dynarmic/externals/catch/docs/release-notes.md
Yang Liu b372dc6157 externals: Update catch2 to v3.5.0
Merge commit 'ba06a404d1437c73ed3ba100d04a217fa69545b8'
2023-12-31 14:00:46 +08:00

80 KiB

Release notes

Contents
3.5.0
3.4.0
3.3.2
3.3.1
3.3.0
3.2.1
3.2.0
3.1.1
3.1.0
3.0.1
2.13.7
2.13.6
2.13.5
2.13.4
2.13.3
2.13.2
2.13.1
2.13.0
2.12.4
2.12.3
2.12.2
2.12.1
2.12.0
2.11.3
2.11.2
2.11.1
2.11.0
2.10.2
2.10.1
2.10.0
2.9.2
2.9.1
2.9.0
2.8.0
2.7.2
2.7.1
2.7.0
2.6.1
2.6.0
2.5.0
2.4.2
2.4.1
2.4.0
2.3.0
2.2.3
2.2.2
2.2.1
2.2.0
2.1.2
2.1.1
2.1.0
2.0.1
Older versions
Even Older versions

3.5.0

Improvements

  • Introduced CATCH_CONFIG_PREFIX_MESSAGES to prefix only logging macros (#2544)
    • This means INFO, UNSCOPED_INFO, WARN and CAPTURE.
  • Section hints in static analysis mode are now const
    • This prevents Clang-Tidy from complaining about misc-const-correctness.
  • from_range generator supports C arrays and ranges that require ADL (#2737)
  • Stringification support for std::optional now also includes std::nullopt (#2740)
  • The Console reporter flushes output after writing benchmark runtime estimate.
    • This means that you can immediately see for how long the benchmark is expected to run.
  • Added workaround to enable compilation with ICC 19.1 (#2551, #2766)
  • Compiling Catch2 for XBox should work out of the box (#2772)
    • Catch2 should automatically disable getenv when compiled for XBox.
  • Compiling Catch2 with exceptions disabled no longer triggers Wunused-function (#2726)
  • random Generators for integral types are now reproducible across different platforms
    • Unlike <rando>, Catch2's generators also support 1 byte integral types (char, bool, ...)
  • random Generators for float and double are now reproducible across different platforms
    • long double varies across different platforms too much to be reproducible
    • This guarantee applies only to platforms with IEEE 754 floats.

Fixes

  • UDL declaration inside Catch2 are now strictly conforming to the standard
    • operator "" _a is UB, operator ""_a is fine. Seriously.
  • Fixed CAPTURE tests failing to compile in C++23 mode (#2744)
  • Fixed missing include in catch_message.hpp (#2758)
  • Fixed CHECK_ELSE suppressing failure from uncaught exceptions(#2723)

Miscellaneous

  • The documentation for specifying which tests to run through commandline has been completely rewritten (#2738)
  • Fixed installation when building Catch2 with meson (#2722, #2742)
  • Fixed catch_discover_tests when using custom reporter and PRE_TEST discovery mode (#2747)
  • catch_discover_tests supports multi-config CMake generator in PRE_TEST discovery mode (#2739, #2746)

3.4.0

Improvements

  • VectorEquals supports elements that provide only == and not != (#2648)
  • Catch2 supports compiling with IAR compiler (#2651)
  • Various small internal performance improvements
  • Various small internal compilation time improvements
  • XMLReporter now reports location info for INFO and WARN (#1251)
    • This bumps up the xml format version to 3
  • Documented that SKIP in generator constructor can be used to handle empty generator (#1593)
  • Added experimental static analysis support to TEST_CASE and SECTION macros (#2681)
    • The two macros are redefined in a way that helps the SA tools reason about the possible paths through a test case with sections.
    • The support is controlled by the CATCH_CONFIG_EXPERIMENTAL_STATIC_ANALYSIS_SUPPORT option and autodetects clang-tidy and Coverity.
  • *_THROWS, *_THROWS_AS, etc now suppress warning coming from __attribute__((warn_unused_result)) on GCC (#2691)
    • Unlike plain [[nodiscard]], this warning is not silenced by void cast. WTF GCC?

Fixes

  • Fixed assertionStarting events being sent after the expr is evaluated (#2678)
  • Errors in TEST_CASE tags are now reported nicely (#2650)

Miscellaneous

  • Bunch of improvements to catch_discover_tests
    • Added DISCOVERY_MODE option, so the discovery can happen either post build or pre-run.
    • Fixed handling of semicolons and backslashes in test names (#2674, #2676)
  • meson build can disable building tests (#2693)
  • meson build properly sets meson version 0.54.1 as the minimal supported version (#2688)

3.3.2

Improvements

  • Further reduced allocations
    • The compact, console, TAP and XML reporters perform less allocations in various cases
    • Removed 1 allocation per entered SECTION/TEST_CASE.
    • Removed 2 allocations per test case exit, if stdout/stderr is captured
  • Improved performance
    • Section tracking is 10%-25% faster than in v3.3.0
    • Assertion handling is 5%-10% faster than in v3.3.0
    • Test case registration is 1%-2% faster than in v3.3.0
    • Tiny speedup for registering listeners
    • Tiny speedup for CAPTURE, TEST_CASE_METHOD, METHOD_AS_TEST_CASE, and TEMPLATE_LIST_TEST_* macros.
  • Contains, RangeEquals and UnorderedRangeEquals matchers now support ranges with iterator + sentinel pair
  • Added IsNaN matcher
    • Unlike REQUIRE(isnan(x)), REQUIRE_THAT(x, IsNaN()) shows you the value of x.
  • Suppressed declared_but_not_referenced warning for NVHPC (#2637)

Fixes

  • Fixed performance regression in section tracking introduced in v3.3.1
    • Extreme cases would cause the tracking to run about 4x slower than in 3.3.0

3.3.1

Improvements

  • Reduced allocations and improved performance
    • The exact improvements are dependent on your usage of Catch2.
    • For example running Catch2's SelfTest binary performs 8k less allocations.
    • The main improvement comes from smarter handling of SECTIONs, especially sibling SECTIONs

3.3.0

Improvements

  • Added MessageMatches exception matcher (#2570)
  • Added RangeEquals and UnorderedRangeEquals generic range matchers (#2377)
  • Added SKIP macro for skipping tests from within the test body (#2360)
    • All built-in reporters have been extended to handle it properly, whether your custom reporter needs changes depends on how it was written
    • skipTest reporter event is unrelated to this, and has been deprecated since it has practically no uses
  • Restored support for PPC Macs in the break-into-debugger functionality (#2619)
  • Made our warning suppression compatible with CUDA toolkit pre 11.5 (#2626)
  • Cleaned out some static analysis complaints

Fixes

  • Fixed macro redefinition warning when NVCC was reporting as MSVC (#2603)
  • Fixed throws in generator constructor causing the whole binary to abort (#2615)
    • Now it just fails the test
  • Fixed missing transitive include with libstdc++13 (#2611)

Miscellaneous

  • Improved support for dynamic library build with non-MSVC compilers on Windows (#2630)
  • When used as a subproject, Catch2 keeps its generated header in a separate directory from the main project (#2604)

3.2.1

Improvements

  • Fix the reworked decomposer to work with older (pre 9) GCC versions (#2571)
    • This required more significant changes to properly support C++20, there might be bugs.

3.2.0

Improvements

  • Catch2 now compiles on PlayStation (#2562)
  • Added CATCH_CONFIG_GETENV compile-time toggle (#2562)
    • This toggle guards whether Catch2 calls std::getenv when reading env variables
  • Added support for more Bazel test environment variables
    • TESTBRIDGE_TEST_ONLY is now supported (#2490)
    • Sharding variables, TEST_SHARD_INDEX, TEST_TOTAL_SHARDS, TEST_SHARD_STATUS_FILE, are now all supported (#2491)
  • Bunch of small tweaks and improvements in reporters
    • The TAP and SonarQube reporters output the used test filters
    • The XML reporter now also reports the version of its output format
    • The compact reporter now uses the same summary output as the console reporter (#878, #2554)
  • Added support for asserting on types that can only be compared with literal 0 (#2555)
    • A canonical example is C++20's std::*_ordering types, which cannot be compared with an int variable, only 0
    • The support extends to any type with this property, not just the ones in stdlib
    • This change imposes 2-3% slowdown on compiling files that are heavy on REQUIRE and friends
    • This required significant rewrite of decomposition, there might be bugs
  • Simplified internals of matcher related macros
    • This provides about ~2% speed up compiling files that are heavy on REQUIRE_THAT and friends

Fixes

  • Cleaned out some warnings and static analysis issues
    • Suppressed -Wcomma warning rarely occurring in templated test cases (#2543)
    • Constified implementation details in INFO (#2564)
    • Made MatcherGenericBase copy constructor const (#2566)
  • Fixed serialization of test filters so the output roundtrips
    • This means that e.g. ./tests/SelfTest "aaa bbb", [approx] outputs Filters: "aaa bbb",[approx]

Miscellaneous

  • Catch2's build no longer leaks -ffile-prefix-map setting to dependees (#2533)

3.1.1

Improvements

  • Added Catch::getSeed function that user code can call to retrieve current rng-seed
  • Better detection of compiler support for -ffile-prefix-map (#2517)
  • Catch2's shared libraries now have SOVERSION set (#2516)
  • catch2/catch_all.hpp convenience header no longer transitively includes windows.h (#2432, #2526)

Fixes

  • Fixed compilation on Universal Windows Platform
  • Fixed compilation on VxWorks (#2515)
  • Fixed compilation on Cygwin (#2540)
  • Remove unused variable in reporter registration (#2538)
  • Fixed some symbol visibility issues with dynamic library on Windows (#2527)
  • Suppressed -Wuseless-cast warnings in REQUIRE_THROWS* macros (#2520, #2521)
    • This was triggered when the potentially throwing expression evaluates to void
  • Fixed "warning: storage class is not first" with nvc++ (#2533)
  • Fixed handling of DL_PATHS argument to catch_discover_tests on MacOS (#2483)
  • Suppressed *-avoid-c-arrays clang-tidy warning in TEMPLATE_TEST_CASE (#2095, #2536)

Miscellaneous

  • Fixed CMake install step for Catch2 build as dynamic library (#2485)
  • Raised minimum CMake version to 3.10 (#2523)
    • Expect the minimum CMake version to increase once more in next few releases.
  • Whole bunch of doc updates and fixes
    • #1444, #2497, #2547, #2549, and more
  • Added support for building Catch2 with Meson (#2530, #2539)

3.1.0

Improvements

  • Improved suppression of -Wparentheses for older GCCs
    • Turns out that even GCC 9 does not properly handle _Pragmas in the C++ frontend.
  • Added type constraints onto random generator (#2433)
    • These constraints copy what the standard says for the underlying std::uniform_int_distribution
  • Suppressed -Wunused-variable from nvcc (#2306, #2427)
  • Suppressed -Wunused-variable from MinGW (#2132)
  • Added All/Any/NoneTrue range matchers (#2319)
    • These check that all/any/none of boolean values in a range are true.
  • The JUnit reporter now normalizes classnames from C++ namespaces to Java-like namespaces (#2468)
    • This provides better support for other JUnit based tools.
  • The Bazel support now understands BAZEL_TEST environment variable (#2459)
    • The CATCH_CONFIG_BAZEL_SUPPORT configuration option is also still supported.
  • Returned support for compiling Catch2 with GCC 5 (#2448)
    • This required removing inherited constructors from Catch2's internals.
    • I recommend updating to a newer GCC anyway.
  • catch_discover_tests now has a new options for setting library load path(s) when running the Catch2 binary (#2467)

Fixes

  • Fixed crash when listing listeners without any registered listeners (#2442)
  • Fixed nvcc compilation error in constructor benchmarking helper (#2477)
  • Catch2's CMakeList supports pre-3.12 CMake again (#2428)
    • The gain from requiring CMake 3.12 was very minor, but y'all should really update to newer CMake

Miscellaneous

  • Fixed SelfTest build on MinGW (#2447)
  • The in-repo conan recipe exports the CMake helper (#2460)
  • Added experimental CMake script to showcase using test case sharding together with CTest
    • Compared to catch_discover_tests, it supports very limited number of options and customization
  • Added documentation page on best practices when running Catch2 tests
  • Catch2 can be built as a dynamic library (#2397, #2398)
    • Note that Catch2 does not have visibility annotations, and you are responsible for ensuring correct visibility built into the resulting library.

3.0.1

Catch2 now uses statically compiled library as its distribution model. This also means that to get all of Catch2's functionality in a test file, you have to include multiple headers.

You probably want to look into the migration docs, which were written to help people coming from v2.x.x versions to the v3 releases.

FAQ

  • Why is Catch2 moving to separate headers?
    • The short answer is future extensibility and scalability. The long answer is complex and can be found on my blog, but at the most basic level, it is that providing single-header distribution is at odds with providing variety of useful features. When Catch2 was distributed in a single header, adding a new Matcher would cause overhead for everyone, but was useful only to a subset of users. This meant that the barrier to entry for new Matchers/Generators/etc is high in single header model, but much smaller in the new model.
  • Will Catch2 again distribute single-header version in the future?
    • No. But we do provide sqlite-style amalgamated distribution option. This means that you can download just 1 .cpp file and 1 header and place them next to your own sources. However, doing this has downsides similar to using the catch_all.hpp header.
  • Why the big breaking change caused by replacing catch.hpp with catch_all.hpp?
    • The convenience header catch_all.hpp exists for two reasons. One of them is to provide a way for quick migration from Catch2, the second one is to provide a simple way to test things with Catch2. Using it for migration has one drawback in that it is big. This means that including it will cause significant compile time drag, and so using it to migrate should be a conscious decision by the user, not something they can just stumble into unknowingly.

(Potentially) Breaking changes

  • Catch2 now uses statically compiled library as its distribution model
    • Including catch.hpp no longer works
  • Catch2 now uses C++14 as the minimum support language version
  • ANON_TEST_CASE has been removed, use TEST_CASE with no arguments instead (#1220)
  • --list* commands no longer have non-zero return code (#1410)
  • --list-test-names-only has been removed (#1190)
    • You should use verbosity-modifiers for --list-tests instead
  • --list* commands are now piped through the reporters
    • The top-level reporter interface provides default implementation that works just as the old one
    • XmlReporter outputs a machine-parseable XML
  • TEST_CASE description support has been removed
    • If the second argument has text outside tags, the text will be ignored.
  • Hidden test cases are no longer included just because they don't match an exclusion tag
    • Previously, a TEST_CASE("A", "[.foo]") would be included by asking for ~[bar].
  • PredicateMatcher is no longer type erased.
    • This means that the type of the provided predicate is part of the PredicateMatcher's type
  • SectionInfo no longer contains section description as a member (#1319)
    • You can still write SECTION("ShortName", "Long and wordy description"), but the description is thrown away
    • The description type now must be a const char* or be implicitly convertible to it
  • The [!hide] tag has been removed.
    • Use [.] or [.foo] instead.
  • Lvalues of composed matchers cannot be composed further
  • Uses of REGISTER_TEST_CASE macro need to be followed by a semicolon
    • This does not change TEST_CASE and friends in any way
  • IStreamingReporter::IsMulti member function was removed
    • This is very unlikely to actually affect anyone, as it was default-implemented in the interface, and only used internally
  • Various classes not designed for user-extension have been made final
    • ListeningReporter is now final
    • Concrete Matchers (e.g. UnorderedEquals vector matcher) are now final
    • All Generators are now final
  • Matcher namespacing has been redone
    • Matcher types are no longer in deeply nested namespaces
    • Matcher factory functions are no longer brought into Catch namespace
    • This means that all public-facing matcher-related functionality is now in Catch::Matchers namespace
  • Defining CATCH_CONFIG_MAIN will no longer create main in that TU.
    • Link with libCatch2Main.a, or the proper CMake/pkg-config target
    • If you want to write custom main, include catch2/catch_session.hpp
  • CATCH_CONFIG_EXTERNAL_INTERFACES has been removed.
    • You should instead include the appropriate headers as needed.
  • CATCH_CONFIG_IMPL has been removed.
    • The implementation is now compiled into a static library.
  • Event Listener interface has changed
    • TestEventListenerBase was renamed to EventListenerBase
    • EventListenerBase now directly derives from IStreamingReporter, instead of deriving from StreamingReporterBase
  • GENERATE decays its arguments (#2012, #2040)
    • This means that str in auto str = GENERATE("aa", "bb", "cc"); is inferred to char const* rather than const char[2].
  • --list-* flags write their output to file specified by the -o flag
  • Many changes to reporter interfaces
    • With the exception of the XmlReporter, the outputs of first party reporters should remain the same
    • New pair of events were added
    • One obsolete event was removed
    • The base class has been renamed
    • The built-in reporter class hierarchy has been redone
  • Catch2 generates a random seed if one hasn't been specified by the user
  • The short flag for --list-tests, -l, has been removed.
    • This is not a commonly used flag and does not need to use up valuable single-letter space.
  • The short flag for --list-tags, -t, has been removed.
    • This is not a commonly used flag and does not need to use up valuable single-letter space.
  • The --colour option has been replaced with --colour-mode option

Improvements

  • Matchers have been extended with the ability to use different signatures of match (#1307, #1553, #1554, #1843)
    • This includes having templated match member function
    • See the rewritten Matchers documentation for details
    • Catch2 currently provides some generic matchers, but there should be more before final release of v3
      • IsEmpty, SizeIs which check that the range has specific properties
      • Contains, which checks whether a range contains a specific element
      • AllMatch, AnyMatch, NoneMatch range matchers, which apply matchers over a range of elements
  • Significant compilation time improvements
    • including catch_test_macros.hpp is 80% cheaper than including catch.hpp
  • Some runtime performance optimizations
    • In all tested cases the v3 branch was faster, so the table below shows the speedup of v3 to v2 at the same task
task debug build release build
Run 1M REQUIRE(true) 1.10 ± 0.01 1.02 ± 0.06
Run 100 tests, 3^3 sections, 1 REQUIRE each 1.27 ± 0.01 1.04 ± 0.01
Run 3k tests, no names, no tags 1.29 ± 0.01 1.05 ± 0.01
Run 3k tests, names, tags 1.49 ± 0.01 1.22 ± 0.01
Run 1 out of 3k tests no names, no tags 1.68 ± 0.02 1.19 ± 0.22
Run 1 out of 3k tests, names, tags 1.79 ± 0.02 2.06 ± 0.23
  • POSIX platforms use gmtime_r, rather than gmtime when constructing a date string (#2008, #2165)
  • --list-* flags write their output to file specified by the -o flag (#2061, #2163)
  • Approx::operator() is now properly const
  • Catch2's internal helper variables no longer use reserved identifiers (#578)
  • --rng-seed now accepts string "random-device" to generate random seed using std::random_device
  • Catch2 now supports test sharding (#2257)
    • You can ask for the tests to be split into N groups and only run one of them.
    • This greatly simplifies parallelization of tests in a binary through external runner.
  • The embedded CLI parser now supports repeatedly callable lambdas
    • A lambda-based option parser can opt into being repeatedly specifiable.
  • Added STATIC_CHECK macro, similar to STATIC_REQUIRE (#2318)
    • When deferred tu runtime, it behaves like CHECK, and not like REQUIRE.
  • You can have multiple tests with the same name, as long as other parts of the test identity differ (#1915, #1999, #2175)
    • Test identity includes test's name, test's tags and test's class name if applicable.
  • Added new warning, UnmatchedTestSpec, to error on test specs with no matching tests
  • The -w, --warn warning flags can now be provided multiple times to enable multiple warnings
  • The case-insensitive handling of tags is now more reliable and takes up less memory
  • Test case and assertion counting can no longer reasonably overflow on 32 bit systems
    • The count is now kept in uint64_t on all platforms, instead of using size_t type.
  • The -o, --out output destination specifiers recognize - as stdout
    • You have to provide it as --out=- to avoid CLI error about missing option
    • The new reporter specification also recognizes - as stdout
  • Multiple reporters can now run at the same time and write to different files (#1712, #2183)
    • To support this, the -r, --reporter flag now also accepts optional output destination
    • For full overview of the semantics of using multiple reporters, look into the reporter documentation
    • To enable the new syntax, reporter names can no longer contain ::.
  • Console colour support has been rewritten and significantly improved
    • The colour implementation based on ANSI colour codes is always available
    • Colour implementations respect their associated stream
      • previously e.g. Win32 impl would change console colour even if Catch2 was writing to a file
    • The colour API is resilient against changing evaluation order of expressions
    • The associated CLI flag and compile-time configuration options have changed
      • For details see the docs for command-line and compile-time Catch2 configuration
  • Added a support for Bazel integration with XML_OUTPUT_FILE env var (#2399)
    • This has to be enabled during compilation.
  • Added --skip-benchmarks flag to run tests without any BENCHMARKs (#2392, #2408)
  • Added option to list all listeners in the binary via --list-listeners

Fixes

  • The INFO macro no longer contains superfluous semicolon (#1456)
  • The --list* family of command line flags now return 0 on success (#1410, #1146)
  • Various ways of failing a benchmark are now counted and reporter properly
  • The ULP matcher now handles comparing numbers with different signs properly (#2152)
  • Universal ADL-found operators should no longer break decomposition (#2121)
  • Reporter selection is properly case-insensitive
    • Previously it forced lower cased name, which would fail for reporters with upper case characters in name
  • The cumulative reporter base stores benchmark results alongside assertion results
  • Catch2's SE handling should no longer interferes with ASan on Windows (#2334)
  • Fixed Windows console colour handling for tests that redirect stdout (#2345)
  • Fixed issue with the random generators returning the same value over and over again

Other changes

  • CATCH_CONFIG_DISABLE_MATCHERS no longer exists.
    • If you do not want to use Matchers in a TU, do not include their header.
  • CATCH_CONFIG_ENABLE_CHRONO_STRINGMAKER no longer exists.
    • StringMaker specializations for <chrono> are always provided
  • Catch2's CMake now provides 2 targets, Catch2 and Catch2WithMain.
    • Catch2 is the statically compiled implementation by itself
    • Catch2WithMain also links in the default main
  • Catch2's pkg-config integration also provides 2 packages
    • catch2 is the statically compiled implementation by itself
    • catch2-with-main also links in the default main
  • Passing invalid test specifications passed to Catch2 are now reported before tests are run, and are a hard error.
  • Running 0 tests (e.g. due to empty binary, or test spec not matching anything) returns non-0 exit code
    • Flag --allow-running-no-tests overrides this behaviour.
    • NoTests warning has been removed because it is fully subsumed by this change.
  • Catch2's compile-time configuration options (CATCH_CONFIG_FOO) can be set through CMake options of the same name
    • They use the same semantics as C++ defines, including the CATCH_CONFIG_NO_FOO overrides,
      • -DCATCH_CONFIG_DEFAULT_REPORTER=compact changes default reporter to "compact"
      • -DCATCH_CONFIG_NO_ANDROID_LOGWRITE=ON forces android logwrite to off
      • -DCATCH_CONFIG_ANDROID_LOGWRITE=OFF does nothing (the define will not exist)

2.13.7

Fixes

  • Added missing <iterator> include in benchmarking. (#2231)
  • Fixed noexcept build with benchmarking enabled (#2235)
  • Fixed build for compilers with C++17 support but without C++17 library support (#2195)
  • JUnit only uses 3 decimal places when reporting durations (#2221)
  • !mayfail tagged tests are now marked as skipped in JUnit reporter output (#2116)

2.13.6

Fixes

  • Disabling all signal handlers no longer breaks compilation (#2212, #2213)

Miscellaneous

  • catch_discover_tests should handle escaped semicolon (;) better (#2214, #2215)

2.13.5

Improvements

  • Detection of MAC and IPHONE platforms has been improved (#2140, #2157)
  • Added workaround for bug in XLC 16.1.0.1 (#2155)
  • Add detection for LCC when it is masquerading as GCC (#2199)
  • Modified posix signal handling so it supports newer libcs (#2178)
    • MINSIGSTKSZ was no longer usable in constexpr context.

Fixes

  • Fixed compilation of benchmarking when min and max macros are defined (#2159)
    • Including windows.h without NOMINMAX remains a really bad idea, don't do it

Miscellaneous

  • The check whether Catch2 is being built as a subproject is now more reliable (#2202, #2204)
    • The problem was that if the variable name used internally was defined the project including Catch2 as subproject, it would not be properly overwritten for Catch2's CMake.

2.13.4

Improvements

  • Improved the hashing algorithm used for shuffling test cases (#2070)
    • TEST_CASEs that differ only in the last character should be properly shuffled
    • Note that this means that v2.13.4 gives you a different order of test cases than 2.13.3, even given the same seed.

Miscellaneous

  • Deprecated ParseAndAddCatchTests CMake integration (#2092)
    • It is impossible to implement it properly for all the different test case variants Catch2 provides, and there are better options provided.
    • Use catch_discover_tests instead, which uses runtime information about available tests.
  • Fixed bug in catch_discover_tests that would cause it to fail when used in specific project structures (#2119)
  • Added Bazel build file
  • Added an experimental static library target to CMake

2.13.3

Fixes

  • Fixed possible infinite loop when combining generators with section filter (-c option) (#2025)

Miscellaneous

  • Fixed ParseAndAddCatchTests not finding TEST_CASEs without tags (#2055, #2056)
  • ParseAndAddCatchTests supports CMP0110 policy for changing behaviour of add_test (#2057)
    • This was the shortlived change in CMake 3.18.0 that temporarily broke ParseAndAddCatchTests

2.13.2

Improvements

  • Implemented workaround for AppleClang shadowing bug (#2030)
  • Implemented workaround for NVCC ICE (#2005, #2027)

Fixes

  • Fixed detection of std::uncaught_exceptions support under non-msvc platforms (#2021)
  • Fixed the experimental stdout/stderr capture under Windows (#2013)

Miscellaneous

  • catch_discover_tests has been improved significantly (#2023, #2039)
    • You can now specify which reporter should be used
    • You can now modify where the output will be written
    • WORKING_DIRECTORY setting is respected
  • ParseAndAddCatchTests now supports TEMPLATE_TEST_CASE macros (#2031)
  • Various documentation fixes and improvements (#2022, #2028, #2034)

2.13.1

Improvements

  • ParseAndAddCatchTests handles CMake v3.18.0 correctly (#1984)
  • Improved autodetection of std::byte (#1992)
  • Simplified implementation of templated test cases (#2007)
    • This should have a tiny positive effect on its compilation throughput

Fixes

  • Automatic stringification of ranges handles sentinel ranges properly (#2004)

2.13.0

Improvements

  • GENERATE can now follow a SECTION at the same level of nesting (#1938)
    • The SECTION(s) before the GENERATE will not be run multiple times, the following ones will.
  • Added -D/--min-duration command line flag (#1910)
    • If a test takes longer to finish than the provided value, its name and duration will be printed.
    • This flag is overridden by setting -d/--duration.

Fixes

  • TAPReporter no longer skips successful assertions (#1983)

2.12.4

Improvements

  • Added support for MacOS on ARM (#1971)

2.12.3

Fixes

  • GENERATE nested in a for loop no longer creates multiple generators (#1913)
  • Fixed copy paste error breaking TEMPLATE_TEST_CASE_SIG for 6 or more arguments (#1954)
  • Fixed potential UB when handling non-ASCII characters in CLI args (#1943)

Improvements

  • There can be multiple calls to GENERATE on a single line
  • Improved fno-except support for platforms that do not provide shims for exception-related std functions (#1950)
    • E.g. the Green Hills C++ compiler
  • XmlReporter now also reports test-case-level statistics (#1958)
    • This is done via a new element, OverallResultsCases

Miscellaneous

  • Added .clang-format file to the repo (#1182, #1920)
  • Rewrote contributing docs
    • They should explain the different levels of testing and so on much better

2.12.2

Fixes

  • Fixed compilation failure if is_range ADL found deleted function (#1929)
  • Fixed potential UB in CAPTURE if the expression contained non-ASCII characters (#1925)

Improvements

  • std::result_of is not used if std::invoke_result is available (#1934)
  • JUnit reporter writes out status attribute for tests (#1899)
  • Suppressed clang-tidy's hicpp-vararg warning (#1921)
    • Catch2 was already suppressing the cppcoreguidelines-pro-type-vararg alias of the warning

2.12.1

Fixes

  • Vector matchers now support initializer list literals better

Improvements

  • Added support for ^ (bitwise xor) to CHECK and REQUIRE

2.12.0

Improvements

  • Running tests in random order (--order rand) has been reworked significantly (#1908)
    • Given same seed, all platforms now produce the same order
    • Given same seed, the relative order of tests does not change if you select only a subset of them
  • Vector matchers support custom allocators (#1909)
  • | and & (bitwise or and bitwise and) are now supported in CHECK and REQUIRE
    • The resulting type must be convertible to bool

Fixes

  • Fixed computation of benchmarking column widths in ConsoleReporter (#1885, #1886)
  • Suppressed clang-tidy's cppcoreguidelines-pro-type-vararg in assertions (#1901)
    • It was a false positive triggered by the new warning support workaround
  • Fixed bug in test specification parser handling of OR'd patterns using escaping (#1905)

Miscellaneous

  • Worked around IBM XL's codegen bug (#1907)
    • It would emit code for destructors of temporaries in an unevaluated context
  • Improved detection of stdlib's support for std::uncaught_exceptions (#1911)

2.11.3

Fixes

  • Fixed compilation error caused by lambdas in assertions under MSVC

2.11.2

Improvements

  • GCC and Clang now issue warnings for suspicious code in assertions (#1880)
    • E.g. REQUIRE( int != unsigned int ) will now issue mixed signedness comparison warning
    • This has always worked on MSVC, but it now also works for GCC and current Clang versions
  • Colorization of "Test filters" output should be more robust now
  • --wait-for-keypress now also accepts never as an option (#1866)
  • Reporters no longer round-off nanoseconds when reporting benchmarking results (#1876)
  • Catch2's debug break now supports iOS while using Thumb instruction set (#1862)
  • It is now possible to customize benchmark's warm-up time when running the test binary (#1844)
    • --benchmark-warmup-time {ms}
  • User can now specify how Catch2 should break into debugger (#1846)

Fixes

  • Fixes missing <random> include in benchmarking (#1831)
  • Fixed missing <iterator> include in benchmarking (#1874)
  • Hidden test cases are now also tagged with [!hide] as per documentation (#1847)
  • Detection of whether libc provides std::nextafter has been improved (#1854)
  • Detection of wmain no longer incorrectly looks for WIN32 macro (#1849)
    • Now it just detects Windows platform
  • Composing already-composed matchers no longer modifies the partially-composed matcher expression
    • This bug has been present for the last ~2 years and nobody reported it

2.11.1

Improvements

  • Breaking into debugger is supported on iOS (#1817)
  • google-build-using-namespace clang-tidy warning is suppressed (#1799)

Fixes

  • Clang on Windows is no longer assumed to implement MSVC's traditional preprocessor (#1806)
  • ObjectStorage now behaves properly in const contexts (#1820)
  • GENERATE_COPY(a, b) now compiles properly (#1809, #1815)
  • Some more cleanups in the benchmarking support

2.11.0

Improvements

  • JUnit reporter output now contains more details in case of failure (#1347, #1719)
  • Added SonarQube Test Data reporter (#1738)
    • It is in a separate header, just like the TAP, Automake, and TeamCity reporters
  • range generator now allows floating point numbers (#1776)
  • Reworked part of internals to increase throughput

Fixes

  • The single header version should contain full benchmarking support (#1800)
  • [.foo] is now properly parsed as [.][foo] when used on the command line (#1798)
  • Fixed compilation of benchmarking on platforms where steady_clock::period is not std::nano (#1794)

2.10.2

Improvements

  • Catch2 will now compile on platform where INFINITY is double (#1782)

Fixes

  • Warning suppressed during listener registration will no longer leak

2.10.1

Improvements

  • Catch2 now guards itself against min and max macros from windows.h (#1772)
  • Templated tests will now compile with ICC (#1748)
  • WithinULP matcher now uses scientific notation for stringification (#1760)

Fixes

  • Templated tests no longer trigger -Wunused-templates (#1762)
  • Suppressed clang-analyzer false positive in context getter (#1230, #1735)

Miscellaneous

  • CMake no longer prohibits in-tree build when Catch2 is used as a subproject (#1773, #1774)

2.10.0

Fixes

  • TEMPLATE_LIST_TEST_CASE now properly handles non-copyable and non-movable types (#1729)
  • Fixed compilation error on Solaris caused by a system header defining macro TT (#1722, #1723)
  • REGISTER_ENUM will now fail at compilation time if the registered enum is too large
  • Removed use of std::is_same_v in C++17 mode (#1757)
  • Fixed parsing of escaped special characters when reading test specs from a file (#1767, #1769)

Improvements

  • Trailing and leading whitespace in test/section specs are now ignored.
  • Writing to Android debug log now uses __android_log_write instead of __android_log_print
  • Android logging support can now be turned on/off at compile time (#1743)
    • The toggle is CATCH_CONFIG_ANDROID_LOGWRITE
  • Added a generator that returns elements of a range
    • Use via from_range(from, to) or from_range(container)
  • Added support for CRTs that do not provide std::nextafter (#1739)
    • They must still provide global nextafter{f,l,}
    • Enabled via CATCH_CONFIG_GLOBAL_NEXTAFTER
  • Special cased Approx(inf) not to match non-infinite values
    • Very strictly speaking this might be a breaking change, but it should match user expectations better
  • The output of benchmarking through the Console reporter when --benchmark-no-analysis is set is now much simpler (#1768)
  • Added a matcher that can be used for checking an exceptions message (#1649, #1728)
    • The matcher helper function is called Message
    • The exception must publicly derive from std::exception
    • The matching is done exactly, including case and whitespace
  • Added a matcher that can be used for checking relative equality of floating point numbers (#1746)
    • Unlike Approx, it considers both sides when determining the allowed margin
    • Special cases NaN and INFINITY to match user expectations
    • The matcher helper function is called WithinRel
  • The ULP matcher now allows for any possible distance between the two numbers
  • The random number generators now use Catch-global instance of RNG (#1734, #1736)
    • This means that nested random number generators actually generate different numbers

Miscellaneous

  • In-repo PNGs have been optimized to lower overhead of using Catch2 via git clone
  • Catch2 now uses its own implementation of the URBG concept
    • In the future we also plan to use our own implementation of the distributions from <random> to provide cross-platform repeatability of random results

2.9.2

Fixes

  • ChunkGenerator can now be used with chunks of size 0 (#1671)
  • Nested subsections are now run properly when specific section is run via the -c argument (#1670, #1673)
  • Catch2 now consistently uses _WIN32 to detect Windows platform (#1676)
  • TEMPLATE_LIST_TEST_CASE now support non-default constructible type lists (#1697)
  • Fixed a crash in the XMLReporter when a benchmark throws exception during warmup (#1706)
  • Fixed a possible infinite loop in CompactReporter (#1715)
  • Fixed -w NoTests returning 0 even when no tests were matched (#1449, #1683, #1684)
  • Fixed matcher compilation under Obj-C++ (#1661)

Improvements

  • RepeatGenerator and FixedValuesGenerator now fail to compile when used with bool (#1692)
    • Previously they would fail at runtime.
  • Catch2 now supports Android's debug logging for its debug output (#1710)
  • Catch2 now detects and configures itself for the RTX platform (#1693)
    • You still need to pass --benchmark-no-analysis if you are using benchmarking under RTX
  • Removed a "storage class is not first" warning when compiling Catch2 with PGI compiler (#1717)

Miscellaneous

  • Documentation now contains indication when a specific feature was introduced (#1695)
    • These start with Catch2 v2.3.0, (a bit over a year ago).
    • docs/contributing.md has been updated to provide contributors guidance on how to add these to newly written documentation
  • Various other documentation improvements
    • ToC fixes
    • Documented --order and --rng-seed command line options
    • Benchmarking documentation now clearly states that it requires opt-in
    • Documented CATCH_CONFIG_CPP17_OPTIONAL and CATCH_CONFIG_CPP17_BYTE macros
    • Properly documented built-in vector matchers
    • Improved *_THROWS_MATCHES documentation a bit
  • CMake config file is now arch-independent even if CMAKE_SIZEOF_VOID_P is in CMake cache (#1660)
  • CatchAddTests now properly escapes [ and ] in test names (#1634, #1698)
  • Reverted CatchAddTests adding tags as CTest labels (#1658)
    • The script broke when test names were too long
    • Overwriting LABELS caused trouble for users who set them manually
    • CMake does not let users append to LABELS if the test name has spaces

2.9.1

Fixes

  • Fix benchmarking compilation failure in files without CATCH_CONFIG_EXTERNAL_INTERFACES (or implementation)

2.9.0

Improvements

  • The experimental benchmarking support has been replaced by integrating Nonius code (#1616)
    • This provides a much more featurefull micro-benchmarking support.
    • Due to the compilation cost, it is disabled by default. See the documentation for details.
    • As far as backwards compatibility is concerned, this feature is still considered experimental in that we might change the interface based on user feedback.
  • WithinULP matcher now shows the acceptable range (#1581)
  • Template test cases now support type lists (#1627)

2.8.0

Improvements

  • Templated test cases no longer check whether the provided types are unique (#1628)
    • This allows you to e.g. test over uint32_t, uint64_t, and size_t without compilation failing
  • The precision of floating point stringification can be modified by user (#1612, #1614)
  • We now provide REGISTER_ENUM convenience macro for generating StringMaker specializations for enums
    • See the "String conversion" documentation for details
  • Added new set of macros for template test cases that enables the use of NTTPs (#1531, #1609)
    • See "Test cases and sections" documentation for details

Fixes

  • UNSCOPED_INFO macro now has a prefixed/disabled/prefixed+disabled versions (#1611)
  • Reporting errors at startup should no longer cause a segfault under certain circumstances (#1626)

Miscellaneous

  • CMake will now prevent you from attempting in-tree build (#1636, #1638)
    • Previously it would break with an obscure error message during the build step

2.7.2

Improvements

  • Added an approximate vector matcher (#1499)

Fixes

  • Filters will no longer be shown if there were none
  • Fixed compilation error when using Homebrew GCC on OS X (#1588, #1589)
  • Fixed the console reporter not showing messages that start with a newline (#1455, #1470)
  • Modified JUnit reporter's output so that rng seed and filters are reported according to the JUnit schema (#1598)
  • Fixed some obscure warnings and static analysis passes

Miscellaneous

  • Various improvements to ParseAndAddCatchTests (#1559, #1601)
    • When a target is parsed, it receives ParseAndAddCatchTests_TESTS property which summarizes found tests
    • Fixed problem with tests not being found if the OptionalCatchTestLauncher variables is used
    • Including the script will no longer forcefully modify CMAKE_MINIMUM_REQUIRED_VERSION
    • CMake object libraries are ignored when parsing to avoid needless warnings
  • CatchAddTests now adds test's tags to their CTest labels (#1600)
  • Added basic CPack support to our build

2.7.1

Improvements

  • Reporters now print out the filters applied to test cases (#1550, #1585)
  • Added GENERATE_COPY and GENERATE_REF macros that can use variables inside the generator expression
    • Because of the significant danger of lifetime issues, the default GENERATE macro still does not allow variables
  • The map generator helper now deduces the mapped return type (#1576)

Fixes

  • Fixed ObjC++ compilation (#1571)
  • Fixed test tag parsing so that [.foo] is now parsed as [.][foo].
  • Suppressed warning caused by the Windows headers defining SE codes in different manners (#1575)

2.7.0

Improvements

  • TEMPLATE_PRODUCT_TEST_CASE now uses the resulting type in the name, instead of the serial number (#1544)
  • Catch2's single header is now strictly ASCII (#1542)
  • Added generator for random integral/floating point types
    • The types are inferred within the random helper
  • Added back RangeGenerator (#1526)
    • RangeGenerator returns elements within a certain range
  • Added ChunkGenerator generic transform (#1538)
    • A ChunkGenerator returns the elements from different generator in chunks of n elements
  • Added UNSCOPED_INFO (#415, #983, #1522)
    • This is a variant of INFO that lives until next assertion/end of the test case.

Fixes

  • All calls to C stdlib functions are now std:: qualified (#1541)
    • Code brought in from Clara was also updated.
  • Running tests will no longer open the specified output file twice (#1545)
    • This would cause trouble when the file was not a file, but rather a named pipe
    • Fixes the CLion/Resharper integration with Catch
  • Fixed -Wunreachable-code occurring with (old) ccache+cmake+clang combination (#1540)
  • Fixed -Wdefaulted-function-deleted warning with Clang 8 (#1537)
  • Catch2's type traits and helpers are now properly namespaced inside Catch:: (#1548)
  • Fixed std{out,err} redirection for failing test (#1514, #1525)
    • Somehow, this bug has been present for well over a year before it was reported

Contrib

  • ParseAndAddCatchTests now properly escapes commas in the test name

2.6.1

Improvements

  • The JUnit reporter now also reports random seed (#1520, #1521)

Fixes

  • The TAP reporter now formats comments with test name properly (#1529)
  • CATCH_REQUIRE_THROWS's internals were unified with REQUIRE_THROWS (#1536)
    • This fixes a potential -Wunused-value warning when used
  • Fixed a potential segfault when using any of the --list-* options (#1533, #1534)

2.6.0

With this release the data generator feature is now fully supported.

Improvements

  • Added TEMPLATE_PRODUCT_TEST_CASE (#1454, #1468)
    • This allows you to easily test various type combinations, see documentation for details
  • The error message for && and || inside assertions has been improved (#1273, #1480)
  • The error message for chained comparisons inside assertions has been improved (#1481)
  • Added StringMaker specialization for std::optional (#1510)
  • The generator interface has been redone once again (#1516)
    • It is no longer considered experimental and is fully supported
    • The new interface supports "Input" generators
    • The generator documentation has been fully updated
    • We also added 2 generator examples

Fixes

  • Fixed -Wredundant-move on newer Clang (#1474)
  • Removed unreachable mentions std::current_exception, std::rethrow_exception in no-exceptions mode (#1462)
    • This should fix compilation with IAR
  • Fixed missing <type_traits> include (#1494)
  • Fixed various static analysis warnings
    • Unrestored stream state in XmlWriter (#1489)
    • Potential division by zero in estimateClockResolution (#1490)
    • Uninitialized member in RunContext (#1491)
    • SourceLineInfo move ops are now marked noexcept
    • CATCH_BREAK_INTO_DEBUGGER is now always a function
  • Fix double run of a test case if user asks for a specific section (#1394, #1492)
  • ANSI colour code output now respects -o flag and writes to the file as well (#1502)
  • Fixed detection of std::variant support for compilers other than Clang (#1511)

Contrib

  • ParseAndAddCatchTests has learned how to use DISABLED CTest property (#1452)
  • ParseAndAddCatchTests now works when there is a whitespace before the test name (#1493)

Miscellaneous

  • We added new issue templates for reporting issues on GitHub
  • contributing.md has been updated to reflect the current test status (#1484)

2.5.0

Improvements

  • Added support for templated tests via TEMPLATE_TEST_CASE (#1437)

Fixes

  • Fixed compilation of PredicateMatcher<const char*> by removing partial specialization of MatcherMethod<T*>
  • Listeners now implicitly support any verbosity (#1426)
  • Fixed compilation with Embarcadero builder by introducing Catch::isnan polyfill (#1438)
  • Fixed CAPTURE asserting for non-trivial captures (#1436, #1448)

Miscellaneous

  • We should now be providing first party Conan support via https://bintray.com/catchorg/Catch2 (#1443)
  • Added new section "deprecations and planned changes" to the documentation
    • It contains summary of what is deprecated and might change with next major version
  • From this release forward, the released headers should be pgp signed (#430)

2.4.2

Improvements

  • XmlReporter now also outputs the RNG seed that was used in a run (#1404)
  • Catch::Session::applyCommandLine now also accepts wchar_t arguments.
    • However, Catch2 still does not support unicode.
  • Added STATIC_REQUIRE macro (#1356, #1362)
  • Catch2's singleton's are now cleaned up even if tests are run (#1411)
    • This is mostly useful as a FP prevention for users who define their own main.
  • Specifying an invalid reporter via -r is now reported sooner (#1351, #1422)

Fixes

  • Stringification no longer assumes that char is signed (#1399, #1407)
    • This caused a Wtautological-compare warning.
  • SFINAE for operator<< no longer sees different overload set than the actual insertion (#1403)

Contrib

  • catch_discover_tests correctly adds tests with comma in name (#1327, #1409)
  • Added a new customization point in how the tests are launched to catch_discover_tests

2.4.1

Improvements

  • Added a StringMaker for std::(w)string_view (#1375, #1376)
  • Added a StringMaker for std::variant (#1380)
    • This one is disabled by default to avoid increased compile-time drag
  • Added detection for cygwin environment without std::to_string (#1396, #1397)

Fixes

  • UnorderedEqualsMatcher will no longer accept erroneously accept vectors that share suffix, but are not permutation of the desired vector
  • Abort after (-x N) can no longer be overshot by nested REQUIRES and subsequently ignored (#1391, #1392)

2.4.0

This release brings two new experimental features, generator support and a -fno-exceptions support. Being experimental means that they will not be subject to the usual stability guarantees provided by semver.

Improvements

  • Various small runtime performance improvements
  • CAPTURE macro is now variadic
  • Added AND_GIVEN macro (#1360)
  • Added experimental support for data generators
  • Added support for compiling and running Catch without exceptions

Fixes

  • Suppressed -Wnon-virtual-dtor warnings in Matchers (#1357)
  • Suppressed -Wunreachable-code warnings in floating point matchers (#1350)

CMake

  • It is now possible to override which Python is used to run Catch's tests (#1365)
  • Catch now provides infrastructure for adding tests that check compile-time configuration
  • Catch no longer tries to install itself when used as a subproject (#1373)
  • Catch2ConfigVersion.cmake is now generated as arch-independent (#1368)
    • This means that installing Catch from 32-bit machine and copying it to 64-bit one works
    • This fixes conan installation of Catch

2.3.0

This release changes the include paths provided by our CMake and pkg-config integration. The proper include path for the single-header when using one of the above is now <catch2/catch.hpp>. This change also necessitated changes to paths inside the repository, so that the single-header version is now at single_include/catch2/catch.hpp, rather than single_include/catch.hpp.

Fixes

  • Fixed Objective-C++ build
  • -Wunused-variable suppression no longer leaks from Catch's header under Clang
  • Implementation of the experimental new output capture can now be disabled (#1335)
    • This allows building Catch2 on platforms that do not provide things like dup or tmpfile.
  • The JUnit and XML reporters will no longer skip over successful tests when running without -s (#1264, #1267, #1310)
    • See improvements for more details

Improvements

  • pkg-config and CMake integration has been rewritten
    • If you use them, the new include path is #include <catch2/catch.hpp>
    • CMake installation now also installs scripts from contrib/
    • For details see the new documentation
  • Reporters now have a new customization point, ReporterPreferences::shouldReportAllAssertions
    • When this is set to false and the tests are run without -s, passing assertions are not sent to the reporter.
    • Defaults to false.
  • Added DYNAMIC_SECTION, a section variant that constructs its name using stream
    • This means that you can do DYNAMIC_SECTION("For X := " << x).

2.2.3

To fix some of the bugs, some behavior had to change in potentially breaking manner. This means that even though this is a patch release, it might not be a drop-in replacement.

Fixes

  • Listeners are now called before reporter
    • This was always documented to be the case, now it actually works that way
  • Catch's commandline will no longer accept multiple reporters
    • This was done because multiple reporters never worked properly and broke things in non-obvious ways
    • This has potential to be a breaking change
  • MinGW is now detected as Windows platform w/o SEH support (#1257)
    • This means that Catch2 no longer tries to use POSIX signal handling when compiled with MinGW
  • Fixed potential UB in parsing tags using non-ASCII characters (#1266)
    • Note that Catch2 still supports only ASCII test names/tags/etc
  • TEST_CASE_METHOD can now be used on classnames containing commas (#1245)
    • You have to enclose the classname in extra set of parentheses
  • Fixed insufficient alt stack size for POSIX signal handling (#1225)
  • Fixed compilation error on Android due to missing std::to_string in C++11 mode (#1280)
  • Fixed the order of user-provided FALLBACK_STRINGIFIER in stringification machinery (#1024)
    • It was intended to be replacement for built-in fallbacks, but it was used after them.
    • This has potential to be a breaking change
  • Fixed compilation error when a type has an operator<< with templated lhs (#1285, #1306)

Improvements

  • Added a new, experimental, output capture (#1243)
    • This capture can also redirect output written via C apis, e.g. printf
    • To opt-in, define CATCH_CONFIG_EXPERIMENTAL_REDIRECT in the implementation file
  • Added a new fallback stringifier for classes derived from std::exception
    • Both StringMaker specialization and operator<< overload are given priority

Miscellaneous

  • contrib/ now contains dbg scripts that skip over Catch's internals (#904, #1283)
    • gdbinit for gdb lldbinit for lldb
  • CatchAddTests.cmake no longer strips whitespace from tests (#1265, #1281)
  • Online documentation now describes --use-colour option (#1263)

2.2.2

Fixes

  • Fixed bug in WithinAbs::match() failing spuriously (#1228)
  • Fixed clang-tidy diagnostic about virtual call in destructor (#1226)
  • Reduced the number of GCC warnings suppression leaking out of the header (#1090, #1091)
    • Only -Wparentheses should be leaking now
  • Added upper bound on the time benchmark timer calibration is allowed to take (#1237)
    • On platforms where std::chrono::high_resolution_clock's resolution is low, the calibration would appear stuck
  • Fixed compilation error when stringifying static arrays of unsigned chars (#1238)

Improvements

  • XML encoder now hex-encodes invalid UTF-8 sequences (#1207)
    • This affects xml and junit reporters
    • Some invalid UTF-8 parts are left as is, e.g. surrogate pairs. This is because certain extensions of UTF-8 allow them, such as WTF-8.
  • CLR objects (T^) can now be stringified (#1216)
    • This affects code compiled as C++/CLI
  • Added PredicateMatcher, a matcher that takes an arbitrary predicate function (#1236)

Others

  • Modified CMake-installed pkg-config to allow #include <catch.hpp>(#1239)
    • The plans to standardize on #include <catch2/catch.hpp> are still in effect

2.2.1

Fixes

  • Fixed compilation error when compiling Catch2 with std=c++17 against libc++ (#1214)
    • Clara (Catch2's CLI parsing library) used std::optional without including it explicitly
  • Fixed Catch2 return code always being 0 (#1215)
    • In the words of STL, "We feel superbad about letting this in"

2.2.0

Fixes

  • Hidden tests are not listed by default when listing tests (#1175)
    • This makes catch_discover_tests CMake script work better
  • Fixed regression that meant <windows.h> could potentially not be included properly (#1197)
  • Fixed installing Catch2ConfigVersion.cmake when Catch2 is a subproject.

Improvements

  • Added an option to warn (+ exit with error) when no tests were ran (#1158)
    • Use as -w NoTests
  • Added provisional support for Emscripten (#1114)
  • Added a way to override the fallback stringifier (#1024)
    • This allows project's own stringification machinery to be easily reused for Catch
  • Catch::Session::run() now accepts char const * const *, allowing it to accept array of string literals (#1031, #1178)
    • The embedded version of Clara was bumped to v1.1.3
  • Various minor performance improvements
  • Added support for DJGPP DOS crosscompiler (#1206)

2.1.2

Fixes

  • Fixed compilation error with -fno-rtti (#1165)
  • Fixed NoAssertion warnings
  • operator<< is used before range-based stringification (#1172)
  • Fixed -Wpedantic warnings (extra semicolons and binary literals) (#1173)

Improvements

  • Added CATCH_VERSION_{MAJOR,MINOR,PATCH} macros (#1131)
  • Added BrightYellow colour for use in reporters (#979)
    • It is also used by ConsoleReporter for reconstructed expressions

Other changes

  • Catch is now exported as a CMake package and linkable target (#1170)

2.1.1

Improvements

  • Static arrays are now properly stringified like ranges across MSVC/GCC/Clang
  • Embedded newer version of Clara -- v1.1.1
    • This should fix some warnings dragged in from Clara
  • MSVC's CLR exceptions are supported

Fixes

  • Fixed compilation when comparison operators do not return bool (#1147)
  • Fixed CLR exceptions blowing up the executable during translation (#1138)

Other changes

  • Many CMake changes
    • NO_SELFTEST option is deprecated, use BUILD_TESTING instead.
    • Catch specific CMake options were prefixed with CATCH_ for namespacing purposes
    • Other changes to simplify Catch2's packaging

2.1.0

Improvements

  • Various performance improvements
    • On top of the performance regression fixes
  • Experimental support for PCH was added (#1061)
  • CATCH_CONFIG_EXTERNAL_INTERFACES now brings in declarations of Console, Compact, XML and JUnit reporters
  • MatcherBase no longer has a pointless second template argument
  • Reduced the number of warning suppressions that leak into user's code
    • Bugs in g++ 4.x and 5.x mean that some of them have to be left in

Fixes

  • Fixed performance regression from Catch classic
    • One of the performance improvement patches for Catch classic was not applied to Catch2
  • Fixed platform detection for iOS (#1084)
  • Fixed compilation when g++ is used together with libc++ (#1110)
  • Fixed TeamCity reporter compilation with the single header version
    • To fix the underlying issue we will be versioning reporters in single_include folder per release
  • The XML reporter will now report WARN messages even when not used with -s
  • Fixed compilation when VectorContains matcher was combined using && (#1092)
  • Fixed test duration overflowing after 10 seconds (#1125, #1129)
  • Fixed std::uncaught_exception deprecation warning (#1124)

New features

  • New Matchers
    • Regex matcher for strings, Matches.
    • Set-equal matcher for vectors, UnorderedEquals
    • Floating point matchers, WithinAbs and WithinULP.
  • Stringification now attempts to decompose all containers (#606)
    • Containers are objects that respond to ADL begin(T) and end(T).

Other changes

  • Reporters will now be versioned in the single_include folder to ensure their compatibility with the last released version

2.0.1

Breaking changes

  • Removed C++98 support
  • Removed legacy reporter support
  • Removed legacy generator support
    • Generator support will come back later, reworked
  • Removed Catch::toString support
    • The new stringification machinery uses Catch::StringMaker specializations first and operator<< overloads second.
  • Removed legacy SCOPED_MSG and SCOPED_INFO macros
  • Removed INTERNAL_CATCH_REGISTER_REPORTER
    • CATCH_REGISTER_REPORTER should be used to register reporters
  • Removed legacy [hide] tag
    • [.], [.foo] and [!hide] are still supported
  • Output into debugger is now colourized
  • *_THROWS_AS(expr, exception_type) now unconditionally appends const& to the exception type.
  • CATCH_CONFIG_FAST_COMPILE now affects the CHECK_ family of assertions as well as REQUIRE_ family of assertions
    • This is most noticeable in CHECK(throws()), which would previously report failure, properly stringify the exception and continue. Now it will report failure and stop executing current section.
  • Removed deprecated matcher utility functions Not, AllOf and AnyOf.
    • They are superseded by operators !, && and ||, which are natural and do not have limited arity
  • Removed support for non-const comparison operators
    • Non-const comparison operators are an abomination that should not exist
    • They were breaking support for comparing function to function pointer
  • std::pair and std::tuple are no longer stringified by default
    • This is done to avoid dragging in <tuple> and <utility> headers in common path
    • Their stringification can be enabled per-file via new configuration macros
  • Approx is subtly different and hopefully behaves more as users would expect
    • Approx::scale defaults to 0.0
    • Approx::epsilon no longer applies to the larger of the two compared values, but only to the Approx's value
    • INFINITY == Approx(INFINITY) returns true

Improvements

  • Reporters and Listeners can be defined in files different from the main file
    • The file has to define CATCH_CONFIG_EXTERNAL_INTERFACES before including catch.hpp.
  • Errors that happen during set up before main are now caught and properly reported once main is entered
    • If you are providing your own main, you can access and use these as well.
  • New assertion macros, *_THROWS_MATCHES(expr, exception_type, matcher) are provided
    • As the arguments suggest, these allow you to assert that an expression throws desired type of exception and pass the exception to a matcher.
  • JUnit reporter no longer has significantly different output for test cases with and without sections
  • Most assertions now support expressions containing commas (ie REQUIRE(foo() == std::vector<int>{1, 2, 3});)
  • Catch now contains experimental micro benchmarking support
    • See projects/SelfTest/Benchmark.tests.cpp for examples
    • The support being experiment means that it can be changed without prior notice
  • Catch uses new CLI parsing library (Clara)
    • Users can now easily add new command line options to the final executable
    • This also leads to some changes in Catch::Session interface
  • All parts of matchers can be removed from a TU by defining CATCH_CONFIG_DISABLE_MATCHERS
    • This can be used to somewhat speed up compilation times
  • An experimental implementation of CATCH_CONFIG_DISABLE has been added
    • Inspired by Doctest's DOCTEST_CONFIG_DISABLE
    • Useful for implementing tests in source files
      • ie for functions in anonymous namespaces
    • Removes all assertions
    • Prevents TEST_CASE registrations
    • Exception translators are not registered
    • Reporters are not registered
    • Listeners are not registered
  • Reporters/Listeners are now notified of fatal errors
    • This means specific signals or structured exceptions
    • The Reporter/Listener interface provides default, empty, implementation to preserve backward compatibility
  • Stringification of std::chrono::duration and std::chrono::time_point is now supported
    • Needs to be enabled by a per-file compile time configuration option
  • Add pkg-config support to CMake install command

Fixes

  • Don't use console colour if running in XCode
  • Explicit constructor in reporter base class
  • Swept out -Wweak-vtables, -Wexit-time-destructors, -Wglobal-constructors warnings
  • Compilation for Universal Windows Platform (UWP) is supported
    • SEH handling and colorized output are disabled when compiling for UWP
  • Implemented a workaround for std::uncaught_exception issues in libcxxrt
    • These issues caused incorrect section traversals
    • The workaround is only partial, user's test can still trigger the issue by using throw; to rethrow an exception
  • Suppressed C4061 warning under MSVC

Internal changes

  • The development version now uses .cpp files instead of header files containing implementation.
    • This makes partial rebuilds much faster during development
  • The expression decomposition layer has been rewritten
  • The evaluation layer has been rewritten
  • New library (TextFlow) is used for formatting text to output

Older versions

1.12.x

1.12.2

Fixes
  • Fixed missing include

1.12.1

Fixes
  • Fixed deprecation warning in ScopedMessage::~ScopedMessage
  • All uses of min or max identifiers are now wrapped in parentheses
    • This avoids problems when Windows headers define min and max macros

1.12.0

Fixes
  • Fixed compilation for strict C++98 mode (ie not gnu++98) and older compilers (#1103)
  • INFO messages are included in the xml reporter output even without -s specified.

1.11.x

1.11.0

Fixes
  • The original expression in REQUIRE_FALSE( expr ) is now reporter properly as !( expr ) (#1051)
    • Previously the parentheses were missing and x != y would be expanded as !x != x
  • Approx::Margin is now inclusive (#952)
    • Previously it was meant and documented as inclusive, but the check itself wasn't
    • This means that REQUIRE( 0.25f == Approx( 0.0f ).margin( 0.25f ) ) passes, instead of fails
  • RandomNumberGenerator::result_type is now unsigned (#1050)
Improvements
  • __JETBRAINS_IDE__ macro handling is now CLion version specific (#1017)
    • When CLion 2017.3 or newer is detected, __COUNTER__ is used instead of
  • TeamCity reporter now explicitly flushes output stream after each report (#1057)
    • On some platforms, output from redirected streams would show up only after the tests finished running
  • ParseAndAddCatchTests now can add test files as dependency to CMake configuration
    • This means you do not have to manually rerun CMake configuration step to detect new tests

1.10.x

1.10.0

Fixes
  • Evaluation layer has been rewritten (backported from Catch 2)
    • The new layer is much simpler and fixes some issues (#981)
  • Implemented workaround for VS 2017 raw string literal stringification bug (#995)
  • Fixed interaction between [!shouldfail] and [!mayfail] tags and sections
    • Previously sections with failing assertions would be marked as failed, not failed-but-ok
Improvements
  • Added libidentify support
  • Added "wait-for-keypress" option

1.9.x

1.9.6

Improvements
  • Catch's runtime overhead has been significantly decreased (#937, #939)
  • Added --list-extra-info cli option (#934).
    • It lists all tests together with extra information, ie filename, line number and description.

1.9.5

Fixes
  • Truthy expressions are now reconstructed properly, not as booleans (#914)
  • Various warnings are no longer erroneously suppressed in test files (files that include catch.hpp, but do not define CATCH_CONFIG_MAIN or CATCH_CONFIG_RUNNER) (#871)
  • Catch no longer fails to link when main is compiled as C++, but linked against Objective-C (#855)
  • Fixed incorrect gcc version detection when deciding to use __COUNTER__ (#928)
    • Previously any GCC with minor version less than 3 would be incorrectly classified as not supporting __COUNTER__.
  • Suppressed C4996 warning caused by upcoming updated to MSVC 2017, marking std::uncaught_exception as deprecated. (#927)
Improvements
  • CMake integration script now incorporates debug messages and registers tests in an improved way (#911)
  • Various documentation improvements

1.9.4

Fixes
  • CATCH_FAIL macro no longer causes compilation error without variadic macro support
  • INFO messages are no longer cleared after being reported once
Improvements and minor changes
  • Catch now uses wmain when compiled under Windows and UNICODE is defined.
    • Note that Catch still officially supports only ASCII

1.9.3

Fixes
  • Completed the fix for (lack of) uint64_t in earlier Visual Studios

1.9.2

Improvements and minor changes
  • All of Approx's member functions now accept strong typedefs in C++11 mode (#888)
    • Previously Approx::scale, Approx::epsilon, Approx::margin and Approx::operator() didn't.
Fixes
  • POSIX signals are now disabled by default under QNX (#889)
    • QNX does not support current enough (2001) POSIX specification
  • JUnit no longer counts exceptions as failures if given test case is marked as ok to fail.
  • Catch::Option should now have its storage properly aligned.
  • Catch no longer attempts to define uint64_t on windows (#862)
    • This was causing trouble when compiled under Cygwin
Other
  • Catch is now compiled under MSVC 2017 using std:c++latest (C++17 mode) in CI
  • We now provide cmake script that autoregisters Catch tests into ctest.
    • See contrib folder.

1.9.1

Fixes
  • Unexpected exceptions are no longer ignored by default (#885, #887)

1.9.0

Improvements and minor changes
  • Catch no longer attempts to ensure the exception type passed by user in REQUIRE_THROWS_AS is a constant reference.
    • It was causing trouble when REQUIRE_THROWS_AS was used inside templated functions
    • This actually reverts changes made in v1.7.2
  • Catch's Version struct should no longer be double freed when multiple instances of Catch tests are loaded into single program (#858)
    • It is now a static variable in an inline function instead of being an externed struct.
  • Attempt to register invalid tag or tag alias now throws instead of calling exit().
    • Because this happen before entering main, it still aborts execution
    • Further improvements to this are coming
  • CATCH_CONFIG_FAST_COMPILE now speeds-up compilation of REQUIRE* assertions by further ~15%.
    • The trade-off is disabling translation of unexpected exceptions into text.
  • When Catch is compiled using C++11, Approx is now constructible with anything that can be explicitly converted to double.
  • Captured messages are now printed on unexpected exceptions
Fixes:
  • Clang's -Wexit-time-destructors should be suppressed for Catch's internals
  • GCC's -Wparentheses is now suppressed for all TU's that include catch.hpp.
    • This is functionally a revert of changes made in 1.8.0, where we tried using _Pragma based suppression. This should have kept the suppression local to Catch's assertions, but bugs in GCC's handling of _Pragmas in C++ mode meant that it did not always work.
  • You can now tell Catch to use C++11-based check when checking whether a type can be streamed to output.
Other notes:
  • We have added VS 2017 to our CI
  • Work on Catch 2 should start soon

1.8.x

1.8.2

Improvements and minor changes
  • TAP reporter now behaves as if -s was always set
    • This should be more consistent with the protocol desired behaviour.
  • Compact reporter now obeys -d yes argument (#780)
    • The format is "XXX.123 s: " (3 decimal places are always present).
    • Before it did not report the durations at all.
  • XML reporter now behaves the same way as Console reporter in regards to INFO
    • This means it reports INFO messages on success, if output on success (-s) is enabled.
    • Previously it only reported INFO messages on failure.
  • CAPTURE(expr) now stringifies expr in the same way assertion macros do (#639)
  • Listeners are now finally documented.
    • Listeners provide a way to hook into events generated by running your tests, including start and end of run, every test case, every section and every assertion.
Fixes:
  • Catch no longer attempts to reconstruct expression that led to a fatal error (#810)
    • This fixes possible signal/SEH loop when processing expressions, where the signal was triggered by expression decomposition.
  • Fixed (C4265) missing virtual destructor warning in Matchers (#844)
  • std::strings are now taken by const& everywhere (#842).
    • Previously some places were taking them by-value.
  • Catch should no longer change errno (#835).
    • This was caused by libstdc++ bug that we now work around.
  • Catch now provides FAIL_CHECK( ... ) macro (#765).
    • Same as FAIL( ... ), but does not abort the test.
  • Functions like fabs, tolower, memset, isalnum are now used with std:: qualification (#543).
  • Clara no longer assumes first argument (binary name) is always present (#729)
    • If it is missing, empty string is used as default.
  • Clara no longer reads 1 character past argument string (#830)
  • Regression in Objective-C bindings (Matchers) fixed (#854)
Other notes:
  • We have added VS 2013 and 2015 to our CI
  • Catch Classic (1.x.x) now contains its own, forked, version of Clara (the argument parser).

1.8.1

Fixes

Cygwin issue with gettimeofday - #define was not early enough

1.8.0

New features/ minor changes
  • Matchers have new, simpler (and documented) interface.
  • Changed console reporter test duration reporting format (#322)
    • Old format: Some simple comparisons between doubles completed in 0.000123s
    • New format: xxx.123s: Some simple comparisons between doubles (There will always be exactly 3 decimal places)
  • Added opt-in leak detection under MSVC + Windows (#439)
    • Enable it by compiling Catch's main with CATCH_CONFIG_WINDOWS_CRTDBG
  • Introduced new compile-time flag, CATCH_CONFIG_FAST_COMPILE, trading features for compilation speed.
    • Moves debug breaks out of tests and into implementation, speeding up test compilation time (~10% on linux).
    • More changes are coming
  • Added TAP (Test Anything Protocol) and Automake reporters.
  • XML reporter now reports filename as part of the Section and TestCase tags.
  • Approx now supports an optional margin of absolute error
Fixes
  • Silenced C4312 ("conversion from int to 'ClassName *") warnings in the evaluate layer.
  • Fixed C4512 ("assignment operator could not be generated") warnings under VS2013.
  • Cygwin compatibility fixes
    • Signal handling is no longer compiled by default.
    • Usage of gettimeofday inside Catch should no longer cause compilation errors.
  • Improved -Wparentheses suppression for gcc (#674)
    • When compiled with gcc 4.8 or newer, the suppression is localized to assertions only
    • Otherwise it is suppressed for the whole TU
  • Fixed test spec parser issue (with escapes in multiple names)
Other
  • Various documentation fixes and improvements

1.7.x

1.7.2

Fixes and minor improvements

Xml:

(technically the first two are breaking changes but are also fixes and arguably break few if any people)

  • C-escape control characters instead of XML encoding them (which requires XML 1.1)
  • Revert XML output to XML 1.0
  • Can provide stylesheet references by extending the XML reporter
  • Added description and tags attributes to XML Reporter
  • Tags are closed and the stream flushed more eagerly to avoid stdout interpolation

Other:

  • REQUIRE_THROWS_AS now catches exception by const& and reports expected type
  • In SECTIONs the file/ line is now of the SECTION. not the TEST_CASE
  • Added std:: qualification to some functions from C stdlib
  • Removed use of RTTI (dynamic_cast) that had crept back in
  • Silenced a few more warnings in different circumstances
  • Travis improvements

1.7.1

Fixes:
  • Fixed inconsistency in defining NOMINMAX and WIN32_LEAN_AND_MEAN inside catch.hpp.
  • Fixed SEH-related compilation error under older MinGW compilers, by making Windows SEH handling opt-in for compilers other than MSVC.
  • Fixed compilation error under MinGW caused by improper compiler detection.
  • Fixed XML reporter sometimes leaving an empty output file when a test ends with signal/structured exception.
  • Fixed XML reporter not reporting captured stdout/stderr.
  • Fixed possible infinite recursion in Windows SEH.
  • Fixed possible compilation error caused by Catch's operator overloads being ambiguous in regards to user-defined templated operators.

1.7.0

Features/ Changes:
  • Catch now runs significantly faster for passing tests
    • Microbenchmark focused on Catch's overhead went from ~3.4s to ~0.7s.
    • Real world test using JSON for Modern C++'s test suite went from ~6m 25s to ~4m 14s.
  • Catch can now run specific sections within test cases.
    • For now the support is only basic (no wildcards or tags), for details see the documentation.
  • Catch now supports SEH on Windows as well as signals on Linux.
    • After receiving a signal, Catch reports failing assertion and then passes the signal onto the previous handler.
  • Approx can be used to compare values against strong typedefs (available in C++11 mode only).
    • Strong typedefs mean types that are explicitly convertible to double.
  • CHECK macro no longer stops executing section if an exception happens.
  • Certain characters (space, tab, etc) are now pretty printed.
    • This means that a char c = ' '; REQUIRE(c == '\t'); would be printed as ' ' == '\t', instead of == 9.
Fixes:
  • Text formatting no longer attempts to access out-of-bounds characters under certain conditions.
  • THROW family of assertions no longer trigger -Wunused-value on expressions containing explicit cast.
  • Breaking into debugger under OS X works again and no longer required DEBUG to be defined.
  • Compilation no longer breaks under certain compiler if a lambda is used inside assertion macro.
Other:
  • Catch's CMakeLists now defines install command.
  • Catch's CMakeLists now generates projects with warnings enabled.

1.6.x

1.6.1

Features/ Changes:
  • Catch now supports breaking into debugger on Linux
Fixes:
  • Generators no longer leak memory (generators are still unsupported in general)
  • JUnit reporter now reports UTC timestamps, instead of "tbd"
  • CHECK_THAT macro is now properly defined as CATCH_CHECK_THAT when using CATCH_ prefixed macros
Other:
  • Types with overloaded && operator are no longer evaluated twice when used in an assertion macro.
  • The use of __COUNTER__ is suppressed when Catch is parsed by CLion
    • This change is not active when compiling a binary
  • Approval tests can now be run on Windows
  • CMake will now warn if a file is present in the include folder but not is not enumerated as part of the project
  • Catch now defines NOMINMAX and WIN32_LEAN_AND_MEAN before including windows.h

1.6.0

Cmake/ projects:
  • Moved CMakeLists.txt to root, made it friendlier for CLion and generating XCode and VS projects, and removed the manually maintained XCode and VS projects.
Features/ Changes:
  • Approx now supports >= and <=
  • Can now use \ to escape chars in test names on command line
  • Standardize C++11 feature toggles
Fixes:
  • Blue shell colour
  • Missing argument to CATCH_CHECK_THROWS
  • Don't encode extended ASCII in XML
  • use std::shuffle on more compilers (fixes deprecation warning/error)
  • Use __COUNTER__ more consistently (where available)
Other:
  • Tweaks and changes to scripts - particularly for Approval test - to make them more portable

Even Older versions

Release notes were not maintained prior to v1.6.0, but you should be able to work them out from the Git history


Home