todo.txt

fbcunit TODO/BUG list:

[ ] write a (short) manual on the library's API

[ ] add unit tests (for fbcunit itself) for all CU_* macros

[ ] for compatibility with existing fbc compiler suite,
	can't check the return value of init & cleanup code
	because fbc test suite doesn't set it.

	1) remove the check, and let user decide how to deal with
		failed init & cleanup code - i.e. CU_ASSERT_FATAL( false )

	OR 2) default to success and let user set a FAIL flag

	OR 3) update fbc test suites

	For either #2 or #3 still todo:
	SUITE_INIT and SUITE_CLEANUP need to set the function 
	return value.  Consider	adding a macro to set the value.

	Current	work around for qb is:

	SUITE( somename )
		SUITE_INIT
			somename.init = true
		END_SUITE_INIT
		SUITE_CLEANUP
			somename.cleanup = true
		END_SUITE_CLEANUP
	END_SUITE


	Current	work around for fb, fblite is:

	SUITE( somename )
		SUITE_INIT
			function = true
		END_SUITE_INIT
		SUITE_CLEANUP
			function = true
		END_SUITE_CLEANUP
	END_SUITE

[ ] allowing adding flags to individual tests, especially:
	TEST_SET_CONSOLE(1) - indicate that test generates console output
	TEST_SET_INTERACTIVE(1) - indicates test is interactive

[ ] add_suite() might be called with same name as a suite already
	added but with different init/cleanup procs.  fbcunit framework
	should maintain a list of init/cleanup procs instead of just one

[ ] SUITE() can be used in more than one module with the same name,
	that's OK, however, the constructor might be duplicated at the
	END_SUITE - It uses the __LINE__ built-in to generate unique 
	identifiers.  This falls apart however,	when END_SUITE is used 
	for the same suite, in multiple files and happens to land on the 
	exact same line number.