Failed and errored are exclusive in the JUnit sense.
Counting the number of total tests slightly differently.
Changing slightly the signature of unit_test_log_formatter::log_build_info to avoid a sticky
state when the function was called only once (happens for instance in JUNIT).
- travis with valgrind, cppcheck, ubsan, codecov, covscan (future)
- appveyor with MSVC 2010 through 2017, cygwin 32/64, mingw 32/64
- README, LICENSE, etc.
- Fixed example Jamfile to build examples based on C++ features
This fixes#146
* develop:
Change log
Documentation updates
Adding the support for skipped test units/suites in the reporter and formatters
<boost/timer.hpp> removal and time-out for suites and major refactoring
Nicer change log
Typo
Change log PR-190
Change log
Correct library name in test runner help screen
Repointing former SGI links
Correcting BOOST_TEST messages in the scope of the Contexts documentation
Fixing floating point dead link
Removing the need for samples from datasets
Change log with TOC
* doc/various-updates-cleanups:
Nicer change log
Typo
Change log PR-190
Repointing former SGI links
Correcting BOOST_TEST messages in the scope of the Contexts documentation
Fixing floating point dead link
Removing the need for samples from datasets
Change log with TOC
* topic/GH-202-boost-timer-deprecation:
Change log
Documentation updates
Adding the support for skipped test units/suites in the reporter and formatters
<boost/timer.hpp> removal and time-out for suites and major refactoring
As boost/timer.hpp is being deprecated, and the needed functionality
is very little, reimplemeted naively a timing class for the purpose
of measuring the test-unit time.
- specific timer class that is cross-platform and agnostic to changes made
to the OS while measuring time (mac, posix and windows). Wall-clock and system/user
time available
- changed framework to work fully with microseconds. Only ::alarm needs seconds
and ceiling is used for signaling time-out
- Windows now signals also time-outs the same way as for other platforms (exception
raised in the test and reported). However, Windows platform do not raise a signal
that terminates the current test
- timeout is not used for failing a test-suite as well. Remaining time calculation is
now made in microseconds. Adding an observer interface for signaling time out and
adding the time outs in the test results
- cleaned up interfaces, clarify the code about the dimension of the different scalar values.
* develop:
Change log
Fixing propagation of decorators to underlying test-case generator
Change log
Added unused variable markup to silence gcc warnings.
Use legacy _vsnprintf on MinGW.
Change log
Making BOOST_TEST_CONTEXT variadic and adding sticky context
Change log
Added documentation on throwing dataset
Change log
Extended fixture messages for template test cases
Dataset test case: allowing fixtures w. setup/teardown
Change log
Relaxed tolerance based comparisons
Change log
Default visibility for enums and rtti objects facility
The BOOST_DATA_TEST_CASE creates an implicit test suite. that consumes
all the decorators that have been declared prior to the BOOST_DATA_TEST_CASE
macro. In order to properly propagate the decorators to the underlying
test-case generator (the data test case), a new "stacked" decorator collector
has been implemented. The new decorator decorator::stack_decorator
allows the user to push the currently stacked decorator to a higher level of the
hierarchy.
The singleton reference variables and a few other globals may not be actually
used in the user's code, which makes gcc emit warnings. This commit marks these
variables as potentially unused, which silences the warnings.
Apparently, some versions of legacy MinGW headers don't provide
the standard vsnprintf function, as Appveyor CI fails for this target.
Thus we use the non-standard MSVC-specific _vsnprintf instead.
To ease the creation of contexts, variadic let us declare several
values in one call (no extra brace) while ticky context let us
add new information while they arrive.
The example shows a random trial making a test fail. We force the failure
for the unit tests, which does not remove the documentation value of the
example.
* develop: (39 commits)
Change log
Rewrite example22 and example23
Change log
Fixing the command line parsing
Change log
Preventing the compiler to optimize out the duplicate strings
Change log
Deprecating boost.test minimal.hpp header
Change log
Using boost/core facility to silence unused variables
Check for non-used variables when NDEBUG is defined
Change log
Removed second declaration
Change log
Fix missing vsnprintf errors on MinGW.
Adding requires to bjam/examples
Change log
Suppress color setup warnings
Change log
Deregistration of global objects at framework shutdown time
...
It is now possible to compare a type that is tolerance based together
with a type that is not necessarily tolerance based but still
arithmetic.
Doc update and tests.
