* topic/loggers-additional-cleanups:
Loggers: now checking for --logger params correctness
Logger: cleaned up and improved the API
being able to add a custom formatter instead of having it in an 'invalid' state
Additional cleanups
Additional cleanup of unused fields
Moving the logger setup into a dedicated function
- now experiencing the same behaviour as pre 1.62
- set_stream and set_threshold_level for the loggers are now mutating all loggers, not only the active ones
- renamed function
- when adding a custom logger, it is added to the set of loggers with CUSTOM
- updated the documentation
* topic/12093-gcc-compilation-issues:
Change log update
gcc 4.6: fixing failing tests
gcc 4.6: fixing issues related to the new cla and C++11 not fully supported
gcc 4.6: fixing issues related to operator<< not properly seen by the compiler
- major reorganization of the output section
- making more clear what a log and what a report is
- checkpoints doc update
- links to API, improved doxygen documentation and reference
- renamed files and sections
- JUNIT additional doc
- removed useless parts
* topic/8707-8834-junit-several-loggers:
Change log update
Documentation
JUnit default output stream
Logger tests: overriding the output_stream_logger matching algorithm and changing the format of the messages sent by the framework
Combined logger: CLA handling and stream/log level per formatter
Junit logger: integration into the set of loggers
JUnit logger: implementation
Multiple logger: entry in progress now being per logger as well
Multiple logger: now enabling a logger by its format
Multiple logger: breaking changes in the API in order to be able to add a logger per format on the fly
Loggers: now handling abort signal as well
Multiple loggers: having several loggers at the same time
Algorithm: replace all strings with/out wildcards
output_test_stream: being able to override the match_pattern method by making it virtual (+ other functions)
output_test_stream: improved messages on error
Framework: fixing the time associated to each test unit
issues with VS placeholders
* topic/unique_data_test_case_name:
Change log update
dataset test case have their own unique name
Generate unique test cases name for data test cases series
Fixes 11845
Each data test case is not creating a test suite with the name provided by the BOOST_DATA_TEST_CASE directive. Inside that test suite, all samples are added with names _0, _1 ...
- Fix counter after copy/move
Rationale: being agnostic to the variable of the unit tests (time, file location)
- the output_test_stream now is being able to replace the content of the stream before it reached the matching algorithm
- several stream processing: file names (cross platform), execution time ... for removing the platform dependant log format
- additional cleanup and fixes for the JUnit implementation (all assertions are now properly logged and flushed at write time in the appropriate section of the file)
- now the framework emits warnings/errors that are signaled by "boost.test framework" instead of a file that is unknown from the user (+1 squashed commit)
- new CLA argument for being able to specify a logger with repetition, as well as its log level and stream
- the specification of the logger are: logger_name[:log_level[:log_sink]]
- empty token are kept: 'logger_name::' is a valid spec
- now being able to specify a stream per format: modification of the stream helper
- the output stream of a particular format now accepts the name of the stream instead of the name of the associated param on the command line
- Make JUnit logger discard the log levels to log all events
- Currently JUnit discards the file directory where the assertion is being logged
- Uses tree visitation at the end of the test suite in order to generate the log file
