For those who have not followed, nd/wildmatch brings another fnmatch-like implementation which can nearly replace fnmatch. System fnmatch() seems to behave differently in some cases. It's better to stay away and use one implementation for all.
I just wanted to see how much work there may be if we go this way. It turns out not much. I haven't checked my dowild() changes carefully. I may have left a bug in '[]' code. There are some minor issues I like dependency on FNM_* macros or wildmatch.h should be incorporated back to git-compat-util.h. But the test suite passes for me. So it's promising. Nguyễn Thái Ngọc Duy (3): wildmatch: make dowild() take arbitrary flags wildmatch: support "no FNM_PATHNAME" mode Convert all fnmatch() calls to wildmatch() builtin/apply.c | 3 ++- builtin/branch.c | 3 ++- builtin/describe.c | 3 ++- builtin/for-each-ref.c | 3 ++- builtin/ls-remote.c | 3 ++- builtin/name-rev.c | 3 ++- builtin/reflog.c | 3 ++- builtin/replace.c | 3 ++- builtin/show-branch.c | 3 ++- builtin/tag.c | 3 ++- diffcore-order.c | 3 ++- dir.c | 6 +++--- refs.c | 3 ++- t/t3070-wildmatch.sh | 27 +++++++++++++++++++++++++++ test-wildmatch.c | 4 +++- tree-walk.c | 5 +++-- wildmatch.c | 25 ++++++++++++++----------- 17 files changed, 74 insertions(+), 29 deletions(-) -- 1.8.0.rc2.23.g1fb49df -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
