Re: [PATCH RFC] parse_object: pass on the original sha1, not the replaced one

From: Nguyen Thai Ngoc Duy <pclouds@gmail.com>
Date: Aug 7, 2010, 04:03 UTC
Message-ID: <AANLkTim9vMiEuBJJdCdGqjTOz8cs-nH+LyJF-rPGOTvS@mail.gmail.com>
In-Reply-To: <201008051341.08632.chriscool@tuxfamily.org>

On Thu, Aug 5, 2010 at 9:41 PM, Christian Couder <chriscool@tuxfamily.org> wrote:

> It looks like parse_commit() is buggy regarding replaced objects. But I am not
> sure how it should be fixed.

It could be fixed the same way you did with parse_object(): replace read_sha1_file() with read_sha1_file_repl(). You would also need to fix parse_tree() and parse_tag(). But..

> Anyway if you use parse_object(), then you don't need parse_commit(). So if
> possible you should use parse_object() instead of both lookup_commit() and
> parse_commit().

That's how those functions are used. For example, in traverse_commit_list(), lookup_*() may be called and uninteresting objects marked UNINTERESTING. Later on in process_{tree,blob,tag}, parse_* may be called if their content is interesting.

To me, the fix above will leave a gap when object->sha1 is the original sha1, until parse_*() is called. It just does not sound good.

Or, you could lookup_replace_object() inside lookup_{object,tree,commit,tag,blob} and update object->sha1. Hm..

-- 
Duy