Re: [WIP] Status update on merge-recursive in C

From: Johannes Schindelin <johannes.schindelin@gmx.de>
Date: Jul 27, 2006, 09:41 UTC
Message-ID: <Pine.LNX.4.63.0607271127520.29667@wbgn013.biozentrum.uni-wuerzburg.de>
In-Reply-To: <7vu053o2g7.fsf@assigned-by-dhcp.cox.net>

Hi,

On Thu, 27 Jul 2006, Junio C Hamano wrote:

Show 16 quoted lines

> Johannes Schindelin <Johannes.Schindelin@gmx.de> writes:
> 
> > On Sun, 9 Jul 2006, Pavel Roskin wrote:
> >
> >> - variable declarations in the middle of the code (allowed by c99 but
> >> not by ANSI C) and incomplete function declarations (no "void").  It's
> >> not C++, let's stick to the C standards.
> >...
> >> I'm attaching a patch that fixes everything except variable
> >> declarations.
> >
> > May I ask you a favour? Can you redo the patch in, say, two weeks? Thank 
> > you!
> 
> Now it's about two weeks so maybe we can apply this on top to
> clean up and prod you to continue ;-)?

*grin*

Actually, we continued a little with it, but I mainly waited what would happen to "git-read-tree --rename"...

I will send an updated patch soon. The remaining TODOs after that are:

$ grep TODO merge-recursive.c 
 * TODO: we should not have to copy the SHA1s around, but rather reference
 * TODO: check if we can just reuse the active_cache structure: it is already
 * TODO: this can be streamlined by refactoring builtin-read-tree.c
 * TODO: this can be streamlined by refactoring builtin-read-tree.c
 * TODO: this can be streamlined by refactoring builtin-write-tree.c
/* General TODO: no javaIfiCation: rename updateCache to update_cache */
 * TODO: once we no longer call external programs, we'd probably be better off
/* General TODO: get rid of all the debug messages */
                /* TODO: refactor, so that 1/2 are not needed */

So, the biggest problem I see: refactoring builtin-read-tree. I really could use some help there -- I already tried 3 times to come up with a nice abstraction for unpack_trees() which could be used as a library function.

Ciao, Dscho