[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Monotone-devel] diff -p, now available from a venge.net near you
From: |
Zack Weinberg |
Subject: |
[Monotone-devel] diff -p, now available from a venge.net near you |
Date: |
Thu, 6 Jul 2006 00:53:39 -0700 |
I pushed out what I've got on diff -p; get it from
net.venge.monotone.diff-p branch. Issues yet to be resolved:
1) I'm using Lua pattern matching to find the enclosing functions. I
did this because the pattern to match comes from a Lua hook, so it
seemed most consistent to use that language's regexp syntax, even if
that syntax is, um, abnormal. I'd be perfectly glad to use
boost::regex instead. Thoughts?
2) lua_hooks::patternmatch: Is this the Right Way To Do It? If not, what is?
3) I boldly added a bunch of short options: -c (--context), -u
(--unified), -p (--show-encloser), -v (--verbose). My fingers
appreciate these, but I can see an argument for not doing -c/-u/-p.
(I am 100% certain -v should mean --verbose. No, it has nothing to do
with the rest of the patch.)
4) Mooore teeeests. I've got some randomly generated files which I
will use to augment the existing diff_output_formats test tomorrow,
when I'm awake, but ideas for really horrible test cases would be
welcome. (Is there a way to override hooks for just one test? It
would be good to be able to put the per-filename encloser-finder
mechanism through its paces.
Meta-issue: How verbose do y'all like your change logs? I'm used to
GCC and their expectation that you mention every. single. function you
changed, in a particular rigidly defined format...
g'night,
zw
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- [Monotone-devel] diff -p, now available from a venge.net near you,
Zack Weinberg <=