On Sun, 2008-11-09 at 02:20 +0100, Neels J Hofmeyr wrote:
> I did find some peculiar behaviour (attached) when changing the very first
> line of a function (which "trickles" down consecutive changes), or when
> changing a function name. Fair enough, I read the diffs anyway.
>
> It's not completely 100% reliable, but it's really really close, considering
> that real-world changes will almost certainly show all the correct function
> names (and maybe some more). What a great help this will be! Thanks a lot!
About reliability: Yes, it's poor. It just uses diff's
"-p/--show-c-function" option which is really hacky and simply prints
the last-seen line that looks like a function declaration before the
hunk it prints. It doesn't work properly for many situations including
added/deleted functions, and when the function declaration appears in
the context lines of the hunk, and when editing the doc string before a
function declaration. Both GNU diff's and svn diff's "-p" options are
like this.
- Julian
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe_at_subversion.tigris.org
For additional commands, e-mail: dev-help_at_subversion.tigris.org
Received on 2008-11-09 21:34:48 CET