Fri, 16 Dec 2011 21:09:40 +0900 windows: use upper() instead of lower() or os.path.normcase() stable
FUJIWARA Katsunori <foozy@lares.dti.ne.jp> [Fri, 16 Dec 2011 21:09:40 +0900] rev 15671
windows: use upper() instead of lower() or os.path.normcase() this patch uses upper() instead of lower() or os.path.normcase() for case folding on Windows(NTFS), because lower-ing causes problems for some languages on it. see below for detail about problem of lower-ing: https://blogs.msdn.com/b/michkap/archive/2005/01/16/353873.aspx
Fri, 16 Dec 2011 21:09:40 +0900 icasefs: avoid normcase()-ing in util.fspath() for efficiency stable
FUJIWARA Katsunori <foozy@lares.dti.ne.jp> [Fri, 16 Dec 2011 21:09:40 +0900] rev 15670
icasefs: avoid normcase()-ing in util.fspath() for efficiency 'dirstate._normalize()', the only caller of 'util.fspath()', has already normcase()-ed path before invocation of it. normcase()-ed root can be cached on dirstate side, too. so, this patch changes 'util.fspath()' API specification to avoid normcase()-ing in it.
Fri, 16 Dec 2011 21:09:40 +0900 icasefs: use util.normcase() instead of lower() or os.path.normcase in fspath stable
FUJIWARA Katsunori <foozy@lares.dti.ne.jp> [Fri, 16 Dec 2011 21:09:40 +0900] rev 15669
icasefs: use util.normcase() instead of lower() or os.path.normcase in fspath this also avoids lower()-ing on each path components by reuse the path normcase()-ed at beginning of function.
(0) -10000 -3000 -1000 -300 -100 -30 -10 -3 +3 +10 +30 +100 +300 +1000 +3000 +10000 +30000 tip