windows file tests, fail and SiB

hi! While fixing a possible in stat, I met a couple of tests which should really be revamped into smaller and nicer versions (Small is Beatiful). One example is ext\standard\tests\file\touch_variation6-win32.phpt. As it is a very good test (covers many cases), it is sadly useless when it comes to figure out what fails where. I wonder if you have the time to do it soonish, that would be amazingly helpful. I could do it as well but I may need more time than the original author :) If not, then I will do it but sadly not for 5.3beta (neither the bug fixes as I can't verify that it does not break more than it should :). Failing tests: ===================================================================== FAILED TEST SUMMARY --------------------------------------------------------------------- Bug #43353 wrong detection of 'data' wrapper [ext\standard\tests\file\bug43353.phpt] Bug #44607 (stream_get_line unable to correctly identify the "ending" in the stream content) [ext\standard\tests\file\bug44607.phpt] Test fopen() function : variation: interesting paths, no use include path [ext\standard\tests\file\fopen_variation10-win32.phpt] Test fopen() function : variation: interesting paths, use include path = true [ext\standard\tests\file\fopen_variation11-win32.phpt] Test parse_ini_file() function : variation: include path searching [ext\standard\tests\file\parse_ini_file_variation3.phpt] Test popen() and pclose function: error conditions [ext\standard\tests\file\popen_pclose_error-win32-debug.phpt] Test stat() functions: usage variations - effects of truncate() [ext\standard\tests\file\stat_variation8-win32.phpt] Test touch() function : variation: various valid and invalid paths [ext\standard\tests\file\touch_variation5-win32.phpt] Test touch() function : variation: various valid and invalid paths [ext\standard\tests\file\touch_variation6-win32.phpt] User-space streams: stream_set_option() [ext\standard\tests\file\userstreams_003.phpt] ===================================================================== Not sure why so much are failing as the last time I run them locally, I was close to 100% passed. That's what I try to clean/fix now. The expected to fail is also amazingly high... Cheers, -- Pierre http://blog.thepimp.net | http://www.libgd.org