- May 28, 2015
-
-
Durham Goode authored
Upstream mercurial has dropped the ignore module and replaced it with 'include:' patterns. Let's do the same in hggit. Ran tests against Mercurial latest (6ac860f700b5) and Mercurial 3.4.
-
- May 14, 2015
-
-
Gregory Szorc authored
filectx.renamed() returns a 2-tuple or None. memfilectx.__init__ expects the copied argument to be either None or a string. Before, we were passing a 2-tuple, leading to the memfilectx storing the wrong type. This eventually resulted in doing a key lookup against a manifest with a 2-tuple, which made manifest.c throw an error.
-
- May 13, 2015
-
-
Sean Farley authored
-
- May 12, 2015
-
-
Sean Farley authored
-
durin42 authored
-
durin42 authored
-
- May 08, 2015
-
-
Siddharth Agarwal authored
Note that one test still fails with 3.4 -- however, it is a bug in core Mercurial, only affects edge cases (broken symlinks) in the test, and is fixed in upstream stable.
-
Siddharth Agarwal authored
The old logic was broken -- it didn't work at the boundary between hg and git commits. The logic in overlayrevlog.parents handles that correctly. This is the last fix required for Mercurial 3.4.
-
Siddharth Agarwal authored
-
- May 03, 2015
-
-
Siddharth Agarwal authored
-
- Apr 29, 2015
-
-
durin42 authored
-
- Apr 22, 2015
-
-
Sean Farley authored
-
Sean Farley authored
This method had a syntax error that would have caused a crash.
-
Sean Farley authored
-
Sean Farley authored
-
Sean Farley authored
-
Sean Farley authored
-
Sean Farley authored
-
Sean Farley authored
-
Sean Farley authored
-
- Apr 17, 2015
-
-
Sean Farley authored
Also cleans up some unused imports.
-
- Apr 16, 2015
-
-
Sean Farley authored
-
- Apr 15, 2015
-
-
Sean Farley authored
-
Sean Farley authored
All errors found with flake8.
-
Sean Farley authored
-
- Apr 10, 2015
-
-
Siddharth Agarwal authored
There really is no point to this -- the sorting is expensive to compute and the structure is never actually used. For a mapfile with 1.5 million entries, this speeds up save_map from 3.6 seconds to 0.87. This is probably the limit of the speedups we can get with pure-Python code. Any further speedups will have to be made by rewriting these bits in C.
-
- Apr 13, 2015
-
-
Siddharth Agarwal authored
There are a couple of places that use StringIO for the added features it has over cStringIO, so we can't replace all its uses wholesale.
-
- Apr 10, 2015
-
-
Siddharth Agarwal authored
For a mapfile with 1.5 million entries, this speeds up save_map from 3.5 seconds to 3.1.
-
Siddharth Agarwal authored
Sorting a list of tuples is much more expensive than sorting a list of strings. For a mapfile with 1.5 million entries, this speeds up save_map from 6 seconds to 3.5.
-
Siddharth Agarwal authored
-
Siddharth Agarwal authored
-
Siddharth Agarwal authored
For a mapfile with 1.5 million entries, this speeds up load_map from 2.3 seconds to 1.8.
-
Siddharth Agarwal authored
For a mapfile with 1.5 million entries, this speeds up perfgitloadmap from 2.5 seconds to 2.3.
-
Siddharth Agarwal authored
In upcoming patches we'll make performance improvements to core routines. This extension will allow us to measure their impact.
-
- Apr 03, 2015
-
-
Siddharth Agarwal authored
In Mercurial, every commit has at least one parent -- root commits have the null revision as their parent. In Git, root commits don't have any parents. This difference needs to be papered over in hg-git for 'hg incoming' to work in Mercurial 3.4+. Note that this doesn't fix all the broken tests in default -- I haven't had time to investigate the others.
-
Siddharth Agarwal authored
Since 2bb13f2b778c in Mercurial, _walkexplicit returns a tuple, so ensure we are up to date and take the normalized path which is the first part of the tuple. Based on a patch by David Soria Parra <davidsp@fb.com>.
-
Siddharth Agarwal authored
-
- Mar 12, 2015
-
-
Siddharth Agarwal authored
The bookmark output has been stable since well before the oldest Mercurial we support, 2.8.
-
- Mar 06, 2015
-
-
Siddharth Agarwal authored
It's the version that ships with the latest Ubuntu LTS as of this writing.
-
- Mar 05, 2015
-
-
Siddharth Agarwal authored
While this has been done since the beginning of time, there's no apparent justification for it. If an imported commit works out to the same hash as an existing one, it simply won't be added to the revlog. The tests all continue to pass. There's already test coverage for reimporting commits in test-pull-after-strip.t. Also, gimport has worked this way all this while.
-