amorfo77/json
1
0
Fork 0
mirror of https://github.com/nlohmann/json.git synced 2026-10-03 13:10:33 +00:00
Code Issues Packages Projects Releases Wiki Activity
5,246 Commits 76 Branches 57 Tags
134b2f0efe973e62251b41aa0ed0bd0d8a8eede9
Commit Graph
30 Commits
Author SHA1 Message Date
Niels Lohmann 371a8a3d9f Merge branch 'json-view/08-view-builder' into json-view/10-view-document 
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

# Conflicts:
#	Makefile
#	cmake/ci.cmake
 2026-10-01 10:19:49 +02:00
Niels Lohmann cb51f80e34 Merge branch 'json-view/04-unicode-escapes' into json-view/08-view-builder 
Signed-off-by: Niels Lohmann <mail@nlohmann.me>
 2026-10-01 10:18:33 +02:00
Niels Lohmann e5a89d671f Fix lint debt: enum-macro NOLINTs, doctest as SYSTEM, no-op analyzer (#5737) 
* Drop stale LCOV_EXCL_LINE from the json_pointer out_of_range.410 throw

The comment said the size_type overflow check in array_index() is only
triggered on special platforms like 32-bit, and the throw was excluded
from coverage. On 64-bit platforms the check is true for SIZE_MAX
itself, and unit-json_pointer.cpp has asserted that case four times
since #5395, so the line is executed in the coverage job. Reword the
comment and remove the exclusion marker so the coverage report notices
if the tests stop reaching it.

Part of #5725

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

* Name all three C-array check aliases in the enum-macro NOLINTs

NLOHMANN_JSON_SERIALIZE_ENUM(_STRICT) suppressed the c-array warning
under modernize-avoid-c-arrays only, but clang-tidy emits the same
diagnostic under the aliases cppcoreguidelines-avoid-c-arrays and
hicpp-avoid-c-arrays too. Any user running those checks got a false
positive at every macro expansion, and our own tests needed a local
NOLINT at each call site to work around it.

Name all three aliases in the four macro comments instead, and drop
the now-redundant c-array names from the five test call-site NOLINTs.
Comment-only change; behavior, the public API, and the ABI do not
change. Ran make amalgamate.

Part of #5725

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

* Include doctest as a SYSTEM directory instead of disabling warnings for all tests

test_main added -Wno-deprecated and -Wno-float-equal as PUBLIC compile
options for every non-MSVC compiler, so they were applied to every
translation unit, library headers included, and silenced the CI
warnings meant to check the library's own -Wfloat-equal pragmas. The
only code that actually needed the suppression was the vendored
doctest.h, which was included as a normal (non-SYSTEM) directory.

Include thirdparty/doctest as SYSTEM for test_main, matching what
tests/abi/CMakeLists.txt already does, and drop the two suppressions
from both targets. Verified locally that unit-comparison,
unit-conversions and unit-constructor1 compile clean with
-Werror -Weverything and doctest as -isystem, and that CMake still
configures with JSON_BuildTests=ON.

Part of #5725

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

* Remove the no-op ci_clang_analyze target

ci_clang_analyze configured the build with the real compiler and only
then wrapped ninja with scan-build. scan-build intercepts compiles by
overriding CC/CXX, but build.ninja already had the compiler path baked
in from the configure step, so every run bypassed the analyzer: CI
logs show "No bugs found" after a normal build, never an analysis.
The job also used Debian's frozen clang-tools-14 rather than the
image's own clang, and CLANG_ANALYZER_CHECKS still named three
valist.* checkers that current clang merged into security.VAList.

ci_clang_tidy already runs every clang-analyzer-* check (via
.clang-tidy's "Checks: '*'") with warnings as errors, so nothing is
lost by removing the dead job. Delete ci_clang_analyze,
CLANG_ANALYZER_CHECKS and the SCAN_BUILD_TOOL lookup from
cmake/ci.cmake, drop it from the ubuntu.yml ci_static_analysis_clang
matrix, and drop the now-unused clang-tools apt package (iwyu stays
for ci_single_binaries). Reword quality_assurance.md and
assurance_case.md, which described the dead job as a working control,
to say the Clang Static Analyzer checks run through clang-tidy.

Verified that `cmake -DJSON_CI=ON` still configures cleanly and that
ci_clang_analyze no longer appears in the generated build or in any
CMake/workflow file.

Part of #5725

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

* Re-enable portability-template-virtual-member-function; remove redundant forwards

.clang-tidy disabled three checks "to get the CI going" (#4489,
2024-11-13): portability-template-virtual-member-function,
bugprone-use-after-move and its alias hicpp-invalid-access-moved.

portability-template-virtual-member-function only flagged
output_stream_adapter::write_character/write_characters; annotate
both with NOLINT and re-enable the check.

bugprone-use-after-move flagged several double forwards that have no
effect at runtime:
- from_json.hpp calls std::forward<BasicJsonType>(j).at(Idx) inside
  pack expansions; at() has no ref-qualified overloads and always
  returns an lvalue reference, so the forward is a no-op. Replace with
  plain j.at(Idx) in all four places.
- the move constructor forwards the whole object to its base class
  and then reads other's members. That is item 9 of #5724 (together
  with its cppcheck suppressions) and is left to that change.
- input_adapters.hpp forwards the container twice on purpose, so the
  begin/end iterator types match adapter_type; annotate with NOLINT
  and a comment instead of changing behavior.

The check still flags the move constructor (see above) and two sites
in at(KeyType&&) (both overloads, json.hpp, in the throw's
string_t(std::forward<KeyType>(key)) after
find(std::forward<KeyType>(key))). Open PR #5689 rewrites that hunk,
so bugprone-use-after-move (and hicpp-invalid-access-moved)
stay disabled for now, with a comment explaining why; re-enable them
once #5689 and the #5724 move-constructor change have landed.

Also resolve the portability-avoid-pragma-once TODO: single_include
never has #pragma once (amalgamate.py strips it) and every supported
compiler accepts it in include/, so keep it disabled with an
explanatory comment instead of a TODO. Fix the stale "json.hpp,
around line 1265" comment in unit-class_parser.cpp, which now points
at the move constructor's actual line.

Behavior, the public API and the ABI do not change. Verified with
clang-tidy 22.1.8 that portability-template-virtual-member-function
now reports nothing, that bugprone-use-after-move/
hicpp-invalid-access-moved report only the known at(KeyType&&) and
move-constructor sites, and that unit-custom-base-class, unit-constructor1,
unit-conversions, unit-element_access2, unit-class_parser and
unit-diagnostic-positions (JSON_DIAGNOSTIC_POSITIONS=1) compile
under ASan/UBSan and pass with the same assertion counts as before.
Ran make amalgamate.

Part of #5725

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

* Fix stale and malformed NOLINT comments

json_sax.hpp named "-warnings-as-errors" in the NOLINT list on the two
JSON_ASSERT(false) lines; that is the suffix clang-tidy appends to a
diagnostic tag under WarningsAsErrors, not a check name, and every
other JSON_ASSERT(false) omits it.

unit-capacity.cpp carried 30 "// NOLINT(misc-const-correctness)"
comments on "json j = ...;" declarations that are all used with
non-const members afterwards, so the check has nothing to report
there.

unit-constructor2.cpp used a blanket "// NOLINT: access after move is
OK here" on a use-after-move that hides every check on the line;
naming bugprone-use-after-move and hicpp-invalid-access-moved keeps
the intent once those checks are re-enabled (#5724).

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

#5725 item 10

* Remove stale .clang-tidy entries

-google-runtime-references disabled a check that neither clang-tidy
22.1.8 nor 23.1.2 lists under --list-checks -checks='*'; it was
removed upstream. The commented-out HeaderFilterRegex line has been
unused since the active HeaderFilterRegex was introduced in #2561
(2021).

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

#5725 item 11

* Remove the GCC C++20 -Wignored-attributes pragma in json.hpp

The pragma (added in #5164) claimed to work around the C++ modules
redefinition errors of #5103, but #5103 is about hard errors (e.g.
"redefinition of std::__is_constant_evaluated()", conflicting
std::integral_constant) that ignoring a warning cannot suppress; they
are traced to GCC PR 124430 and reproduce with <map> or <string>
instead of json.hpp too. A GCC 16.2 -std=gnu++20 -fmodules build
following #5103's repro steps still fails with the pragma in place,
and a build of all test TUs with GCC_CXXFLAGS (which enable
-Wignored-attributes) and the pragma removed produces no such
warning. The block only hid a warning class from GCC C++20 users
while suggesting #5103 was handled.

Overlaps #5610, whose hunks touch the closing half of this pragma to
insert the json_literals.hpp include.

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

#5725 item 9

* Fix stale doxygen comments hidden by the -Wdocumentation pragma

macro_scope.hpp ignores -Wdocumentation and -Wdocumentation-unknown-command
for the whole library, which also hides genuine documentation mistakes:

- detail::unescape() documented "@return unescaped string" but returns
  void and unescapes its argument in place; reworded to
  "@param[in,out] s string to unescape in place" and dropped the
  bogus @return.
- basic_json::get()'s copy-conversion overload wrote "converted to
  @tparam ValueType" inside @return, which Doxygen and Clang parse as
  a second, malformed @tparam; changed to "@a ValueType", matching the
  two other get() overloads a few lines above that already use it.

This narrows the gap the -Wdocumentation pragma needs to cover; fully
replacing the Doxygen-only commands it also hides (item 2c) is left
for after #5267.

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

#5725 item 2

* Fix -Wextra-semi-stmt at its actual source, not assert()

clang_flags.cmake blamed the global -Wno-extra-semi-stmt on assert(),
but assert() expands to an expression under glibc and libc++ and does
not trigger this warning. unit-assert_macro.cpp overrides JSON_ASSERT
with "{if (!(x)) ++assert_counter; }", a bare block followed by a
semicolon at every JSON_ASSERT(...) call site in the library; that
was the actual source of 151 of the 208 -Wextra-semi-stmt sites found
in a Clang 22 -Weverything sweep of the test suite with the flag
removed. Switched to the standard do/while(false) macro idiom, which
does not expand to a statement-plus-semicolon, and corrected the
comment to name the remaining source instead: vendored Doctest's
CAPTURE(x) shim, which already ends in a semicolon.

Verified with clang++ -Wextra-semi-stmt (plus the file's other CI
ignores) that unit-assert_macro.cpp now compiles without any
-Wextra-semi-stmt diagnostic.

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

#5725 item 8 (step 1 of 2; step 2 covers the CAPTURE() call sites)

* Drop the redundant semicolon from CAPTURE() call sites; remove -Wno-extra-semi-stmt

doctest_compatibility.h defines CAPTURE(x) as DOCTEST_CAPTURE(x); (with
a trailing semicolon baked into the macro), specifically so call sites
do not need to add one themselves; most of the ~267 call sites already
follow that convention. The remaining 64 call sites across 20 files
wrote "CAPTURE(x);" anyway, turning into a statement plus an empty
statement and triggering -Wextra-semi-stmt. Dropped the redundant
semicolon at each of those sites.

With item 6 having already made vendored Doctest a SYSTEM include, and
this the last known source of -Wextra-semi-stmt findings, removed the
flag from clang_flags.cmake entirely.

Verified with clang++ -Wextra-semi-stmt (plus the file's other CI
ignores) that all 20 touched files, plus a file with no CAPTURE() use
(unit-json_pointer.cpp), compile without any -Wextra-semi-stmt
diagnostic.

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

#5725 item 8 (step 2 of 2)

* Switch ci_static_analysis_clang off the frozen LLVM 22 dev image

ubuntu.yml pinned the clang-tidy/clang-tidy-sanitizer/single-binaries
job to silkeh/clang:dev, a tag last pushed 2026-02-18 that reports
"clang version 22.0.0 (...+20251015...)", a pre-release snapshot from
before the LLVM 22 release; the maintainer now updates dev-unstable,
22, and latest instead. Switched to silkeh/clang:22, matching the
other clang jobs on :latest.

Verified with clang-tidy 22.1.8 (the image's actual version) against
this repository's .clang-tidy and library headers what the release
image newly reports compared to :dev:

- readability-redundant-typename fires at ~250 sites across the
  _cpp20-relevant conversion/to_chars headers; the library targets
  C++11 and keeps the typenames, so the check is disabled in
  .clang-tidy, matching how the file already handles checks that
  don't fit a C++11 codebase.
- misc-anonymous-namespace-in-header fires on the two anonymous
  namespaces in from_json.hpp and to_json.hpp; added the alias to
  their existing NOLINT (cert-dcl59-cpp, fuchsia-header-anon-namespaces,
  google-build-namespaces).
- bugprone-std-namespace-modification fires on every addition to
  namespace std: the std::hash, std::formatter and std::swap
  overloads in json.hpp, and the std::tuple_size/std::tuple_element
  specializations in iteration_proxy.hpp (this last file is not named
  in #5725's item 5, found by actually running clang-tidy 22.1.8
  against the current tree). All six are legal, deliberate additions
  to namespace std (explicit/partial specializations of std types, or
  the pre-C++20 std::swap overload); annotated each with the check
  name next to its existing cert-dcl58-cpp NOLINT.
- modernize-avoid-c-style-cast reported nothing new.

Also added clang++-22/21, clang-tidy-22/21, g++-16 and gcov-16 to the
find_program search lists in ci.cmake so a local "maximal warnings"
configure prefers the current toolchain version over an older one on
PATH.

#5725 item 5

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

* Regenerate cmake/gcc_flags.cmake for GCC 16.2.0

GCC_CXXFLAGS was generated for GCC 15.1.0, but ci_test_gcc and
ci_test_gcc_cxx{11..26} now run in gcc:latest, currently GCC 16.2.0,
so the "maximal warnings" job was missing warnings introduced since
15.1.0 while carrying entries GCC 16 treats as duplicates or no-ops.

Regenerated with https://github.com/nlohmann/gcc_flags (patched
locally to not crash on an option whose "-x c++ <opt> -" probe fails
before it reads stdin, e.g. -Wabi=; the tool otherwise raises
BrokenPipeError instead of recording the option as an error) run
against g++ 16.2.0 in the official gcc:16 Docker image, keeping the
documented -Wno-* exclusions and the same alphabetical placement
scheme as before.

Also added three GCC 16 warnings the generator cannot discover on its
own because it only probes value ranges/lists it finds in the -Q
option name itself, not in the enum choices --help=warnings documents
separately:

- -Wbidi-chars=any, -Wleading-whitespace=spaces: manually verified
  these compile cleanly with g++ 16.2.0.
- -Wstrict-flex-arrays: deliberately NOT added, unlike the other two.
  Without -fstrict-flex-arrays (which the library does not enable, as
  it would change codegen for flexible array members), GCC prints
  "'-Wstrict-flex-arrays' is ignored when '-fstrict-flex-arrays' is
  not present" on every translation unit, and under our -Werror that
  note itself aborts the build. This differs from the harmless
  no-op warnings already kept in the file (-Whsa, -Wsynth,
  -Wunreachable-code, -Wunsafe-loop-optimizations), which emit
  nothing; #5725 item 7 named -Wstrict-flex-arrays as one of the
  flags GCC 16 adds, but did not anticipate this failure mode.

Verified: compiled the library header and a representative set of
test translation units (including ones touched by items 1, 3, 8, 9,
10 of this issue) with the regenerated GCC_CXXFLAGS plus -Werror
under g++ 16.2.0 at -std=c++11 through -std=c++26, with zero warnings;
ran the full local test suite (129/129 passing, unrelated to this
compiler) as a regression check. CI must still confirm the actual
ci_test_gcc / ci_test_standards_gcc targets end to end, since this was
verified with direct g++ invocations rather than through the CMake/
CXXFLAGS environment-variable plumbing in ci.cmake.

#5725 item 7

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

* Avoid std::basic_string<CharType> for non-character output_adapter CharType

output_adapter<CharType, StringType> defaulted StringType to
std::basic_string<CharType>, and (with JSON_NO_IO undefined) always
declared a std::basic_ostream<CharType>&-taking constructor. For
CharType with no non-deprecated std::char_traits specialization (only
std::uint8_t is ever used this way, by the binary writers), simply
naming either type - as an unused default template argument, or as an
unused, never-called constructor's parameter type - instantiates
std::char_traits<CharType> merely to name it, which some standard
libraries mark deprecated: with the library-wide -Wdocumentation
pragma (item 2's other half, left for a later commit) temporarily
removed, an Apple clang 21 / libc++ TU calling json::to_cbor(j, vec)
with std::vector<std::uint8_t>& got one -Wdeprecated-declarations
warning per binary writer at the old output_adapters.hpp:193.

Replaced the eager std::basic_string<CharType> / std::basic_ostream
<CharType> defaults with a bool-tagged partial specialization (not
std::conditional, which requires naming both branches' types up
front regardless of which is selected, reproducing the same warning)
that only ever names std::basic_string<CharType> / std::basic_ostream
<CharType> when CharType is actually one of char, wchar_t, char16_t,
char32_t, or (with __cpp_lib_char8_t) char8_t. For any other
CharType, output_adapter's StringType and ostream-constructor
parameter fall back to two distinct empty placeholder types, kept
distinct so the two constructor overloads do not collide into a
single redeclaration.

Public API / behavior: passing a std::basic_string<std::uint8_t>& or
std::basic_ostream<std::uint8_t>& directly to a binary writer's
output_adapter now fails to compile instead of compiling with a
deprecation warning; this was neither documented nor tested. All
documented uses (std::vector<CharType>, std::basic_ostream<CharType>
and StringType for character CharType) are unaffected.

Verified with Apple clang 21 / libc++, with the two -Wdocumentation*
"ignored" pragma lines in macro_scope.hpp temporarily removed and
-std=c++11/c++20 plus the project's -Weverything flag set: calling
to_cbor/to_msgpack/to_ubjson/to_bjdata/to_bson/to_bon8 on a
std::vector<std::uint8_t> now produces no char_traits<unsigned char>
(or any other) deprecation warning, while the char-based string- and
ostream-adapter paths, and a to_cbor/from_cbor round trip, still
compile and run correctly; also verified with GCC 16.2.0. Ran the
full local test suite, including the binary-format unit tests
(unit-cbor, unit-msgpack, unit-ubjson, unit-bjdata, unit-bson,
unit-bon8, unit-binary_writer_sinks, unit-binary_formats,
unit-custom-binary-type): 129/129 passing.

#5725 item 2 (step a)

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

* Remove the library-wide -Wdocumentation pragma; fix what it hid

macro_scope.hpp / macro_unscope.hpp pushed and popped a Clang
diagnostic region over the entire library that ignored -Wdocumentation
and -Wdocumentation-unknown-command. Removed both pragmas and fixed
every finding a full -Wdocumentation (which implies
-Wdocumentation-unknown-command and -Wdocumentation-deprecated-sync)
build reports, so the library now compiles clean under Clang's
documentation checks without a blanket suppression. Overlaps #5267,
which is still open and edits a nearby doc block (json.hpp's
get()/get_impl() @return, already fixed in the item 2 step (b) commit
of this branch); this commit does not touch that block again.

Unknown Doxygen alias commands (Doxyfile removed in #3071, so these
were never rendered by anything) rewritten as plain prose, keeping the
same information:
- @requirement REQ-JSON-01 / REQ-JSON-02 (iter_impl.hpp,
  json_reverse_iterator.hpp): now "This class satisfies the following
  concept requirements (REQ-JSON-0N):".
- @liveexample{prose,example-id} (three sites in json.hpp): kept the
  prose, dropped the command wrapper and the trailing example-id
  (docs/mkdocs/docs/examples/*.cpp still exist and are used directly
  by the rendered docs, not through this in-header alias) and
  unescaped the "\," commas that were only needed for the old alias's
  comma-separated argument syntax.
- @complexity X (json.hpp x4, json_pointer.hpp x2, serializer.hpp x1):
  now "Complexity: X".

Backslash sequences Clang's comment lexer tried to parse as commands,
escaped to render as literal backslashes:
- lexer.hpp get_codepoint(): two `\u` occurrences.
- binary_reader.hpp get_bson_cstr() / get_bson_cstr_bulk(): two
  `\x00` occurrences.
- serializer.hpp: three `\uXXXX` occurrences (constructor @param,
  append_codepoint_to_string_buffer() @brief, and the ensure_ascii
  member comment).

One finding remained after all of the above: Clang reports
"declaration is marked with '@deprecated' command but does not have a
deprecation attribute" on the deprecated sax_parse(span_input_adapter&&, ...)
overload, even though JSON_HEDLEY_DEPRECATED_FOR does expand to
__attribute__((deprecated(...))) for Clang. Several isolated
reproductions of this exact declaration shape - doc comment,
template<>, two stacked __attribute__ macros, an overload set sharing
the name - did not reproduce the warning, so this looks like a
Clang comment/declaration-association quirk specific to this overload
inside the much larger basic_json class template, not an actual
documentation defect. Rather than keep the pragma library-wide for one
Clang false positive, added a tightly scoped
-Wdocumentation-deprecated-sync push/pop around just that overload.

Verified with Apple clang 21 and the project's actual -Weverything
flag set (cmake/clang_flags.cmake) on the full header at -std=c++11
and -std=c++20: zero -Wdocumentation* diagnostics. Also compiled
clean with GCC 16.2.0 (the pragmas are already __clang__-gated, so
this only confirms no unrelated breakage). Ran make check-amalgamation
and the full local test suite: 129/129 passing.

#5725 item 2 (step c)

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

* Take the JSON value by const reference in the array and tuple from_json paths

Review feedback on #5737 (gregmarr): once the no-op std::forward calls
are gone, the forwarding references have no purpose. from_json_fn
passes the value as const BasicJsonType&, so these functions were only
ever instantiated with a const lvalue anyway.

The std::array, std::pair and std::tuple overloads of from_json and
their helpers now take const BasicJsonType& and pass j on unchanged.
Because the deduced BasicJsonType is now the plain type, tuple_type and
the static_assert name const BasicJsonType& explicitly, so the
reference checks are unchanged: get<std::tuple<const std::string&>>()
still works, and get<std::tuple<std::string&>>() still fails the same
static_assert. from_json_tuple_get_impl keeps its forwarding reference,
since tuple_type calls it through std::declval.

Behavior, the public API and the ABI do not change. unit-conversions,
unit-constructor1, unit-udt, unit-udt_macro, unit-regression1/2/3,
unit-deserialization, unit-noexcept, unit-items, unit-allocator,
unit-custom-object-type, unit-ordered_json2 and
unit-brace-init-copy-semantics pass at C++11, C++17 and C++20 with
unchanged assertion counts. Ran make amalgamate.

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

---------

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
 2026-10-01 07:37:47 +02:00
Niels Lohmann b54ed188e6 Remove test debt: dead guards, discarded results, and unreferenced files (#5732) 
* Run the README test case in JSON_FastTests jobs

The "README" test case was marked doctest::skip() when the tests
moved from Catch to doctest in 2019, where it replaced Catch's hidden
tag. It is not slow (17 assertions, about 0.00 s), but cmake/test.cmake
only passes --no-skip when JSON_FastTests is off, so the per-compiler
ci_test_*_cxxNN matrix, macOS, Windows Release/ARM, icpc, icpx and
nvhpc compiled the README examples without running them.

Drop the skip decorator so every job runs the case. Test-only change.

Part of #5713

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

* Remove stale clang ranges guards in unit-iterators2.cpp

The "algorithms" and "views" sections were guarded by clang/libstdc++
checks written for a clang 15 (04/2022) bug. The first guard's
condition contradicts its own comment: it skips clang+libc++ and
keeps clang+libstdc++. Both sections already sit inside
`#if JSON_HAS_RANGES`, which macro_scope.hpp excludes for the
toolchains these guards targeted, so the inner guards never let the
sections run on the platforms they meant to protect and are
redundant on the rest. Verified locally with Apple clang 21/libc++
and clang 16.0.6/libstdc++ 12 (Docker): both pass all 1355
assertions with the guards removed.

Part of #5713

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

* Fix copy-pasted CBOR half-float checks; enable stale encode checks

In the RFC 8949 Appendix A test case, the decode checks for
5.960464477539063e-8 (0xf9 0x00 0x01) and 0.00006103515625
(0xf9 0x04 0x00) were copy-pasted from the neighboring -4.0 example,
so those two half-float byte sequences were never actually decoded
and checked, and -4.0 was checked three times instead. The two
float32 encode checks for 100000.0 and 3.4028234663852886e+38 were
commented out before the writer supported emitting float32 and are
now verified to match byte for byte, so they are enabled. The
remaining commented-out half-precision to_cbor checks are collapsed
into a single explanatory comment, since the writer never emits
half-precision floats.

Part of #5713

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

* Assert on the result of STL container conversions in tests

The "object-like STL containers" and "array-like STL containers"
sections converted json values into std::map, unordered_map,
multimap, unordered_multimap, list, forward_list, array, valarray,
vector, deque, set and unordered_set and discarded the result, so
these ~60 conversions only proved that the code compiles and does
not throw; a conversion that dropped or reordered elements would
still pass. Bind each result and compare it against the expected
container. Also fix a copy-paste slip in the deque section
(`j2.get<std::deque<double>>()` instead of j3, so j3's doubles were
never converted to a deque), and remove the dead
`// CHECK(m5["one"] == "eins")` comments that referred to a variable
that did not exist by asserting the equivalent through the bound
result.

Part of #5713

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

* Deduplicate SaxCountdown and other test helpers across formats

SaxCountdown was copied byte-for-byte into six binary-format test
files (unit-cbor.cpp, unit-msgpack.cpp, unit-ubjson.cpp,
unit-bjdata.cpp, unit-bon8.cpp, unit-bson.cpp), about 370 redundant
lines. Move it into tests/src/sax_countdown.hpp (namespace utils,
alongside test_utils.hpp and round_trip_corpus.hpp) and include it
from all six.

trait_test_arg and the "value_in_range_of trait"
TEST_CASE_TEMPLATE_DEFINE were duplicated between unit-32bit.cpp and
unit-bjdata.cpp; the trait is a detail/meta trait, not specific to
either file. Move it into tests/src/value_in_range_of_test.hpp;
unit-32bit.cpp keeps its own include, since JSON_32bitTest=ONLY
builds only that file. Each file keeps its own
TEST_CASE_TEMPLATE_INVOKE list.

sax_no_exception and the "issue #2824" section were duplicated in
unit-regression2.cpp and unit-disabled_exceptions.cpp. Drop the copy
from unit-regression2.cpp; unit-disabled_exceptions.cpp already
covers the no-exceptions case that #2824 was about, and
ci_test_noexceptions reruns it.

No behavior change. Verified by building and running unit-cbor,
unit-msgpack, unit-ubjson, unit-bjdata, unit-bon8, unit-bson,
unit-32bit, unit-regression2 and unit-disabled_exceptions against
include/ (clang++ -std=c++11, ASan/UBSan where applicable); assertion
counts are unchanged from before the refactor.

Part of #5714

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

* Remove the unreferenced vendored libFuzzer

tests/thirdparty/Fuzzer (155 files, ~776 KB of vendored Apache-2.0
LLVM code from the 2016 OSS-Fuzz import) is not referenced by any
CMakeLists, Makefile or workflow: the fuzz drivers link against
-fsanitize=fuzzer or the repo's own
tests/src/fuzzer-driver_afl.cpp. Its vendored README only points at
llvm.org's own libFuzzer docs. Being dead code, it also adds noise
to the flawfinder code-scanning workflow, which scans the whole
tree. Remove the directory and its .reuse/dep5 entry.

Part of #5714

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

* Remove unreferenced 2016 benchmark and fuzz reports

tests/reports (1.6 MB) holds AFL status pages and plots from
2016-08-29 and 2016-10-02, and a nativejson-benchmark snapshot from
2016 with links to rawgit.com, which shut down in 2019. Nothing
references this directory: no doc, README section, script or
workflow points at it, and it describes a ten-years-old, pre-2.0
snapshot of the library.

Part of #5714

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

* Run the CBOR, MessagePack, BSON and BON8 round-trip invariants in CI

tests/src/round_trip_corpus.hpp exists so that the byte-stability
invariant the fuzzer drivers check also runs on a fixed corpus in CI,
instead of only at OSS-Fuzz. So far only the UBJSON and BJData drivers
had a matching unit test; the CBOR, MessagePack, BSON and BON8 drivers
assert the same invariant (assert(to_X(j2) == vec)) but nothing ran it
outside OSS-Fuzz.

Add "<FORMAT> round-trip invariants" test cases to unit-cbor.cpp,
unit-msgpack.cpp, unit-bson.cpp and unit-bon8.cpp, modeled on the
UBJSON case: seed j1 from the corpus (skipping values that do not
survive the format's own round trip, as the fuzzer drivers only ever
see values from_X() actually produced), then require from_X(to_X(j1))
not to throw and check to_X(j2) == to_X(j1). BSON only serializes
objects, so non-object corpus values are skipped. Update the comments
in round_trip_corpus.hpp and tests/fuzzing.md to name all six formats.

The stream-versus-contiguous check in the BON8 driver is left out, as
#5601 reworks it.

A local probe confirms no violations on the current corpus (CBOR 3849
checked, MessagePack 3909, BSON 2958, BON8 3841 - matching the counts
already recorded for this probe in the issue).

Closes #5714 item 1.

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

* Fix fuzzer driver step lists to match the checks the code performs

The header comment of six of the seven binary-format fuzzer drivers
listed an invariant the code does not check: CBOR, MessagePack, BSON
and BON8 said "assert(j1 == j2)", but the code checks byte stability,
assert(to_X(j2) == vec). UBJSON and BJData still described the old
"assert(j1 == j2/j3/j4)" byte-exact check from before PR #5494 replaced
it with a use_size/use_type-aware round trip (UBJSON) and a
value-stability check (BJData); BJData's added paragraph already
explained the new check, but the step list above it did not.

Also remove a dead branch in fuzzer-parse_bson.cpp: from_bson() is
called with allow_exceptions = true, so it throws instead of returning
a discarded value, and the "if (j1.is_discarded()) return 0;" guard
could never trigger. Drop the unused <iostream> include from all seven
drivers and <sstream> from all but fuzzer-parse_bon8.cpp, which is the
only one that uses std::istringstream.

Overlaps #5601, which edits all seven drivers in the same hunks.

Closes #5714 item 4.

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

* Silence the CMP0169 deprecation in cmake_fetch_content, fix stale guards

tests/cmake_fetch_content/project calls the single-argument
FetchContent_Populate(json) after FetchContent_Declare(), which CMake
3.30 deprecated as CMP0169. Since the project declares
cmake_minimum_required(VERSION 3.11...3.14), the policy stays unset,
so every configure with a current CMake prints the deprecation
warning. The test is kept on purpose: it is the only coverage of the
FetchContent_Populate + add_subdirectory pattern for CMake 3.11-3.13
users, which the docs still describe as supported. Explicitly set
CMP0169 to OLD, with a comment explaining why.

Also fix two stale version guards:
- tests/cmake_fetch_content/CMakeLists.txt guarded the test with
  VERSION_GREATER "3.11.0", which is dead now that tests/CMakeLists.txt
  requires CMake 3.13.
- tests/cmake_fetch_content2/CMakeLists.txt guarded with
  VERSION_GREATER "3.14.0", which skips exactly 3.14.0, the first
  version with FetchContent_MakeAvailable. Change it to
  VERSION_GREATER_EQUAL "3.14".

Verified locally: `ctest -R cmake_fetch_content` passes with CMake
4.1, and the CMP0169 deprecation warning that appeared before this
change is gone.

Closes #5714 item 5.

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

* Make the CMake integration-test wrappers consistent

The six tests/cmake_* integration-test wrappers had drifted:

- Only cmake_import and cmake_import_minver forwarded
  -A "${CMAKE_GENERATOR_PLATFORM}" to the inner configure, and none
  forwarded -T "${CMAKE_GENERATOR_TOOLSET}". The Windows workflow
  configures the outer build with -A Win32 -T ClangCL, so without
  forwarding, the inner projects of cmake_add_subdirectory,
  cmake_fetch_content, cmake_fetch_content2 and
  cmake_target_include_directories built with the generator defaults
  instead of matching the outer build's platform and toolset. Forward
  both consistently from all six wrappers.
- cmake_fetch_content and cmake_fetch_content2 passed
  -Dnlohmann_json_source to their inner projects, which never read it
  (CMake warns "manually-specified variables were not used"); the
  inner projects fetch their own copy of the library instead. Drop it.
- tests/CMakeLists.txt set JSON_FORCED_GLOBAL_COMPILE_OPTIONS from the
  matching environment variable but never read the cache variable
  again; the lines right below it read $ENV{JSON_FORCED_GLOBAL_COMPILE_OPTIONS}
  directly, like the LINK_OPTIONS counterpart already does. Remove the
  dead set().

This changes which platform and toolset the Win32 and ClangCL CI jobs
build the four newly-forwarding wrappers' inner projects with, which
may surface new failures there; CI has to confirm those jobs.

Verified locally with Ninja (empty -A ""/-T "" is accepted): all 12
cmake_* tests still pass, and the inner fetch_content configures no
longer warn about the unused variable.

Closes #5714 item 6.

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

* Turn the #972 fifo_map regression test into a real test

The #972 regression test in unit-regression1.cpp only built a
my_json array from a string literal (the original crash) and had no
CHECK, so the fifo_map object type it exists to demonstrate was never
exercised. Meanwhile the docs recommend fifo_map for keeping object
keys in insertion order (object_order.md, template_parameters.md),
and nothing tested that recommendation.

Extend the section: after the original array assignment, parse an
object with my_json::parse() (not via the "..."_json UDL, which
returns a plain nlohmann::json and would exercise the cross-basic_json
conversion constructor instead of the parser's own key insertion -
and, as tried locally, does not keep fifo order for this stateful
comparator) and check that dump() keeps insertion order, and that it
survives erase() and inserting a new key.

Also narrow thirdparty/fifo_map off the include path of every other
test-* target: it was a PUBLIC include directory of test_main, even
though unit-regression1.cpp is its only user. Add a small
fifo_map_include INTERFACE library with that include directory and
attach it to test-regression1 only via json_test_set_test_options().

Verified locally (test-regression1_cpp11, default build and
-fsanitize=address,undefined): the new checks pass; `git grep fifo_map
tests` still only finds unit-regression1.cpp and the vendored header.

Closes #5714 item 7.

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

* Document the vendored doctest.h patch; fix stale doctest_compatibility.h comments

tests/thirdparty/doctest/doctest.h is doctest 2.4.12, imported in
#4771. Two weeks later, #4801 hand-edited translateActiveException()
to declare "String res;" inside the translator loop instead of before
it, so a translator that does not match does not leave a previous
translator's result in "res" for the next iteration to see. Nothing
recorded this, so re-vendoring doctest.h from upstream would silently
drop the fix. Add a comment at the patched site naming the version,
the PR and the reason, so a future re-vendor knows to re-apply it.

Also fix two stale comments in doctest_compatibility.h:
- The DOCTEST_THREAD_LOCAL comment referenced Xcode 6/7, which is no
  longer supported; reword it to explain why the define must stay
  regardless (it keeps doctest's own thread_local usage out of the way
  of the same Clang/MinGW crash that JSON_NO_THREAD_LOCAL works around
  in the library, see ci_test_no_thread_local).
- The <iosfwd> include's comment justified it with tests that define
  "private" as "public"; no test under tests/src does that any more
  (removed by #2352). Reword the comment instead of dropping the
  include, since confirming it is safe to drop needs the full CI
  matrix including MSVC 2015+.

Verified locally that tests/src/unit-readme.cpp still builds and
passes 17/17 with these headers.

Closes #5714 item 9.

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

* Stop compiling unit-wstring.cpp out entirely on classic ICC

tests/src/unit-wstring.cpp wrapped the whole file in
#ifndef __INTEL_COMPILER, with the comment "ICPC errors out on
multibyte character sequences in source files". The ci_icpc job
(intel/oneapi-hpckit:2023.2.1) still exists, so that job ran none of
the wstring/u16string/u32string input adapter tests, including the
malformed-input checks #5704 (open) extends.

Only 9 lines contained non-ASCII bytes: the three *_is_utf16()/
*_is_utf32() probe functions, and three std::wstring/u16string/
u32string literals plus their narrow-string dump() expectations.
Rewrite all of them with \u/\U escapes in the wide/u16/u32 literals
and \x escapes (split into separate string-literal tokens so a
following byte is never read as part of the same hex escape, e.g.
"\xE1\x83\x85" "a") in the narrow ones. Remove the
#ifndef __INTEL_COMPILER/#endif guard along with it.

The *_is_utf16()/*_is_utf32() probes compared a raw multibyte literal
against an escape-based one to detect a compiler that misreads the
source file's encoding; with no raw literals left to misread, the
comparison is now tautological, so drop the probes and the "if"
guards around each SECTION's body instead of leaving them in as dead
checks.

The same non-ASCII-in-source-and-in-a-narrow-comparison pattern
existed once more in unit-deserialization.cpp's "Using _json with
char8_t literals #4945" test: a raw emoji character in a u8R"(...)"
literal, guarded by a check_utf8() that returned false for ICC (same
reason) and for Windows without the active UTF-8 code page. Rewrite
the literal with a \U escape and compare it against a \x-escaped
expectation instead of a second raw literal, and drop check_utf8()
and the now-unused <windows.h> include along with the guard.

Verified locally (clang, -std=c++11 and -std=c++20,
-fsanitize=address,undefined, and a plain build): test-wstring keeps
18/18 assertions and unit-deserialization keeps 466/466 (c++11) and
477/477 (c++20) assertions, matching this branch before the change
exactly - no coverage was gained or lost, only the source-encoding
dependency was removed. ci_icpc has to confirm classic ICC actually
builds and passes test-wstring now; if it does not, that is a real
finding, not a reason to restore the guard.

Overlaps #5704 (open), which edits unit-wstring.cpp inside the
previously-guarded region (an include near the top, checks in the
invalid-string sections, and a new section at the end).

Closes #5713 item 5.

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

---------

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
 2026-10-01 07:33:07 +02:00
Niels Lohmann d71a494367 Fix CI: json_view tests without exceptions, on clang 3.6, and single header 
- ci_test_noexceptions: the helpers that compare the exceptions of
  json_document::parse() and json::parse() catch them outside a
  CHECK_THROWS, so with JSON_NOEXCEPTION the first parse error aborted
  the test. Compile those comparisons only with exceptions, as
  unit-class_parser.cpp does.
- ci_test_gcc: -Werror=unused-result for CHECK_THROWS_AS(json_document::
  parse(...)); assign the result to a dummy document.
- ci_test_compilers_clang (3.6): `const json_view invalid;` needs a
  user-provided default constructor there (CWG 253); value-initialize it.
- ci_test_single_header: json_view.hpp now exists as a single header and
  contains the internal view headers, so unit-json_view_builder.cpp
  includes it instead of the detail headers in that mode, and the test
  is built again with the single header.
- Regenerate single_include/nlohmann/json_view.hpp for the builder change
  merged from json-view/08-view-builder.

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
 2026-09-30 20:58:49 +02:00
Niels Lohmann 0a64e6c99b Fix CI: MSVC C4127 in the view builder and the single-header test build 
- msvc (Win32, /W4 /WX) reported C4127 (conditional expression is
  constant) for `TrailingCommas && cur() == ']'` and the like when the
  option is off. Route the template arguments through a static enabled()
  function, as json.hpp's nesting_depth_exhausted() does.
- ci_test_single_header compiled unit-json_view_builder.cpp against
  single_include/, which does not contain the internal
  nlohmann/detail/view headers. Build that test only with the multiple
  headers.

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
 2026-09-30 20:19:13 +02:00
Niels Lohmann 1e101ecac1 Add BON8 support (#2998) 
* Add BON8 support

Add to_bon8/from_bon8 and input_format_t::bon8 for BON8, a binary format
that uses the byte values that cannot begin a UTF-8 character as type
markers, so strings need no length prefix. It is the most compact of the
supported binary formats on the benchmark files.

The reader is non-recursive like the other binary readers. A string ends
at the first byte that cannot continue it, so the reader hands the one or
two bytes it reads past a string back to the value that follows. The
writer produces the canonical representation of the specification, except
for NFC normalization; its output is identical to that of the reference
implementation (HikoGUI) on all files of the test data.

The round-trip tests need the .bon8 files of json_test_data 3.2.0.

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

* Address review comments

- Reuse detail::validate_one_utf8 to check strings in to_bon8; the error
  now names the first byte of the invalid sequence.
- Document that to_bon8 leaves bytes in the output adapter on an
  exception, and that string_open is only an output of write_bon8_marker.
- Explain why the pushback buffer of the BON8 reader cannot overflow.

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

* Select the BON8 float prefix by type

get_bon8_float_prefix only depends on the type of its argument, so make
the type a template parameter instead of passing an unused value.

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

* Rename a test variable that Flawfinder mistakes for read()

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

* Fix the BON8 CI failures

- compare the float in write_bon8_float with number_float_t constants,
  so GCC does not warn about a float-to-double conversion
- mark check_bon8_utf8's context as used when exceptions are disabled
- choose the compact float prefix in a helper rather than with nested
  conditional operators (clang-tidy)
- use auto for the cast in the BON8 integer reader (clang-tidy)
- write the int32 minimum test values as long long literals (MSVC C4146)

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

* Amalgamate

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

* Read BON8 strings in bulk from contiguous input

- copy the valid UTF-8 of a string in one step when the input is
  contiguous (twitter.json is read in 1.68 instead of 2.52 ms,
  jeopardy.json in 196 instead of 297 ms, close to CBOR and MessagePack)
- share the new valid_utf8_prefix() with the writer's UTF-8 check, which
  now skips ASCII 8 bytes at a time
- let the fuzzer check that contiguous and stream input give the same
  value or error, and test both paths in the unit tests
- clarify that a second 0xFF after a string is an empty string

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

* Link the BON8 functions from the other binary format pages

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

* Name the bulk scan flag after the input, not BON8

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

* Read BSON keys in bulk from contiguous input

BSON keys (and array indices) are C-style strings, which were read byte
by byte. For contiguous input they are now read up to their \x00-byte in
one step, using the same bulk_scan flag as BON8 strings: twitter.json is
read in 1.46 instead of 2.01 ms, citm_catalog.json in 2.93 instead of
3.33 ms, jeopardy.json in 182 instead of 207 ms. canada.json, whose keys
are almost all one-digit array indices, takes 2 % longer.

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

* Fix the BON8 CI failures of the bulk-read tests

- skip the contiguous-versus-stream tests of BON8 strings and BSON keys
  when exceptions are disabled: they catch the parse errors of invalid
  input, and without exceptions the library aborts instead
- use static_cast for the int64 test value (google-readability-casting)

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

* Move the explicit basic_json instantiation into its own test file

Linking test-regression3_cpp20 with clang and MinGW failed with
"relocation truncated to fit: IMAGE_REL_AMD64_REL32 against `.rdata'",
as test-regression2 did before #5511. The explicit instantiation of
basic_json<> for #4825 compiles every member function, including the
BON8 reader and writer, into that object, and it was already close to
the limit (2,226,104 bytes on develop, 2,234,960 with BON8; clang -O1,
C++20).

Give the instantiation a file of its own: unit-regression3 is now
1,594,736 bytes and unit-explicit_instantiation 1,095,064. The new file
mentions JSON_HAS_CPP_17 and JSON_HAS_CPP_20 so it keeps being built
for the C++17 standard the regression was about.

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

* Convert the bytes of the BON8 test strings explicitly

The str() helper constructed a std::string from a byte range, which
converts each unsigned char implicitly; -fsanitize=integer reports that
for bytes of 0x80 and above (ci_test_clang_sanitizer).

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

---------

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
 2026-09-27 16:56:21 +02:00
Niels Lohmann 98278dc3f6 Fix CI: disable MSVC warning C5285 for the vendored doctest (#5577) 
The windows-11-arm runner now ships MSVC 19.51, which reports doctest's
forward declaration of std::tuple as C5285 ("cannot declare a
specialization for 'std::tuple'"). With /WX this breaks the msvc-arm64
job on develop and on every open pull request. Disable the warning for
the test targets, like the other MSVC warnings already disabled there.

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
 2026-09-25 17:54:40 +02:00
Niels LohmannandClaude Sonnet 5 56b3ee566c Fix stack overflow when copying a deeply nested value (#5387) (#5389) 
* Bound the descent of the copy constructor

basic_json's copy constructor copied objects and arrays by handing the
container to its own copy constructor, which copy-constructs every element
and so reaches this constructor again, once per nesting level. A value
nested deeply enough exhausted the call stack and terminated the process
with a segmentation fault - no exception, nothing the caller could catch.
Parsing such a value works, as the parser is iterative, and so does
destroying one, as #1436 made destruction iterative.

Bound how far the copy descends rather than take the call stack away from
it. The first levels are copied exactly as they were - the containers copy
their own elements, which is by far the fastest way to fill them - and only
once the copy has descended 128 levels is the value below it finished
without the call stack, through an explicit worklist. Copying can therefore
no longer exhaust the stack, however deeply a value is nested, while a value
nested less deeply than the bound - all but a vanishing minority - is copied
by the very same code as before and pays only for one counter.

That counter lives in thread_local storage, as one shared between threads
would be raced. JSON_NO_THREAD_LOCAL switches it off for toolchains without
thread_local; copying then goes through the worklist right away, which
yields the same values but is measurably slower.

The deferred values are completed before the copy they belong to returns, so
a value copied while another copy is going on - by a custom base class, say -
is unaffected by the copy it is nested in.

operator= takes its argument by value, so copy assignment is fixed as well.

Copying is as fast as it was, within measurement noise (medians of 9
interleaved runs, clang -O3): -1.3% for an array of strings, +0.0% for a
flat object, +0.1% for a flat array of numbers, +0.3% for nested arrays,
+0.6% for nested objects and +1.2% for a twitter-like document. Copying a
three-key object costs about ten nanoseconds more, the counter. Deferring
every level instead, rather than only those below the bound, measured
between 3% and 9% slower depending on the shape of the value.

This fixes #5387 for the copy constructor. dump() is still recursive.

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

* Test the copy constructor's iterative path in CI

The copy constructor descends into 128 levels before it finishes a value
without the call stack, so the iterative path is otherwise only reached
by the few tests that nest deeper than that.

JSON_NO_THREAD_LOCAL switches the descent off, which sends every value
down that path. Running the whole test suite that way covers it with
every object type, string type, allocator, and base class the suite
already exercises. The new ci_test_no_thread_local target does that; the
macro had no build coverage at all before.

Copying a nested value also has to carry over what the element-wise copy
constructor would have copied: the parents that JSON_DIAGNOSTICS relies
on, and the positions that JSON_DIAGNOSTIC_POSITIONS reports. Both are
now checked on either side of the descent bound, for objects and arrays.
Neither was tested before, and dropping either one makes the new tests
fail.

Also quantify what JSON_NO_THREAD_LOCAL costs a copy instead of calling
it "measurably slower".

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

* Split the regression tests so that they keep linking

Linking test-regression2 fails with "relocation truncated to fit:
IMAGE_REL_AMD64_REL32 against `.rdata'" once its object grows past what
the MinGW linker copes with, and the copy constructor's helpers push it
over: the object grows by 6.3%, from 4,654,128 to 4,944,920 bytes at -O0,
and develop links at the smaller of the two.

Building the tests optimized shrinks the object enough to link, but the
binaries clang 11.0.1 and clang 18.1.8 then produce crash before doctest
prints its first line - 39 of 102 tests on clang 18 - so the objects have
to become smaller rather than denser.

Moving the test cases that follow "regression tests 2" into a file of
their own brings that object to 4,687,888 bytes, which is 0.7% above the
size that links today rather than 6.3%. Both files still build for C++11,
C++17 and C++20, and run the same 9 test cases and 135 assertions as
before, now spread over two binaries.

New regression tests belong in unit-regression3.cpp from here on, which
is what CONTRIBUTING.md now says.

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

* Do not use thread_local storage with Clang targeting MinGW

Every test that copies a value segfaults there - 42 of 105 on clang
11.0.1, 39 of 102 on clang 18.1.8 - while the same tests pass with GCC
targeting MinGW, with Clang targeting MSVC, and with every other
toolchain the library is tested on. The counter that bounds the copy
constructor's descent is the library's first use of thread_local, so
that job had never exercised it before.

JSON_NO_THREAD_LOCAL already covers toolchains without thread_local
storage, and copying yields the same values with it, only more slowly.
Define it for this one automatically.

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

* Balance the warning suppression the split separated

unit-regression2.cpp opens a DOCTEST_CLANG_SUPPRESS_WARNING_PUSH block at
the top and closed it at the very bottom, which the split moved into
unit-regression3.cpp: one file was left with a push and no pop, the other
with a pop and no push, which clang reports as an error.

Give each file the pair it needs.

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

* Check both shapes without a C-style array

clang-tidy rejects the array the two shapes were iterated over
(cppcoreguidelines-avoid-c-arrays). The array only existed because astyle
reformats a range-for over a braced initializer list into something
unreadable; naming the two cases avoids both.

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

* Split the regression tests far enough to leave room

The first split left unit-regression2.cpp 0.7% below the size develop
links at, which the comparison change in the follow-up immediately used
up: the MinGW linker fails on test-regression2_cpp20 again, naming
copy_shallow and to_partial_ordering among the relocations it cannot fit.

Move the sections from "issue #2067" on, and the helper types they use,
so that the file stops being the one that decides whether the tests can
be linked at all. At -O0 and C++20, unit-regression2.cpp is now 2,964,944
bytes against develop's 4,708,248, and 3,070,568 bytes with the follow-up
applied - roughly a third smaller either way, rather than a fraction of a
percent larger.

The 135 assertions are the same ones as before, now spread over three
test cases in two files.

Also silence the clang-tidy findings the deep-nesting tests draw: the
copies they make are what is being tested, and the reserve() computation
gets its parentheses.

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

* Move the #4804 alias to the file that uses it

The split left the json_4804 alias behind in unit-regression2.cpp while
the test case that uses it went to unit-regression3.cpp, which does not
build for C++17 and C++20 as a result.

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

* Include <span> where the split moved its only use

The #2546 test case guards itself with __has_include(<span>), but the
include itself sat in unit-regression2.cpp's preamble and stayed behind,
so the section compiled without a declaration wherever the guard passed -
which nvhpc reported and libc++ builds do not, as they skip the section
altogether.

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

* Keep the descent bookkeeping in one place

Copying carried a depth count, a depth limit and a guard of its own, and
the comparison in the follow-up added a second set beside them. Neither
operation needs its own: they are never nested inside one another by the
library - copying a value does not compare one, and comparing two values
does not copy them - and where user code nests them anyway, sharing the
count only ends a descent sooner than it had to.

So there is now one nesting_depth(), one nesting_depth_limit() and one
nesting_depth_guard, which the follow-up uses instead of adding its own.
Inverting the test in copy_structured leaves the too-deep case and the
no-thread-local case as the same code.

The guard takes the count rather than looking it up, because the caller
has looked it up already to test it against the limit, and reaching
thread-local storage twice on the path that is taken almost every time is
worth avoiding.

The switch that copies the value of anything that is not an object or an
array was written twice - once in the copy constructor, once in
copy_shallow - so that adding a value_t meant editing both, and missing
one would have been silent. It is copy_leaf_value now, and inlined: both
callers have already sorted the containers out, and folding that test into
the switch is what keeps a value made mostly of numbers copying as fast as
it did.

Copying canada.json, citm_catalog.json and twitter.json is within 0.6% of
what it was before, measured as a paired ratio over 18 interleaved rounds
against a run-to-run spread of 0.3%.

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

* Check that an abandoned copy can still be destroyed

Copying a value without the call stack builds the copy from the top down,
and every value whose own copy has not been made yet stays a null value
until it is. That is what lets a copy be abandoned half-built: the
destructor finds nothing but complete values and null ones.

Nothing tested it. Failing an allocation part-way through a copy of a
deeply nested value does, with the allocator the file already has for
exactly this kind of test.

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

* Name the test's locals so Flawfinder stops matching them

The code scanning job reports CWE-362 - "check when opening files" - for
a test that opens no files: Flawfinder matched a local variable called
open. Rename it and its partner.

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

* Keep the descent guard's bookkeeping self-contained

nesting_depth_limit() and nesting_depth_guard were only used inside
the JSON_NO_THREAD_LOCAL-guarded branch of copy_structured(), but were
defined unconditionally. Move them inside the #ifndef, and have the
guard look up the depth and test it against the limit itself (via
okay()) instead of making the caller do it - the caller no longer
needs to touch nesting_depth() at all. Also shrink the thread-local
counter to std::uint8_t, matching what its own doc comment already
argued.

Addresses gregmarr's review comments on #5389.

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

* Make nesting_depth_guard usable regardless of JSON_NO_THREAD_LOCAL

nesting_depth_limit() and nesting_depth() stay behind #ifndef
JSON_NO_THREAD_LOCAL, since a descent cannot be bounded without a
per-thread count. But the guard itself now always exists, becoming a
no-op that is never okay() under that macro - the same way the bound
is already reached on every call without one. copy_structured() no
longer needs to know which case it is in.

This is what lets #5390 reuse the guard for comparison, which cannot
test JSON_NO_THREAD_LOCAL where the macro-based operators use it: the
guard now carries that distinction itself instead of requiring every
caller to.

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

* Silence VS2015's C4503 for the custom-base-class test

The deep-copy support added for #5387 lengthened the mangled name of
std::allocator_traits<...>::construct for the test's map type past
VS2015's limit, which /WX turns into a build failure even though the
name is only used for (now-truncated) debug info.

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

* Remove dead unused-parameter casts from copy_metadata()

@gregmarr asked whether the static_cast<void> pair in the
JSON_DIAGNOSTIC_POSITIONS-off branch was needed for an empty
json_base_class_t. It isn't: src and dst are already referenced
unconditionally by the base-class copy above, so no -Wunused-parameter
warning fires either way (checked with -Wall -Wextra
-Wunused-parameter, JSON_DIAGNOSTIC_POSITIONS 0 and 1).

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

* Fix CI: build custom array types without a fill constructor, re-amalgamate

copy_array_level() built the destination array with the fill
constructor array_t(count, value), which is not part of the array
container interface the library otherwise assumes (e.g. custom
ArrayTypes that only provide a default and an iterator-pair
constructor, as covered by unit-custom-array-type.cpp). Default-
construct the array and resize() it instead, matching how the rest
of the codebase already grows array_t.

Also re-run the amalgamation, which had fallen out of sync with
include/nlohmann/json.hpp.

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

---------

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>
 2026-09-24 17:01:15 +02:00
Niels Lohmann 35cdd5f408 De-duplicate the diagnostic-positions test files via define-based recompilation (#5474) 
tests/src/unit-class_parser_diagnostic_positions.cpp and
tests/src/unit-diagnostic-positions-only.cpp were maintained as near-copies
of unit-class_parser.cpp and unit-diagnostic-positions.cpp respectively, and
had drifted: trailing-comma handling, the #5342 filter-array/filter-value
sections, and the cross-input-adapter diagnostics test were never ported to
the positions-enabled copy.

Fold the position-specific assertions into the base files, guarded by
file a second time with the relevant macro set via CMake COMPILE_DEFINITIONS
(mirroring the existing test-comparison_legacy pattern) instead of
maintaining a separate source file. This removes the duplication and, as a
side effect, closes the coverage gaps above since the full test file now
compiles under JSON_DIAGNOSTIC_POSITIONS=1 as well.

Fixes #5417

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
 2026-09-11 08:18:43 +02:00
Niels LohmannandClaude Opus 4.8 4a93aa4e2f Speed up parsing of contiguous input (numbers, strings, UTF-8) (#5283) 
* Speed up number parsing in the lexer (fast paths from the fast_float/simdjson world)

The number scanner converted its already-validated digit buffer with
std::strtoull/std::strtoll/std::strtod. Those pull in locale and errno
machinery and dominate number-heavy parsing (strtod runs at ~6 M/s).

Replace them with dedicated parsers over the validated buffer:

- parse_integer_unsigned / parse_integer_signed: accumulate digits with
  overflow detection, falling back to the float path on overflow exactly
  as the strtoull/strtoll round-trip check did. Overflow behavior is
  unchanged for narrower or wider custom number types.

- parse_float_fast: Clinger's exact fast path for `double` (<=19 significant
  digits, |exp10| <= 22, significand < 2^53), where significand * 10^exp is
  exact under IEEE round-to-nearest. This is the same fast path used by
  fast_float/simdjson. It is bit-identical to strtod on this subset and
  declines (falling back to strtod) otherwise. Only `double` uses it; float
  and long double keep std::strtof/std::strtold via a templated overload.

Measured on representative data (g++ 13, -O3):
  - integers:  DOM parse +11%, SAX +25-34%
  - floats:    DOM parse +37%, SAX +70%  (clang: float DOM ~1.9x)

No dependencies added; header-only and C++11-clean. Existing parser,
lexer, conversion and deserialization unit tests pass unchanged; a
3M-value random-double fuzz matches strtod bit-for-bit.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AXcDtEma2PjxgmPS9cQGzA
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Add SWAR bulk string scanning for contiguous input (simdjson-style)

scan_string() read the input one character at a time through the input
adapter and classified every byte with a large switch. For contiguous
byte buffers we can instead scan 8 bytes at a time with a SWAR word test
that finds the first byte needing individual handling (the closing quote,
an escape, a control character, or a non-ASCII UTF-8 byte) and bulk-append
the ordinary run in one go.

- input adapters expose supports_bulk_scan / bulk_data / bulk_remaining /
  bulk_skip for provably-contiguous, same-type, 1-byte iterator ranges
  (raw pointers in every standard; std::string/std::vector/std::array and
  friends additionally in C++20 via std::contiguous_iterator).
- the lexer gains a bulk_scan capability (gated on lazy_token_string so
  bypassing the per-character capture cannot lose error diagnostics) and a
  scan_string_bulk() fast path; streaming/wide/user adapters are unchanged
  and keep the byte-at-a-time scanner.

The run contains no newline (all bytes < 0x20 are treated as special), so
position bookkeeping stays exact, and error tokens are still reconstructed
lazily from the consumed byte range. The SWAR special-byte test is pure
uint64_t arithmetic - no intrinsics, no runtime dispatch, C++11-clean.

Measured on representative data, pointer input, g++ 13 -O3
(string values discarded by accept() see the largest gains):

  long ASCII strings:  DOM +4.5x,  SAX +14x,   accept +17x  (to ~2 GB/s)
  short strings:       DOM +15%,   SAX +62%,   accept +85%
  escape-heavy:        DOM +31%,   SAX +26%,   accept +28%

Same-input parity verified: 200k randomized documents (escapes, multibyte
UTF-8, surrogate pairs) accept/parse identically via the contiguous SWAR
path and the streaming byte path; unit lexer/parser/diagnostic-position/
deserialization/conversions suites pass unchanged.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AXcDtEma2PjxgmPS9cQGzA
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Validate UTF-8 in the bulk string scanner (portable ~2-3x on non-ASCII text)

The SWAR bulk string path stopped at the first non-ASCII byte and handed
every multibyte character to the byte-at-a-time scanner, whose per-byte
get()/next_byte_in_range()/add() machinery runs at roughly half the speed
of validating straight from the buffer. As a result, dense non-ASCII text
(CJK, emoji, accented Latin) parsed ~10-15x slower than ASCII.

Fold well-formed UTF-8 into the bulk run: scan_string_bulk() now, on a
non-ASCII lead byte, validates one sequence with validate_one_utf8() -
which mirrors scan_string()'s per-byte switch ranges exactly (rejecting
overlong forms, surrogates, and out-of-range code points) - and appends it
in place, continuing until the closing quote, an escape, a control byte,
or an ill-formed sequence. All error handling still defers to the byte
path, so error messages and positions are byte-for-byte unchanged.

Because only well-formed content is fast-pathed and every rejection falls
through to the existing scanner, behavior is identical; the win is purely
throughput. Measured on pointer input (accept, string values discarded):

  content        g++ 13         clang 18
  dense CJK      277 -> 648     ~605  MB/s   (~2.3x)
  dense emoji    299 -> 857     ~702  MB/s   (~2.6-2.9x)
  mixed 90% ASCII 246 -> 331    ~334  MB/s   (~1.35x)
  pure ASCII     unchanged (~3.2 / 4.1 GB/s)

Verified: 2,000,000 randomized documents built from arbitrary bytes
(overlong, surrogate, truncated, out-of-range sequences) accept/reject and
parse identically via the contiguous path and the streaming byte path;
lexer/parser/diagnostic-position/deserialization/conversions suites pass
unchanged. Pure C++11, no intrinsics.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AXcDtEma2PjxgmPS9cQGzA
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Route contiguous byte containers through the pointer adapter (fast paths in C++11)

json::parse(std::string) - the most common entry point - did not benefit
from the contiguous fast paths (bulk string scanning, UTF-8 bulk
validation, memcpy for binary formats) in C++11..17: std::string::iterator
is a library wrapper, not a raw pointer, and pre-C++20 there is no portable
way to prove it contiguous, so supports_bulk_scan was false. Only raw
pointers, string literals, and C-arrays (and, in C++20, anything modelling
std::contiguous_iterator) took the fast path.

Detect contiguous single-byte containers (std::string, std::vector<char>,
std::vector<std::uint8_t>, std::string_view, ...) via is_contiguous_byte_
container and route them through an iterator_input_adapter built from
data()/data()+size(). The generic iterator-based container overload is
constrained to exclude these, so the two overloads are disjoint and there
is no ambiguity (a plain competing overload loses to the greedy
forwarding-reference container overload on reference binding, and a factory
partial-specialization is ambiguous - both were tried and rejected).

The pointer keeps the container's own element type, so char_type - and
therefore all parsing behavior - is byte-for-byte identical to the iterator
path (const char* for std::string, const std::uint8_t* for
std::vector<std::uint8_t>); only the raw pointer additionally turns on the
fast paths. Lifetimes are unchanged: the container outlives the adapter for
the full parse expression, exactly as the iterators it replaces did.

Measured, C++11, json::parse/accept(std::string), g++ 13:

  long ASCII strings:  accept 201 -> 3200 MB/s  (~16x),  parse 174 -> 1444
  dense CJK:           accept 263 ->  697 MB/s  (~2.6x)
  short strings:       accept 163 ->  243 MB/s  (~1.5x)

Verified: char_type preserved for std::string (char) and
std::vector<std::uint8_t> (uint8_t); CBOR/MsgPack round-trips from
std::vector<std::uint8_t> unchanged; 1,000,000 randomized documents accept
and parse identically via std::string and via std::istream;
deserialization/user-defined-input/parser/lexer/conversions/diagnostic-
position suites pass (20,480 assertions); warning-clean on g++ and clang in
C++11/17/20.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AXcDtEma2PjxgmPS9cQGzA
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Add optional simdutf backend for bulk UTF-8 validation (JSON_USE_SIMDUTF)

The bulk string scanner validates UTF-8 straight from a contiguous buffer.
The scalar validator caps at ~0.3-0.7 GB/s on non-ASCII text; a SIMD
validator reaches several GB/s. Rather than hand-rolling SIMD UTF-8
validation (easy to get subtly wrong - a from-scratch SSE attempt rejected
valid CJK), wire in the vetted simdutf library behind an opt-in switch.

simdutf is not header-only (it ships simdutf.cpp and uses runtime CPU
dispatch), so it is not vendored: defining JSON_USE_SIMDUTF includes
<simdutf.h> and routes the bulk validator through simdutf::validate_utf8;
the project supplies and links simdutf. Undefined (the default), nothing
external is included and the portable C++11 scalar path is used, so the
library stays header-only and its baseline behavior is unchanged.

Design keeps behavior identical either way:
- scan_string_bulk() now finds the run up to the next quote/escape/control
  byte (non-ASCII allowed) and validates it in one shot; on the rare
  validation failure it recomputes the exact valid prefix with the scalar
  helper, so ill-formed input still falls through to the byte path and is
  reported at the same position with the same message.
- the per-sequence scalar path is factored into scalar_string_bulk_run()
  and is the default backend; the refactor is behavior-preserving and does
  not change scalar throughput.

Verified: default and JSON_USE_SIMDUTF builds accept/reject/parse
identically across 2,000,000 arbitrary-byte documents and 1,000,000
mixed-escape/UTF-8 documents (differential fuzz vs the streaming byte
path); lexer/parser/diagnostic-position/deserialization suites pass under
both configurations (20,188 assertions with the backend enabled);
warning-clean on g++ and clang, C++11 and C++20, both configurations.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AXcDtEma2PjxgmPS9cQGzA
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Add a contiguous fast path for scanning numbers

scan_number() reads a number one character at a time through the input
adapter (get()) and appends each byte to token_buffer (add()) before
converting. For contiguous input, the per-character get()/add() overhead
dominates: it is roughly two thirds of the time spent on number-heavy
parsing, far more than the value conversion itself.

Add scan_number_bulk_contiguous(), which parses the whole number token
straight from the input buffer: it validates and classifies the extent
with the same grammar as scan_number()'s state machine, materializes
token_buffer in one copy (substituting the locale decimal point exactly as
scan_number() does), advances the adapter, and reuses the shared
convert_number() tail. On anything it does not recognize as a well-formed
number it makes no state change and returns token_type::uninitialized, so
the caller falls back to scan_number(), which then produces the exact
diagnostic. Errors and their positions are therefore unchanged.

The conversion tail is factored out of scan_number() into convert_number()
so both scanners share it; the fast path is selected by tag dispatch on the
existing bulk_scan capability, so streaming/wide/user adapters are
unaffected.

Measured on pointer input, g++ 13 -O3:
  - integers: parse +65%, accept +98%
  - floats:   parse +39%, accept +70%

Verified: 2,000,000 randomized number documents (including overflow-range
integers, long digit strings and %.17g doubles) parse identically via the
contiguous path and the streaming byte path, matching value, type and
round-trip text; the locale suite and existing parser/lexer/conversions/
deserialization tests pass; a new "lexer number fast path" test checks
contiguous-vs-streaming parity, token classification, and that malformed
numbers are rejected identically. Pure C++11, no intrinsics.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AXcDtEma2PjxgmPS9cQGzA
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Move byte-level scan/parse helpers out of lexer.hpp

lexer.hpp had grown by ~600 lines of byte-level helpers that have no
dependency on the lexer's template parameters and clutter the state
machine. Move them, unchanged, into two focused headers as free functions
in namespace detail:

- number_parse.hpp: parse_integer_unsigned/parse_integer_signed (now
  templated on the number type) and parse_float_fast (Clinger's exact
  double fast path, with the decimal point passed as an argument instead of
  read from a lexer member).
- string_scan.hpp: the SWAR string helpers (is_string_special,
  swar_string_special, find_string_special, validate_one_utf8,
  scalar_string_bulk_run) and the backend-dispatched string_bulk_run,
  including the optional simdutf include and find_string_delimiter.

lexer.hpp now includes these and calls the free functions; the methods that
touch lexer state (scan_string, scan_number, scan_string_bulk,
scan_number_bulk_contiguous, convert_number) stay put. This is a pure code
move with no behavior change: lexer.hpp drops from 2357 to 1934 lines, the
now-unused <cstdint>/<cstring>/<limits> includes are removed, and the
free-function form makes the SWAR helpers reusable elsewhere (e.g. the
serializer's string escaping).

Verified: default and JSON_USE_SIMDUTF builds compile; 2,000,000 number and
2,000,000 arbitrary-byte-string differential-fuzz documents parse
identically to before; lexer/parser/conversions/deserialization/locale/
diagnostic-position suites pass (20,576 assertions); warning-clean on g++
and clang in C++11/17/20; the amalgamation regenerates and passes
check-amalgamation.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AXcDtEma2PjxgmPS9cQGzA
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Fix clang-tidy findings and document JSON_USE_SIMDUTF in the nav

- number_parse.hpp: use std::array for the powers-of-ten table
  (avoid-c-arrays) and `auto` for the cast-initialized result
  (modernize-use-auto), matching the codebase style (cf. the serializer's
  utf8d table). Indexing casts keep the -Wsign-conversion build clean.
- add JSON_USE_SIMDUTF to the mkdocs navigation so the macro page is
  reachable.

No behavior change; clang-tidy is clean on the new headers and the
amalgamation is regenerated.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AXcDtEma2PjxgmPS9cQGzA
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Fix number fast path for custom string types without assign()

The contiguous number fast path materialized token_buffer with
token_buffer.assign(data, len), but string_t is only required to provide
the minimal interface the rest of the lexer uses (push_back, append,
clear, operator[], ...). Custom string types such as the test's alt_string
do not implement assign(), so scan_number_bulk_contiguous() failed to
compile for them (unit-alt-string), breaking the gcc/clang standards and
old-compiler CI jobs.

reset() already clears token_buffer, so fill it with append() - which
alt_string and std::string both provide and which the string fast path
already relies on - instead of assign().

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AXcDtEma2PjxgmPS9cQGzA
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Satisfy clang-tidy: parenthesize math and drop unused forwarding reference

The CI clang-tidy (newer than the locally available version) reported two
additional checks on the new code:

- readability-math-missing-parentheses: parenthesize the (a * b) + c digit
  accumulations in number_parse.hpp.
- cppcoreguidelines-missing-std-forward: the contiguous-byte-container
  input_adapter overload took a forwarding reference but only reads
  data()/size() and never forwards it. It is already disjoint from the
  generic container overload via SFINAE, so a plain const& is correct and
  clearer (and keeps the container alive for the whole parse just as before).

No behavior change; char_type and routing are unchanged (std::string and
std::vector<std::uint8_t> still take the pointer adapter with char/uint8_t
char_type), CBOR/MsgPack round-trips and the 2M number fuzz still pass.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AXcDtEma2PjxgmPS9cQGzA
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Use std::from_chars (Eisel-Lemire) for float conversion when available

The Clinger fast path is exact only for the "easy" subset (<=19 significant
digits, |exp10| <= 22); high-precision and scientific floats fall through to
strtod, where the failed Clinger attempt actually makes parsing a net loss.
std::from_chars implements the Eisel-Lemire algorithm in modern standard
libraries: locale-independent, correctly rounded, and fast over the whole
value range.

convert_number() now tries parse_float_from_chars() first (guarded by
__cpp_lib_to_chars, so C++11 and libc++-without-float-support keep the
Clinger + strtod path unchanged), then Clinger, then strtof. from_chars is
used only when it consumes the entire token; a partial parse means a non-'.'
locale decimal point, and an under-/overflow (result_out_of_range) also
declines - in both cases the existing strtod fallback supplies the exact
value and the well-defined +/-inf/0 the parser expects, side-stepping the
P4168 divergence between implementations. float and long double now get the
fast path too (Clinger was double-only).

Measured, C++17, g++ 13 -O3, json::parse/accept:
  - canada-style floats: ~unchanged (Clinger already covered them)
  - high-precision (17 digits):  parse 2.1x, accept 2.5x
  - scientific (17 digits + exp): parse 3.6x, accept 4.1x

Verified: C++11 (Clinger/strtod) and C++17 (from_chars) parse every value -
including subnormals, boundary values, and 1e9999/1e-9999 over-/underflow -
to bit-identical results; 2M number-fuzz clean; conversions/deserialization/
locale/number-fast-path suites pass in both C++11 and C++17; clang-tidy
clean; warning-clean on g++ and clang in C++11/17/20.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AXcDtEma2PjxgmPS9cQGzA
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Guard from_chars use on JSON_HAS_CPP_17, not just __cpp_lib_to_chars

libstdc++ 15 defines __cpp_lib_to_chars even in C++14 mode (via
bits/version.h pulled in by other headers), but <charconv> is only included
under JSON_HAS_CPP_17. That made parse_float_from_chars() reference
std::from_chars without the header in C++14 builds, breaking gcc-latest,
icpx, and the offline-testdata jobs.

Gate the use on JSON_HAS_CPP_17 && __cpp_lib_to_chars so it matches the
include condition exactly; C++11/14 always take the scalar fallback.
Verified by forcing __cpp_lib_to_chars in a C++14 build: the guard
suppresses std::from_chars and it compiles. C++17 behavior is unchanged.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AXcDtEma2PjxgmPS9cQGzA
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Disable Clinger float fast path under extended FP precision (x87)

The contiguous number fast path uses a Clinger-style exact algorithm
(significand * 10^scale in double arithmetic), which is only correctly
rounded when double operations are evaluated in true 53-bit precision.
On the x87 FPU used by 32-bit x86 (FLT_EVAL_METHOD == 2) the single
multiply/divide is computed in 80-bit and then double-rounded to double,
so a small fraction of values land 1 ULP off.

This surfaced as test-cbor_cpp11 and test-msgpack_cpp11 failing on the
mingw (x86) job for regression/floats.json: the C++17 builds pass because
they take the correctly-rounded std::from_chars path, while C++11 falls
back to parse_float_fast(). A 5M-sample check over shortest round-trip
decimals reproduces it: 0 divergences with 53-bit doubles, ~1 in 25 000
with 80-bit intermediates; declining to std::strtod fixes all of them.

Guard parse_float_fast() on FLT_EVAL_METHOD so it declines whenever the
platform evaluates doubles in extended precision, letting the caller use
the correctly-rounded std::from_chars / std::strtod path instead. On
mainstream x86-64/ARM64 (FLT_EVAL_METHOD == 0) the fast path is unchanged.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AXcDtEma2PjxgmPS9cQGzA
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Guard <charconv> include with __has_include for GCC 7

GCC 7 sets __cplusplus to the C++17 value under -std=gnu++1z, so
JSON_HAS_CPP_17 is defined, but its libstdc++ ships no <charconv> header
(added in GCC 8; floating-point from_chars in GCC 11). The unconditional
"#if defined(JSON_HAS_CPP_17) #include <charconv>" therefore failed to
compile there: "fatal error: charconv: No such file or directory" in the
ci_test_compilers_gcc (7) job.

Wrap the include in __has_include(<charconv>), mirroring the library's
existing handling of <version> and <filesystem> in macro_scope.hpp. When
the header is absent, __cpp_lib_to_chars stays undefined and
parse_float_from_chars() takes its scalar fallback, so the from_chars use
site (already gated on __cpp_lib_to_chars) is never reached. GCC 8-10,
which have <charconv> but no floating-point from_chars, are unaffected:
they include the header but still take the fallback. GCC 11+ is unchanged.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AXcDtEma2PjxgmPS9cQGzA
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Restore the column when ungetting a newline

The contiguous number fast path never reads the character that terminates
a number token, while scan_number() reads it and then ungets it. When that
character is a newline, get() has already cleared chars_read_current_line,
and unget() could only restore lines_read - leaving the column at 0. The
two paths therefore reported different columns for the same document:

    json::parse("[01\n]")      -> line 1, column 3
    json::parse(stringstream)  -> line 1, column 0

Remember the column the newline was read at so unget() can restore it.
Both paths now report the position the offending token actually starts at,
which also fixes the pre-existing column-0 artifact for streaming input.

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

* Extend the bulk scan fast paths to sized sentinels

supports_bulk_scan required IteratorType and SentinelType to be the same
type, which excluded std::counted_iterator paired with std::default_sentinel_t
- the combination #5268 had already enabled for the memcpy fast path. Such
input fell back to the byte-at-a-time scanner even though it is contiguous
and its remaining length is computable in O(1).

Factor the "distance is computable in O(1)" test into sentinel_is_sized and
use it for iterator_is_contiguous, supports_seek, and supports_bulk_scan
alike, and share the std::ranges::distance/std::distance dispatch through a
remaining_count() helper.

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

* Document JSON_USE_SIMDUTF on the macro overview page

The macro was only listed in the API macro index; add it to the supported
macros overview alongside the other JSON_USE_* macros, and note that it
selects between two definitions of the same inline function and so must be
defined identically in every translation unit.

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

* Amalgamate source code

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

* Cover the counted-iterator bulk scan paths

Sized sentinels newly reach the bulk string/number scanners and the
seek-based token reconstruction, so exercise both:

- diagnostics that quote the offending token, which are rebuilt from the
  consumed input via copy_consumed_range()
- inputs whose count ends before the underlying buffer does, including a
  closing quote that exists only behind the count, a cut inside an 8-byte
  SWAR stride, and a cut inside a UTF-8 sequence

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

* Fix the SentinelType example on iterator_input_adapter

The comment offered "a C++20 sentinel or counted_iterator" as examples of a
SentinelType, but std::counted_iterator is the IteratorType - the sentinel it
pairs with is std::default_sentinel_t. #5268 corrected the same wording in the
API documentation and left the code comment behind.

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

* Do not discard the parse result in the error position check

json::parse is declared warn_unused_result, and CHECK_THROWS_WITH_AS
evaluates its expression as a discarded statement, so the assertion broke
the -Werror builds (GCC -Werror=unused-result, MSVC C4834 under /WX).
Compare against the helper that already captures the message instead.

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

* Lock the two number grammars together with a parity test

The JSON number grammar is encoded twice: as the scan_number() state machine
and as the contiguous fast path. The fast path declining on anything it does
not recognize keeps most divergence harmless, but if it ever accepted
something the state machine rejects the result would be a silent correctness
bug, and the existing test only pinned a hand-written list of numbers.

Enumerate every string of length 1..4 over "01.eE+-" (2800 tokens) and
require both paths to agree on the parsed value and on the exact error
message. Verified to fail if the fast path's grammar is perturbed.

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

* Skip token_buffer for integers on the contiguous path

An integer token does not need token_buffer: the number_integer and
number_unsigned SAX callbacks take only the value, and the overflow
diagnostic rebuilds the text from the input via get_token_string(). Convert
straight from the input buffer and materialize the token only for the
floating-point tail, which still needs a NUL-terminated buffer for strtod.

JSON_DIAGNOSTIC_POSITIONS derives a number's start position from
get_string().size(), so the copy is kept when that is enabled.

The integer dispatch is factored into convert_integer() and shared with
convert_number(), so both scanners keep using one implementation.

Integer-heavy input, 400k values, -O3:

              parse    accept
  gcc 16      +14%     +23%
  clang       +15%     +22%

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

* Amalgamate source code

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

* Cover the bulk string and UTF-8 scanners

These paths had no dedicated tests and rested on differential fuzzing only.
Add three sections, all comparing the contiguous scanner against the
byte-at-a-time one on the parsed value and on the exact error message:

- every string of length 1..3 over an alphabet of ordinary ASCII, both
  specials, a control byte, escape characters, UTF-8 lead and continuation
  bytes, and a byte that is never valid - each at offset 0 and offset 9, so
  the bulk scanner sees them with and without a run behind them
- every kind of run-ending byte at each offset across two 8-byte SWAR words,
  so multibyte sequences also straddle the word boundary
- the boundaries of every range validate_one_utf8() recognizes: shortest and
  longest encodings, overlongs, both ends of the surrogate block, U+10FFFF
  and just past it, and truncated sequences

Verified to fail if the bulk validator accepts surrogates, and if the SWAR
word test stops detecting control characters.

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

* Format the new string fast path test with astyle

The pinned astyle expands a braced-init-list used as a range-for range onto
several lines; hoist the two offsets into a named vector instead, which reads
better and leaves nothing for astyle to reformat.

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

* Fix shadowed locals and guard the exception-dependent tests

Two problems in the tests added for the bulk scanners, both found by CI:

- the inner `const json j` in the counted-iterator diagnostics shadowed the
  one declared at test-case scope, which -Wshadow rejects on GCC and clang
  and C4456 rejects on MSVC under /WX; rename them
- the new parity checks parse deliberately invalid input, which calls
  std::abort() rather than throwing when JSON_NOEXCEPTION is defined, so
  they would have crashed the no-exception build; guard them the way the
  other tests do

json::accept() does not abort, so the UTF-8 range assertions stay compiled
without exceptions and keep covering validate_one_utf8() there.

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

* Gate the C++20 iterator classification on JSON_HAS_RANGES

Making supports_bulk_scan depend on iterator_is_contiguous meant the trait
is now instantiated for every adapter, not only when get_elements() is
called. On standard libraries with an incomplete <ranges> that is fatal:
libstdc++ 10 evaluates std::contiguous_iterator<std::counted_iterator<T*>>
by calling std::to_address, which needs an operator-> its counted_iterator
does not have, so satisfaction checking is a hard error rather than false.
Reported by clang 14 + libstdc++ 10.

JSON_HAS_RANGES already encodes exactly this ("libstdc++ < 11 has incomplete
C++20 ranges", #4440), so require it for the C++20 branch. Affected
toolchains fall back to the pointer-only test and the byte-at-a-time
scanner, which parses identically, just without the bulk fast paths.

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

* Amalgamate source code

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

* Satisfy clang-tidy in the new bulk scanner tests

- give the helper lambdas an explicit std::string return type and return
  braced initializer lists (modernize-return-braced-init-list)
- replace the C-style array of test cases with a std::vector
  (modernize-avoid-c-arrays)
- silence pro-type-member-init on the two brace-initialized aggregates;
  default member initializers would stop them being aggregates in C++11

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

* Avoid escaped literals in the counted-iterator diagnostics list

clang-tidy reads "[\"\\ud834\"]" as a literal better written raw, and the
two literals written next to each other in "[\"a\x01""b\"]" as a missing
comma. The concatenation was there to stop the hex escape swallowing the
following character; build those documents from explicit bytes instead and
use raw strings elsewhere. The byte sequences are unchanged.

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

* Reattach convert_number's documentation

@gregmarr spotted that convert_integer() was inserted between convert_number()
and its doc block, leaving convert_integer() with two stacked blocks and
convert_number() with none. Comment only; no code change.

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

* Address review comments on the simdutf backend

string_bulk_run() had the same `return scalar_string_bulk_run(...)` in both
arms of the `#if`. The simdutf arm already falls through when validation
fails, so a single return after the `#endif` says the same thing.

The JSON_USE_SIMDUTF example showed `#include <simdutf.h>`, which
string_scan.hpp already does under the same guard; users only have to put the
header on the include path and link the library, not include it themselves.

Set the version history entry to 3.13.0, matching the other macro pages
documenting unreleased features.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MXDi7NTMKAmoArUZSKMc4T
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Pin the number error position against a non-newline terminator

The comment claimed the reported column is the one the offending token starts
at. It is the column reached after the token's last character - which is the
actual point of the unget() change: a number terminated by a newline now
reports what the same number terminated by a space always did.

Assert that equality directly, and add a multi-character token where the start
and end columns differ, so the invariant cannot be read off a single-character
example.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MXDi7NTMKAmoArUZSKMc4T
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Do not repeat the integer conversion that just failed

scan_number_bulk_contiguous() converts an integer token straight from the input
buffer. When the value does not fit, it materializes token_buffer and calls
convert_number(), which tried the very same integer conversion again before
falling back to floating point.

Recording the outcome in number_type skips the second attempt. The resulting
token type and value are unchanged: convert_number() reached the float tail
either way.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MXDi7NTMKAmoArUZSKMc4T
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Require a container's value_type to match what data() points at

is_contiguous_byte_container accepted any type with a data() returning a
pointer to a single-byte integral plus a size(). That is duck typing: the two
members say nothing about size() counting the units data() points at.

A type where it does not - fixed-size records, say - was routed to the
pointer-based adapter and parsed as [data(), data() + size()) bytes, silently
truncating input the iterator-based adapter had read in full:

    struct record_buffer {
        using value_type = std::array<char, 4>;
        std::string bytes;
        const char* data() const;                        // raw bytes
        std::size_t size() const;                        // in records
        const char* begin() const; const char* end() const;
    };
    json::parse(record_buffer{"[1,2,3,4,5]"});           // parse error at column 3

Requiring the container's own value_type to be that same element type ties the
two together. Every contiguous standard container satisfies it, so std::string,
std::vector<char>, std::array<char, N> and std::string_view keep the fast path;
anything else falls back to the iterator-based adapter, which is always correct.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MXDi7NTMKAmoArUZSKMc4T
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Only compile the simdutf backend from C++17 on

simdutf.h rejects anything below C++17 with an #error, so defining
JSON_USE_SIMDUTF in a C++11 or C++14 translation unit did not fail with a
message about simdutf being unavailable - it failed to compile at all, taking
the library's C++11 support with it. Nothing caught this because no build ever
compiled that path.

Gate the include and both uses on JSON_HAS_CPP_17, the same way number_parse.hpp
gates std::from_chars. Below C++17 the macro now has no effect and the scalar
validator runs; it accepts and rejects exactly the same input, so the macro is
safe to set project-wide even when some translation units use an older standard.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MXDi7NTMKAmoArUZSKMc4T
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Test the simdutf backend in CI

JSON_USE_SIMDUTF was documented and shipped but never compiled by anything in
the repository, so nothing held the backend to the behavior the docs promise.

Add JSON_TestSimdutf (OFF by default), which fetches simdutf and defines
JSON_USE_SIMDUTF for every test target, and a ci_test_simdutf target that runs
the whole suite in that configuration. Because simdutf needs C++17, the suite is
built at C++11 as well, so one job covers both the scalar fallback with the
macro defined and simdutf itself.

The dependency hangs off test_main, whose usage requirements every test target
inherits. The library target and the installed CMake package are deliberately
untouched: making nlohmann_json link simdutf would put a find_dependency() in
the exported package, which is a separate decision.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MXDi7NTMKAmoArUZSKMc4T
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Warn when JSON_TestSimdutf cannot reach the simdutf backend

simdutf needs C++17: without it the dependency does not even compile, and with
a C++17 compiler but no C++17-or-later standard under test it builds and then
goes unused. Either way the option silently did nothing useful, or broke the
configure step outright.

Resolve the tested standards first, then check them: when none of them can
reach simdutf, skip the dependency and say so, naming which of the two reasons
applies and how to fix it. The tests then run against the scalar validator,
which is what would have happened anyway.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MXDi7NTMKAmoArUZSKMc4T
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Use record_buffer::data() so clang does not flag it unneeded

The record_buffer test type declares data() and size() so the
is_contiguous_byte_container trait can see both and still reject the
type on its value_type. data() was never called, so clang's
-Wunneeded-member-function (under -Weverything -Werror) failed the
C++20 build. Assert that data() points at the underlying bytes: it
ODR-uses the member and documents the property the type is meant to
demonstrate.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XAYM1qhSA2FDaDcGfPW3fG
Signed-off-by: Niels Lohmann <mail@nlohmann.me>

* Skip the float fast path when it cannot succeed

parse_float_fast() (Clinger) needs a significand below 2^53, so it always
declines once the mantissa has 17 or more significant digits. convert_number()
called it unconditionally, so those numbers were walked an extra time before
strtod had to run anyway. On streaming input, where scanning is byte-at-a-time
and there is no compensating win, that made canada.json about 6% slower than
develop.

Derive the significant-digit count from token_buffer indices - the digits are
not scanned again - and skip the call when it is guaranteed to decline. Both
scanners pass the offset where the mantissa ends; the count only has to be
corrected for a leading "0", which the JSON grammar admits nowhere else. The
integer path returns before the check, so integer-heavy input is unaffected.

Values are unchanged: this only avoids an attempt that would have failed.
Verified bit-exact against develop over every number in canada.json,
floats.json, signed_ints.json, unsigned_ints.json, small_signed_ints.json,
citm_catalog.json and twitter.json, for both the contiguous and the streaming
scanner.

  parse, streaming     develop    before     after
  canada.json           19.4ms    20.5ms    19.3ms
  floats.json          135.9ms   131.8ms   128.0ms

  parse, contiguous    develop    before     after
  canada.json           15.5ms    12.9ms    11.7ms
  floats.json           98.6ms    69.8ms    66.7ms

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

---------

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
 2026-09-10 17:20:48 +02:00
Niels Lohmann c05347dd54 Undefine the four JSON_HEDLEY_* macros that leak after including json.hpp (#5475) 
* Undefine the four JSON_HEDLEY_* macros that leak after including json.hpp

include/nlohmann/detail/macro_unscope.hpp includes hedley_undef.hpp to
#undef every JSON_HEDLEY_* macro so none of them leak into the including
translation unit. Four macros were missing from that list and therefore
stayed defined after #include <nlohmann/json.hpp>:

- JSON_HEDLEY_PRAGMA
- JSON_HEDLEY_PREDICT_TRUE
- JSON_HEDLEY_PREDICT_FALSE
- JSON_HEDLEY_CLANG_HAS_DECLSPEC_ATTRIBUTE

hedley_undef.hpp is generated (via `make update_hedley`) by grepping
hedley.hpp for its own internal `#undef JSON_HEDLEY_X` redefinition
guards. JSON_HEDLEY_PRAGMA/PREDICT_TRUE/PREDICT_FALSE have no such guard
in upstream Hedley, so they were never picked up. The guard for
JSON_HEDLEY_CLANG_HAS_DECLSPEC_ATTRIBUTE also has an upstream typo
(`JSON_HEDLEY_CLANG_HAS_DECLSPEC_DECLSPEC_ATTRIBUTE`), so hedley_undef.hpp
was undefining the wrong (never-defined) name.

Fixes:
- include/nlohmann/thirdparty/hedley/hedley_undef.hpp: corrected the
  DECLSPEC_ATTRIBUTE typo and added the three missing #undef lines,
  keeping the file's alphabetical ordering.
- Makefile (update_hedley target): changed hedley_undef.hpp generation to
  extract macro names directly from every `#define JSON_HEDLEY_...` in
  hedley.hpp instead of from existing `#undef` guards, so a future
  `make update_hedley` run undefines every macro Hedley actually defines,
  even ones without a pre-existing redefinition guard. This was not run
  in this PR (it would also pull in an unrelated upstream Hedley sync);
  hedley_undef.hpp was hand-patched instead and single_include was
  regenerated with `make amalgamate`.
- tests/src/unit-no-macro-leak.cpp: new regression test (picked up
  automatically by tests/CMakeLists.txt's existing unit-*.cpp glob) that
  includes json.hpp and then #ifdef/#error-checks every JSON_HEDLEY_*
  macro name, so any future leak of any of the 151 vendored macros fails
  the build, not just the four fixed here.

Fixes #5408.

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

* Derive the JSON_HEDLEY_* leak-check test from hedley.hpp at build time

tests/src/unit-no-macro-leak.cpp previously hardcoded a static list of
~151 #ifdef/#error checks, one per JSON_HEDLEY_* macro name known at the
time it was written. That list would silently go stale the next time
`make update_hedley` pulls in a vendor update that adds, removes, or
renames a macro, since nothing would force it to be regenerated.

Add cmake/scripts/gen_hedley_undef_check.cmake, which derives the full
list of JSON_HEDLEY_* macro names directly from
include/nlohmann/thirdparty/hedley/hedley.hpp:

- tests/CMakeLists.txt uses it (MODE=checks) to (re)generate
  hedley_undef_checks.inc at configure and build time, and wires the
  generating custom target as a dependency of the test-no-macro-leak_cpp*
  targets so it can never build against a stale copy. unit-no-macro-leak.cpp
  now just #include-s the generated file inside its TEST_CASE instead of
  carrying the checks itself.
- The Makefile's `update_hedley` target now delegates hedley_undef.hpp
  generation to the same script (MODE=undef, new `update_hedley_undef`
  target), so the vendored header, the generated #undef list, and the
  generated test checks are all derived from the same extraction logic and
  cannot drift apart.

This mirrors the approach taken independently in #5415 for the same
issue (#5408), credited there to a self-regenerating mechanism that
"can never drift again" -- ported into this branch instead of the
static list originally proposed here.

Verified with a local CMake configure + build + ctest, both against
include/ (JSON_MultipleHeaders=ON) and against the amalgamated
single_include/nlohmann/json.hpp (JSON_MultipleHeaders=OFF), and by
temporarily deleting a #undef line from hedley_undef.hpp to confirm the
generated test actually fails on a real leak.

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

* Fix REUSE compliance failure in gen_hedley_undef_check.cmake

The generated file's embedded banner contains the literal text
'SPDX-License-Identifier: MIT' as part of the *content* being written
to hedley_undef.hpp, not as this .cmake script's own REUSE header (it
is already covered by the blanket 'Files: *' rule in .reuse/dep5).

The reuse tool matched that embedded line as an SPDX tag for the
script itself and failed to parse the trailing 'MIT\n")' as a valid
SPDX License Expression, breaking ci_reuse_compliance. Wrap the
embedded banner in REUSE-IgnoreStart/REUSE-IgnoreEnd comments, as
recommended by the tool's own diagnostic output.

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

---------

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
 2026-09-09 09:48:17 +02:00
Niels Lohmann 855f511db4 Fix stale Clang -Weverything suppression comments; drop -Wno-missing-noreturn (#5250) 
* Fix stale Clang -Weverything suppression comments; eliminate -Wno-missing-noreturn

cmake/clang_flags.cmake claimed -Wno-unsafe-buffer-usage was needed only
for Doctest and that -Wno-missing-noreturn had "no way to silence...
otherwise" (PR #4871, which never actually attempted a source fix).
Neither held up under investigation (todo 130):

- -Wno-unsafe-buffer-usage is pervasive (208 distinct sites across 19
  files measured with clang trunk in silkeh/clang:dev), spanning the
  library's own low-level numeric/buffer code (to_chars, serializer,
  lexer, binary reader/writer, input adapters, json_pointer) as well as
  vendored Doctest itself (96 of the 208 sites). A source-level fix is
  not feasible at this scale; the comment now says so instead of
  blaming Doctest alone.

- -Wno-missing-noreturn had exactly two real trigger sites, both
  genuinely and unconditionally non-returning: a test-only throwing
  allocator (tests/src/unit-allocator.cpp) and, previously undiscovered,
  wide_string_input_adapter::get_elements<T>() in
  include/nlohmann/detail/input/input_adapters.hpp. Verified this isn't
  a wider pattern by checking all 160 JSON_THROW call sites in the
  library for functions whose entire body is an unconditional throw.
  Annotated both ([[noreturn]] in the test file, since JSON_HEDLEY_NO_RETURN
  is #undef'd by the time test code runs; JSON_HEDLEY_NO_RETURN in the
  library file, its first real use anywhere in the codebase) and
  dropped the suppression entirely.

single_include/nlohmann/json.hpp regenerated via `make amalgamate`;
`make check-amalgamation` passes.

Verified in Docker (silkeh/clang:dev, matching the ci_static_analysis_clang
CI job): baseline builds clean, and the full 194-target test suite builds
with zero warnings under the corrected CLANG_CXXFLAGS (-Wno-missing-noreturn
no longer in the list). Also sanity-compiled and ran unit-allocator.cpp and
unit-wstring.cpp on host Apple Clang to confirm behavior is unchanged.

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

* Fix MSVC C4702 warning caused by JSON_HEDLEY_NO_RETURN on get_elements()

PR #5250 annotated wide_string_input_adapter::get_elements<T>() with
JSON_HEDLEY_NO_RETURN (it unconditionally throws). On MSVC this expands to
__declspec(noreturn), and MSVC correctly determined that the code following
its call in binary_reader.hpp is unreachable for that instantiation, firing
C4702 under /W4 /WX in the msvc, msvc-vs2026, and msvc-arm64 Debug jobs.

Clang doesn't flag this case, so the Docker verification for #5250 (which
only checked Clang -Weverything) didn't catch it.

This is the same warning class already tolerated for Release builds since
PR #5216, where MSVC's optimizer independently found the same dead code
after /Od was removed. Extend that existing /wd4702 suppression to Debug
builds too, instead of reverting the noreturn annotation.

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

---------

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
 2026-07-09 20:18:28 +02:00
Niels Lohmann 8d7e0046f4 Add std::format and fmt support (#5224) 
* ✨ add std::format and fmt support

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

* ♻️ reorganize PR

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

* 💚 fix build

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

* 💚 fix build

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

* 💚 fix build

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

---------

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
 2026-07-02 15:59:36 +02:00
Niels Lohmann 272411c5e6 Overwork project infrastructure (#5218) 
* 📝 overwork project infrastructure

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

* 🐛 fix GCC16 issue

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

* 🐛 fix GCC16 issue

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

* 🐛 only build module for GCC

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

* 🐛 fix build

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

* 📝 fix documentation

Closes #5012: fix the error_handler_t::ignore wording

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

* 📝 fix documentation

Closes #4354: fix "Custom data source" example

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

---------

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
 2026-06-30 18:09:06 +02:00
risa2000andRichard Musil c37f82e563 Remove forced /Od flag in MSVC Release build. (#5216) 
Add /wd4702 to disable warning C4702: unreachable code in MSVC Release build.

Signed-off-by: Richard Musil <risa2000x@gmail.com>
Co-authored-by: Richard Musil <risa2000x@gmail.com>
 2026-06-28 17:39:07 +02:00
risa2000andRichard Musil 11aa5f944d Make std::filesystem::path conversion to/from UTF-8 encoded string explicit (#4631) 
* Make std::filesystem::path conversion to/from UTF-8 encoded JSON string explicit.

Signed-off-by: Richard Musil <risa2000x@gmail.com>

* Experimental: Changing C++ standard detection logic to accommodate potential corner cases.

Signed-off-by: Richard Musil <risa2000x@gmail.com>

* Drop C++ standard tests for compilers which do not implement required features.

Signed-off-by: Richard Musil <risa2000x@gmail.com>

* Drop C++ standard tests for MSVC versions which do not implement required features.

Signed-off-by: Richard Musil <risa2000x@gmail.com>

---------

Signed-off-by: Richard Musil <risa2000x@gmail.com>
Co-authored-by: Richard Musil <risa2000x@gmail.com>
 2025-04-04 10:59:23 +02:00
Craig Scott fac07e22c5 Accept NEW CMake policies up to CMake 3.14 (#4112) 
Starting with CMake 3.27, deprecation warnings are issued
when asking for policy settings for CMake 3.4 or earlier.
The cmake_minimum_required() command accepts a version
range, which allows NEW policy settings up to the upper end
of that range to be used, but without raising the minimum
CMake version above the bottom of that range. This means
NEW policy settings will be used where available, without
requiring them. This change updates the project's
cmake_minimum_required() calls to use a version range to
extend the upper policy version to 3.14 where it wasn't already
at that version or higher. This prevents the deprecation warning
from CMake 3.27, and gives breathing space before a future
CMake release will start issuing similar deprecation warnings
again.
 2023-09-25 09:31:26 +02:00
Raphael Grimmandbarcode 233d233439 Fix CI issues (#3906) 
Co-authored-by: barcode <barcode@example.com>
 2023-01-31 19:23:37 +01:00
Florian Albrechtskirchinger 0c7a18374c Reimplement value() access functions (#3663) 
* Reimplement value() access functions

* Merges the 'const char *' with the 'ValueType &&' overloads.
* Fixes ambiguities when default value is 0.
* Fixes 'no matching function' error when specifying ValueType template
  parameter.
* Fixes incorrect template parameter order in previous overloads.

* Add additional value() tests

* Make JSON_MultipleHeaders visible to unit tests

Define the macro JSON_TEST_USING_MULTIPLE_HEADERS to 0/1 depending on
JSON_MultipleHeaders.

* Add type_traits unit test

* Update documentation
 2022-08-07 13:54:55 +02:00
Florian Albrechtskirchinger a2578d1d50 Disable exceptions on ICPC (#3621) 2022-07-30 22:10:49 +02:00
Florian Albrechtskirchinger d909f80960 Add versioned, ABI-tagged inline namespace and namespace macros (#3590) 
* 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_*
 2022-07-30 21:59:13 +02:00
Florian Albrechtskirchinger 5520761433 Fix Unicode test timeout (for real this time!) (#3614) 
* Use TIMEOUT instead of TIMEOUT_AFTER_MATCH to set test timeout
 2022-07-28 21:51:11 +02:00
Florian Albrechtskirchinger 71bdaf5748 Enable overriding test properties and set Unicode test timeouts (#3580) 
* CMake: Add TEST_PROPERTIES to json_test_set_test_options()

Allow overriding test properties via json_test_set_test_options().

* CMake: Set timeouts for Unicode tests
 2022-07-17 13:43:32 +02:00
Florian Albrechtskirchinger e80945da2c CI: Enable 32bit unit test (3) (#3532) 
* Enable JSON_MultipleHeaders by default

* CI: Add single-header build

* CI: Enable 32bit unit test

* Fix "-Wuseless-cast" warnings

* Remove coverage exclusion

* Fix 32bit test case
 2022-06-14 08:54:12 +02:00
Florian AlbrechtskirchingerandQianqian Fang 48a102c2c5 Fix ndarray dimension signedness, fix ndarray length overflow (2); add 32bit unit test (#3523) 
* Fix ndarray dimension signness, fix ndarray length overflow, close #3519

* detect size overflow in ubjson and bjdata

* force reformatting

* Fix MSVC compiler warning

* Add value_in_range_of trait

* Use value_in_range_of trait

* Correct 408 parse_errors to out_of_range

* Add 32bit unit test

The test can be enabled by setting JSON_32bitTest=ON.

* Exclude unreachable lines from coverage

Certain lines are unreachable in 64bit builds.

Co-authored-by: Qianqian Fang <fangqq@gmail.com>
 2022-06-09 08:22:58 +02:00
Florian Albrechtskirchinger 6b97599a27 Fix C++20/gcc-12 issues (Part 2) (#3446) 
* Add C++20 3-way comparison operator and fix broken comparisons

Fixes #3207.
Fixes #3409.

* Fix iterators to meet (more) std::ranges requirements

Fixes #3130.
Related discussion: #3408

* Add note about CMake standard version selection to unit tests

Document how CMake chooses which C++ standard version to use when
building tests.

* Update documentation

* CI: add legacy discarded value comparison

* Fix internal linkage errors when building a module
 2022-05-29 13:08:06 +02:00
Niels LohmannChaoqi ZhangQianqian Fang
 		6a7392058e Complete documentation for 3.11.0 (#3464) 
* 👥 update contributor and sponsor list

* 🚧 document BJData format

* 🚧 document BJData format

* 📝 clarified documentation of [json.exception.parse_error.112]

* ✏️ adjust titles

* 📝 add more examples

* 🚨 adjust warnings for index.md files

* 📝 add more examples

* 🔥 remove example for deprecated code

* 📝 add missing enum entry

* 📝 overwork table for binary formats

* ✅ add test to create table for binary formats

* 📝 fix wording in example

* 📝 add more examples

* Update iterators.md (#3481)

* ✨ add check for overloads to linter #3455

* 👥 update contributor list

* 📝 add more examples

* 📝 fix documentation

* 📝 add more examples

* 🎨 fix indentation

* 🔥 remove example for destructor

* 📝 overwork documentation

* Updated BJData documentation, #3464 (#3493)

* update bjdata.md for #3464

* Minor edit

* Fix URL typo

* Add info on demoting ND array to a 1-D optimized array when singleton dimension

Co-authored-by: Chaoqi Zhang <prncoprs@163.com>
Co-authored-by: Qianqian Fang <fangqq@gmail.com>
 2022-05-17 13:08:56 +02:00
Florian AlbrechtskirchingerandNiels Lohmann bdc21ad1a7 Add build step for ICPC (with fixes) (#3465) 
* ⬆️ Doctest 2.4.7

* 👷 add CI step for ICPC

* 👷 add CI step for ICPC

* 👷 add CI step for ICPC

* ⬇️ downgrade to Doctest 2.4.6

* 👷 add CI step for ICPC

* 👷 add CI step for ICPC

* 👷 add CI step for ICPC

* 👷 add CI step for ICPC

* 👷 add CI step for ICPC

* 🔇 suppress warning #2196: routine is both "inline" and "noinline"

* Re-enable <filesystem> detection on ICPC

* Limit regression test for #3070 to Clang and GCC >=8.4

* Disable deprecation warnings on ICPC

* Disable regression test for #1647 on ICPC (C++20)

* Fix compilation failure of regression test for #3077 on ICPC

* Disable wstring unit test on ICPC

Fixes:
  error 913: invalid multibyte character sequence

* Add ICPC to README

Co-authored-by: Niels Lohmann <mail@nlohmann.me>
 2022-05-01 22:46:45 +02:00
Niels Lohmann b21c345179 Reorganize directories (#3462) 
* 🚚 move files
* 🚚 rename doc folder to docs
* 🚚 rename test folder to tests
 2022-05-01 09:41:50 +02:00