summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAge
* - templated test cases example is doneonqtam2017-05-16
| | | | - simplified a bit the cmake
* - added an undocumented flag to omit the "skipped" part in the output ↵onqtam2017-05-16
| | | | | | | | summary - for convenience for the all_features example - grouped examples by project - changed names of output from tests - added source files for some features without examples
* - removed another unnecessary example that slowed down builds - now only the ↵onqtam2017-05-16
| | | | | | python script is left - fixing clang builds
* output is now like the old oneonqtam2017-05-16
|
* added playground project and made (again) the common.cmake behave when ↵onqtam2017-05-16
| | | | included multiple times
* - added test outputsonqtam2017-05-16
|
* everything compiles!onqtam2017-05-16
|
* cleaning up examplesonqtam2017-05-16
|
* added example how to filter out doctest options from the command line in the ↵onqtam2017-05-16
| | | | docs and removed it from the examples folder
* - added prefix for the TEST_MODE optiononqtam2017-05-16
| | | | - removed the cpp11 option
* fixing appveyor builds - the name of the solution had changed in my previous ↵onqtam2017-05-16
| | | | commits
* fixed link to small logoonqtam2017-05-16
|
* adding first draft of logoonqtam2017-05-16
|
* setting the cmake version of the project properlyonqtam2017-05-16
|
* examples can no longer be built using cmake separately - they require cmake ↵onqtam2017-05-16
| | | | invocation from the root of the repository (and now use the modern way of handling include dependencies)
* reworked cmake a bitonqtam2017-05-16
| | | | | - examples/tests now guarded with a single option - no longer overriding add_executable/add_library
* moved all cmake-related stuff to scripts/cmakeonqtam2017-05-16
|
* Read version from scripts/version.txt in CMakeLists.txtGiuseppe Roberti2017-05-16
|
* add DOCTEST_SKIP_COVERAGE and DOCTEST_SKIP_EXAMPLESGiuseppe Roberti2017-05-16
|
* add cmake package configuration files creationGiuseppe Roberti2017-05-16
|
* moved stuff around (compacted in one folder)onqtam2017-05-16
|
* - removed warning that is included in Wall/Wextra (although not documented ↵onqtam2017-05-16
| | | | | | in the online docs of GCC) - removed old unnecessary stuff
* improved compile times (not measured though...) - in commit ↵onqtam2017-05-16
| | | | d321e9baf9b4904da74ab3d7404b2d41aa50fb9c 1 year ago I introduced the always_false() function to silence the C4127 MSVC warning (conditional expression is constant) for constructs such as "do {} while(false)" - but that added an unnecessary function call. Now I fix that issue by using the "owl" technique - see this SO question: http://stackoverflow.com/a/1947233/3162383
* fixing -Wunused-parameter warning introduced in last commitonqtam2017-05-16
|
* simplified exception translation a bit - relates #12onqtam2017-05-16
|
* - added CII badgeonqtam2017-05-16
| | | | | - tiny docs update - removed tests - trying to figure out WHY ON EARTH ONLY GCC 5 COMPLAINS ABOUT A STRICT OVERFLOW WARNING where there shouldn't be any - see this CI build: https://travis-ci.org/onqtam/doctest/builds/221092038
* windows builds should be fixed! problem was that colors are disabled for the ↵onqtam2017-05-16
| | | | code coverage project and windows.h wasn't getting included - but it was needed for the recently added SEH support
* fixing clang buildsonqtam2017-05-16
|
* fixing builds with -fno-exceptions ...onqtam2017-05-16
|
* this should hopefully fix linux builds...onqtam2017-05-16
|
* added "struct" infront of "sigaction" - seems necessary - see this SO ↵onqtam2017-05-16
| | | | question: http://stackoverflow.com/questions/14213270/sigaction-both-struct-and-function
* fixing builds for old MSVC versions - where the is_constructible<> trait ↵onqtam2017-05-16
| | | | doesn't work - adding a default double overload - relates previous commit b68f6d7f6dfd7d37f5b78596007adca96da45588
* Approx epsilon() and scale() can not take floats and strong typedefs of ↵onqtam2017-05-16
| | | | double as well - relates #62
* added crash handling: signals on UNIX platforms or structured exceptions on ↵onqtam2017-05-16
| | | | | | | | Windows - closes #63 - taken from here: https://github.com/philsquared/Catch/blob/master/include/internal/catch_fatal_condition.hpp - see history here: https://github.com/philsquared/Catch/commits/master/include/internal/catch_fatal_condition.hpp - current latest commit in catch: cd6de9cd34351b12b09480d2f6fc8cdd4248862b
* preserving context from ```INFO()``` contexts when the test case ends from ↵onqtam2017-05-16
| | | | an exception - relates #48 #23 (also see this Catch PR: https://github.com/philsquared/Catch/pull/876 )
* clarified stuff in the docsonqtam2017-05-16
|
* handling the newly introduced asser_with_message macros properly when ↵onqtam2017-05-16
| | | | exceptions are disabled - relates #48
* silencing this warning again to fix my builds... I don't think it could end ↵onqtam2017-05-16
| | | | up silencing user code since by the time the evaluation of operator== (or whatever) comes to the doctest templates all operators with higher precedence should have already finished (regarding the expression decomposition)
* added comment about the use of operator << instead of <= for the expression ↵onqtam2017-05-16
| | | | decomposition
* all clang/gcc builds should pass now!onqtam2017-05-16
|
* removed unnecessary silencing of Wdouble-promotion for gcc/clangonqtam2017-05-16
|
* simplified is_constructible<> to accept only one argument and removed ↵onqtam2017-05-16
| | | | pointer support - builds with old MSVC versions should pass now - relates #62
* fixed the constructor - should now work for types from which double can be ↵onqtam2017-05-16
| | | | | | constructed - see here: http://stackoverflow.com/questions/43306848 relates #62
* Added support to Approx for strong typedefs of double - fixes #62onqtam2017-05-16
| | | | - added is_constructible<> trait which for gcc/clang works as expected but for MSVC accepts only 1 argument for construction and versions of MSVC older than 2012 will always return false - so strong typedefs of double won't work with Approx there.
* Revert "- silenced warnings for unused variables in assert macros when ↵onqtam2017-05-16
| | | | | | DOCTEST_CONFIG_DISABLE is used" This reverts commit 6b61e8aa3818c5ea100cedc1bb48a60ea10df6e8.
* - silenced warnings for unused variables in assert macros when ↵onqtam2017-05-16
| | | | | | | DOCTEST_CONFIG_DISABLE is used - couldn't do it only for the THROWS/THROWS_AS/NOTHROW macros - relates #61
* added 2 more notes about the templated test casesonqtam2017-05-16
|
* the TEST_CASE_TEMPLATE_INSTANTIATE and TYPE_TO_STRING macros now properly ↵onqtam2017-05-16
| | | | require a semicolon after their use
* builds should pass now... the return type of sizeof() is tricky - it gets ↵onqtam2017-05-16
| | | | stringified differently in some environments
* - moved contributing guidelines to root folderonqtam2017-05-16
| | | | - added template for issues and PRs
generated by cgit v1.2.3 (git 2.25.1) at 2024-07-02 06:46:11 +0000