keep converted object keys alive while writing UBJSON and BJData (#5791)

2026-10-09 17:05:36 +02:00

unit-binary_utf8_strict.cpp

Fix CI on develop after merging the ready-to-merge PRs (#5754)

2026-10-04 17:18:19 +02:00

unit-binary_writer_sinks.cpp

Write binary formats through reserved sink space and number runs (#5885)

2026-10-23 06:52:58 +02:00

unit-bjdata_round_trip_draft2.cpp

Split the slowest Valgrind tests out of unit-large_json and unit-bjdata (#5862)

2026-10-17 08:28:28 +02:00

unit-bjdata_round_trip_draft3.cpp

Split the slowest Valgrind tests out of unit-large_json and unit-bjdata (#5862)

2026-10-17 08:28:28 +02:00

unit-bjdata.cpp

Read BJData ND-arrays of float16, restrict ND-array char to 0..127, correct BJData docs (#5931)

2026-10-26 18:25:46 +01:00

unit-bon8.cpp

Throw type_error.321 for discarded values in to_bon8 (#5802)

2026-10-11 08:22:00 +02:00

unit-brace-init-copy-semantics.cpp

Copy a pair-shaped array value under JSON_BRACE_INIT_COPY_SEMANTICS (#5701)

2026-10-07 07:39:05 +02:00

unit-bson.cpp

Reconcile -Weffc++ and readability-redundant-member-init in unit-bson.cpp (#5953)

2026-10-28 07:11:04 +01:00

unit-byte_container_with_subtype.cpp

Test suite: cover untested macro configs, std::formatter branches, patch_inplace, and fix a duplicate TEST_CASE name (#5492)

2026-09-16 20:15:46 +02:00

unit-capacity.cpp

Fix lint debt: enum-macro NOLINTs, doctest as SYSTEM, no-op analyzer (#5737)

2026-10-01 07:37:47 +02:00

unit-cbor.cpp

Skip CBOR tags in front of map keys, validate indefinite text chunks, fix CBOR docs (#5927)

2026-10-26 18:25:44 +01:00

unit-class_const_iterator.cpp

Add tests for uncovered code paths (#5581)

2026-09-27 14:19:22 +02:00

unit-class_iterator.cpp

📄 adjust year (#5044)

2026-01-01 20:00:39 +01:00

unit-class_lexer.cpp

Consolidate duplicated detail helpers and bodies (#5891) (#5901)

2026-10-25 06:17:50 +01:00

unit-class_parser.cpp

Fix implicit-conversion sanitizer failure in unit-class_parser.cpp (#5960)

2026-10-30 09:56:06 +01:00

unit-comparison.cpp

Make the value lifecycle inline-friendly (#5840)

2026-10-15 20:01:21 +02:00

unit-concepts.cpp

📄 adjust year (#5044)

2026-01-01 20:00:39 +01:00

unit-constructor1.cpp

from_json: number-first dispatch for arithmetic types, loop for std::array (#5902)

2026-10-25 06:17:51 +01:00

unit-constructor2.cpp

Fix CI failures on develop from the API performance stack (#5860)

2026-10-16 19:16:21 +02:00

unit-convenience.cpp

Stop dump() from heap-allocating its output adapter per call (#5449)

2026-09-11 08:22:50 +02:00

unit-conversions1.cpp

Fix CI on develop after merging the ready-to-merge PRs (#5754)

2026-10-04 17:18:19 +02:00

unit-conversions2.cpp

Fix CI on develop after #5600, #5607, and #5755 (#5764)

2026-10-06 07:22:55 +02:00

unit-custom-array-type.cpp

Fix silent wrong results for custom number types (#5805)

2026-10-21 08:36:31 +00:00

unit-custom-base-class.cpp

Use the with_*_t aliases in the remaining tests and docs (#5790)

2026-10-09 16:56:14 +02:00

unit-custom-binary-type.cpp

Derive the lexer position lazily for seekable input (#5809)

2026-10-13 17:16:52 +02:00

unit-custom-key-type-c-str.cpp

Restore v3.12.0 support for custom object key types (#5795)

2026-10-11 08:24:31 +02:00

unit-custom-key-type-explicit.cpp

Restore v3.12.0 support for custom object key types (#5795)

2026-10-11 08:24:31 +02:00

unit-custom-key-type-full.cpp

Restore v3.12.0 support for custom object key types (#5795)

2026-10-11 08:24:31 +02:00

unit-custom-key-type-no-eq.cpp

Restore v3.12.0 support for custom object key types (#5795)

2026-10-11 08:24:31 +02:00

unit-custom-key-type-to-json.cpp

Restore v3.12.0 support for custom object key types (#5795)

2026-10-11 08:24:31 +02:00

unit-custom-number-types.cpp

Write binary formats through reserved sink space and number runs (#5885)

2026-10-23 06:52:58 +02:00

unit-custom-object-key-type.cpp

Move custom object key tests out of unit-cbor.cpp and unit-msgpack.cpp (#5798)

2026-10-10 00:05:02 +02:00

unit-custom-object-type.cpp

Use the with_*_t aliases in tests, examples, and docs (#5787)

2026-10-08 16:11:16 +02:00

unit-delete_deprecated_functions.cpp

Fix CI on develop after #5600, #5607, and #5755 (#5764)

2026-10-06 07:22:55 +02:00

unit-deserialization.cpp

Skip the indentation after a newline 8 bytes at a time (#5810)

2026-10-13 17:16:52 +02:00

unit-diagnostic-positions.cpp

Document clang-tidy exclusions, drop unneeded ones, and clean up doc examples (#5909)

2026-10-25 12:07:51 +01:00

unit-diagnostics-optimized.cpp

Fix CI on develop after #5585 (#5779)

2026-10-07 16:37:19 +02:00

unit-diagnostics.cpp

Fix CI failures on develop from the API performance stack (#5860)

2026-10-16 19:16:21 +02:00

unit-disable-tuple-reference-conversion.cpp

Add JSON_DISABLE_TUPLE_REFERENCE_CONVERSION to fix std::tuple conversions (#5598)

2026-09-30 20:53:56 +02:00

unit-disabled_exceptions.cpp

Document clang-tidy exclusions, drop unneeded ones, and clean up doc examples (#5909)

2026-10-25 12:07:51 +01:00

unit-element_access1.cpp

Throw std::length_error for operator[](SIZE_MAX) instead of corrupting the array (#5687)

2026-09-30 20:07:06 +02:00

unit-element_access2.cpp

Document clang-tidy exclusions, drop unneeded ones, and clean up doc examples (#5909)

2026-10-25 12:07:51 +01:00

unit-enum_keyed_maps_default.cpp

Store maps with enum keys as objects (opt-in) (#5600)

2026-10-04 17:54:05 +02:00

unit-enum_keyed_maps.cpp

Store maps with enum keys as objects (opt-in) (#5600)

2026-10-04 17:54:05 +02:00

unit-explicit_instantiation.cpp

Document clang-tidy exclusions, drop unneeded ones, and clean up doc examples (#5909)

2026-10-25 12:07:51 +01:00

unit-file_input.cpp

Make parsing from std::istream and FILE* fast (#5828)

2026-10-19 22:00:58 +02:00

unit-format-as.cpp

Document clang-tidy exclusions, drop unneeded ones, and clean up doc examples (#5909)

2026-10-25 12:07:51 +01:00

unit-guard-page.cpp

Test that the contiguous fast paths never read outside the input (#5826)

2026-10-20 22:24:09 +02:00

unit-hash.cpp

Document clang-tidy exclusions, drop unneeded ones, and clean up doc examples (#5909)

2026-10-25 12:07:51 +01:00

unit-inspection.cpp

📄 adjust year (#5044)

2026-01-01 20:00:39 +01:00

unit-items-cpp17.cpp

Reduce test-suite compile time: extract tiny per-standard test content; drop redundant legacy-comparison CI job (#5481)

2026-09-09 09:52:58 +02:00

unit-items.cpp

Add JSON_DELETE_DEPRECATED_FUNCTIONS to delete the deprecated functions (#5755)

2026-10-04 22:11:34 +02:00

unit-iterators1.cpp

Fix CI failures on develop from the API performance stack (#5860)

2026-10-16 19:16:21 +02:00

unit-iterators2.cpp

Remove test debt: dead guards, discarded results, and unreferenced files (#5732)

2026-10-01 07:33:07 +02:00

unit-iterators3.cpp

📄 adjust year (#5044)

2026-01-01 20:00:39 +01:00

unit-json_patch.cpp

Fix self-aliasing in json_pointer::operator/= and patch_inplace (#5936)

2026-10-28 20:23:55 +01:00

unit-json_pointer.cpp

Fix self-aliasing in json_pointer::operator/= and patch_inplace (#5936)

2026-10-28 20:23:55 +01:00

unit-json_view_builder.cpp

json_view: decode chains of unicode escapes in a loop (#5969)

2026-10-31 13:54:02 +01:00

unit-json_view_compare.cpp

json_view: comparison without temporaries and a faster image float check (#5890)

2026-10-30 15:19:42 +01:00

unit-json_view_dump.cpp

json_view: comparison without temporaries and a faster image float check (#5890)

2026-10-30 15:19:42 +01:00

unit-json_view_edit_values.cpp

Fix the VS2015 build of unit-json_view_edit_values.cpp (#5947)

2026-10-28 19:46:28 +01:00

unit-json_view_edit.cpp

json_view: check_utf8 skips well-formed runs with the parser's string scan (#5926)

2026-10-30 15:17:55 +01:00

unit-json_view_from_json.cpp

json_view: source positions in materialize() with JSON_DIAGNOSTIC_POSITIONS (#5945)

2026-10-28 20:24:49 +01:00

unit-json_view_image.cpp

json_view: comparison without temporaries and a faster image float check (#5890)

2026-10-30 15:19:42 +01:00

unit-json_view_macros.cpp

Add json_document and json_view: node index, parser, and document (#5620)

2026-10-21 10:21:27 +02:00

unit-json_view_positions.cpp

Fix the VS2015 build of unit-json_view_positions.cpp (#5957)

2026-10-28 22:08:55 +01:00

unit-json_view.cpp

json_view: source positions in materialize() with JSON_DIAGNOSTIC_POSITIONS (#5945)

2026-10-28 20:24:49 +01:00

unit-large_json_binary_writers.cpp

Split the slowest Valgrind tests out of unit-large_json and unit-bjdata (#5862)

2026-10-17 08:28:28 +02:00

unit-large_json_conversions.cpp

Split the slowest Valgrind tests out of unit-large_json and unit-bjdata (#5862)

2026-10-17 08:28:28 +02:00

unit-large_json.cpp

Split the slowest Valgrind tests out of unit-large_json and unit-bjdata (#5862)

2026-10-17 08:28:28 +02:00

unit-locale-cpp.cpp

Speed up the lexer: own float parser, string scan, and \u table (#5738)

2026-10-14 17:38:36 +02:00

unit-merge_diff.cpp

support JSON Merge Patch (RFC 7396) diff creation (#4965)

2026-10-12 09:33:43 +02:00

unit-merge_patch.cpp

docs: state merge_patch complexity and RFC 7396 limitations; test non-object, discarded, binary cases (#5938)

2026-10-26 17:11:55 +01:00

unit-meta.cpp

📄 adjust year (#5044)

2026-01-01 20:00:39 +01:00

unit-modifiers.cpp

Fix CI on develop after merging the ready-to-merge PRs (#5754)

2026-10-04 17:18:19 +02:00

unit-msgpack-cpp17.cpp

Reduce test suite runtime and run the Unicode tests everywhere (#5605)

2026-10-06 22:53:03 +02:00

unit-msgpack-custom-types.cpp

Fix silent wrong results for custom number types (#5805)

2026-10-21 08:36:31 +00:00

unit-msgpack.cpp

Fix silent wrong results for custom number types (#5805)

2026-10-21 08:36:31 +00:00

unit-no_automatic_udls.cpp

Document clang-tidy exclusions, drop unneeded ones, and clean up doc examples (#5909)

2026-10-25 12:07:51 +01:00

unit-no_io_and_user_exceptions.cpp

Document clang-tidy exclusions, drop unneeded ones, and clean up doc examples (#5909)

2026-10-25 12:07:51 +01:00

unit-no-macro-leak.cpp

Undefine the four JSON_HEDLEY_* macros that leak after including json.hpp (#5475)

2026-09-09 09:48:17 +02:00

unit-no-mem-leak-on-adl-serialize.cpp

Test suite: cover untested macro configs, std::formatter branches, patch_inplace, and fix a duplicate TEST_CASE name (#5492)

2026-09-16 20:15:46 +02:00

unit-noexcept.cpp

Make comparisons with scalars noexcept only when the conversion is (#5751)

2026-10-02 17:25:17 +02:00

unit-ordered_json2.cpp

Use the with_*_t aliases in the remaining tests and docs (#5790)

2026-10-09 16:56:14 +02:00

unit-ordered_json.cpp

Make the DOM parser cheaper per container and per member (#5882)

2026-10-24 07:41:21 +02:00

unit-ordered_map.cpp

Fix CI on develop after merging the ready-to-merge PRs (#5754)

2026-10-04 17:18:19 +02:00

unit-pointer_access.cpp

Fix remaining red CI on develop: Infer null dereferences, clang-tidy redundant init (#5943)

2026-10-27 14:31:31 +01:00

unit-precise-stream-position.cpp

Make parsing from std::istream and FILE* fast (#5828)

2026-10-19 22:00:58 +02:00

unit-qt_keywords.cpp

Do not use Qt's keyword macros as identifiers (#5866)

2026-10-19 22:00:31 +02:00

unit-readme.cpp

Remove test debt: dead guards, discarded results, and unreferenced files (#5732)

2026-10-01 07:33:07 +02:00

unit-reference_access.cpp

📄 adjust year (#5044)

2026-01-01 20:00:39 +01:00

unit-regression1.cpp

Fix regression1 operator[](json_pointer) test under JSON_NO_THREAD_LOCAL (#5952)

2026-10-27 20:31:53 +01:00

unit-regression2.cpp

Document clang-tidy exclusions, drop unneeded ones, and clean up doc examples (#5909)

2026-10-25 12:07:51 +01:00

unit-regression3.cpp

Document clang-tidy exclusions, drop unneeded ones, and clean up doc examples (#5909)

2026-10-25 12:07:51 +01:00

unit-same-number-types.cpp

Allow the same type for number_integer_t and number_unsigned_t (#5815)

2026-10-15 20:04:42 +02:00

unit-serialization.cpp

Write strings, scalars and indentation in dump() with fewer calls (#5883)

2026-10-23 06:52:41 +02:00

unit-std-format.cpp

Document clang-tidy exclusions, drop unneeded ones, and clean up doc examples (#5909)

2026-10-25 12:07:51 +01:00

unit-stl_helpers.cpp

Fix develop CI: bit_cast ADL ambiguity and artifact-less comment runs (#5894)

2026-10-22 20:22:53 +02:00

unit-stream-bulk-scan.cpp

Faster std::istream input: inline get-area path, binary reads, unbuffered streambufs (#5884)

2026-10-23 06:48:00 +02:00

unit-testsuites.cpp

Fix CI failures on develop from the API performance stack (#5860)

2026-10-16 19:16:21 +02:00

unit-throw_messages.cpp

Fix CI failures on develop from the API performance stack (#5860)

2026-10-16 19:16:21 +02:00

unit-to_chars.cpp

Use standard library helpers where available, with C++11 fallbacks (#5876)

2026-10-21 20:14:59 +02:00

unit-to_json-rvalue.cpp

Fix compiler warnings on develop from the API performance stack (#5858)

2026-10-15 22:42:46 +02:00

unit-type_traits.cpp

Remove dead metaprogramming and duplicated code in traits and pointers (#5728)

2026-10-04 11:50:42 +02:00

unit-ubjson.cpp

Validate the UBJSON/BJData container type marker for count 0; fix UBJSON docs (#5929)

2026-10-26 18:25:45 +01:00

unit-udl.cpp

📄 adjust year (#5044)

2026-01-01 20:00:39 +01:00

unit-udt_macro.cpp

Support zero-member types in NLOHMANN_DEFINE_TYPE_* macros (#4041) (#5272)

2026-09-25 20:44:30 +02:00

unit-udt.cpp

Add with_*_t alias templates to create basic_json types with changed template parameters (#5758)

2026-10-06 22:33:29 +02:00

unit-unicode.cpp

Add tests for error_handler_t::keep in dump() (#4555)

2026-10-07 19:17:57 +02:00

unit-user_defined_input.cpp

Fix lint debt: enum-macro NOLINTs, doctest as SYSTEM, no-op analyzer (#5737)

2026-10-01 07:37:47 +02:00

unit-windows_h.cpp

📄 adjust year (#5044)

2026-01-01 20:00:39 +01:00

unit-wstring.cpp

Fix CI on develop after merging the ready-to-merge PRs (#5754)

2026-10-04 17:18:19 +02:00

unit.cpp

Fix CI jobs that check less than they claim; move arm64 to GitHub (#5733)

2026-09-30 22:49:24 +02:00

value_in_range_of_test.hpp

Fix lint debt: enum-macro NOLINTs, doctest as SYSTEM, no-op analyzer (#5737)

2026-10-01 07:37:47 +02:00