| commit | a061f5cfd6cfc06840ebe9c5231039d19823a736 | |
| author | mhagger <mhagger@be7e6eca-30d4-0310-a8e5-ac0d63af7087> | |
| Sat, 14 Jan 2006 15:00:55 +0000 (14 15:00 +0000) | ||
| committer | mhagger <mhagger@be7e6eca-30d4-0310-a8e5-ac0d63af7087> | |
| Sat, 14 Jan 2006 15:00:55 +0000 (14 15:00 +0000) | ||
| tree | 259c0ea93ea319b9ef5380ba1f4fecd659acbb68 | treesnapshot (tar.gz zip) |
| parent | b23bad778cd0a69ed9779a44dc464c01695a0514 | commitdiff |
Don't sort arguments before passing them to cvs2svn.
Previously, arguments were sorted before passing them to cvs2svn.
This was apparently to avoid superfluous conversions in the case of
calling ensure_conversion() twice with the same arguments but in
different order.
However, order-dependent options might be interesting (e.g., see
discussion of proposed --include option), in which case sorting the
arguments would be bad. And who's going to call ensure_conversion()
twice with the same arguments but in different order? (Even if this
happens, it only results in a redundant conversion.)
* run-tests.py (ensure_conversion): Pass arguments to cvs2svn in the
order that they appear in ARGS.
git-svn-id: http://cvs2svn.tigris.org/svn/cvs2svn/trunk@1803 be7e6eca-30d4-0310-a8e5-ac0d63af7087
Previously, arguments were sorted before passing them to cvs2svn.
This was apparently to avoid superfluous conversions in the case of
calling ensure_conversion() twice with the same arguments but in
different order.
However, order-dependent options might be interesting (e.g., see
discussion of proposed --include option), in which case sorting the
arguments would be bad. And who's going to call ensure_conversion()
twice with the same arguments but in different order? (Even if this
happens, it only results in a redundant conversion.)
* run-tests.py (ensure_conversion): Pass arguments to cvs2svn in the
order that they appear in ARGS.
git-svn-id: http://cvs2svn.tigris.org/svn/cvs2svn/trunk@1803 be7e6eca-30d4-0310-a8e5-ac0d63af7087
| run-tests.py | diffblobblamehistory |