Closed tompittson closed 5 years ago
Thanks for reporting, I'll investigate!
@tompittson - Thanks for raising this, good find. I have fixed this bug which was manifesting when **
was being matched against the last character of a test string causing in some cases a false positive match. I have added test coverage for this, and it's fixed in 2.1.2
or latest alpha.
Correction: version 2.1.3
ended up being the released hotfix version as I also had to fix an issue with the appveyor build.
@dazinator thanks for looking at this, I have tried the newest versions (2.1.3 and 3.0.1) and the original issue is fixed but it seems to have introduced a couple of additional issues.
"a/b" is now not matched by pattern "a/**/b"
"abc/def" matched against the pattern "abc/**" throws an 'Index was outside the bounds of the array' exception.
Hmm.. apologies for this.. It seems I was too confident in my existing test coverage when making those tweaks to fix that last bug.. I'll add these test cases, and release an update!
Thanks for your patience with this one, I have now fixed this regression in 2.1.4 - and added your repro examples to the regression suite.
Hello,
Trying something out with your library I found the following pattern doesn't match as expected.
I have tried this with 2.1.1 stable and 3.0.0 alpha045 versions. I noticed there was a similar bug (#33) which mentions a GlobParseOptions setting around invalid path characters but I can't see that setting in the GlobOptions at the latest versions.