Re: Bug Report: git add

From: Junio C Hamano <gitster@pobox.com>
Date: Apr 7, 2011, 01:09 UTC
Message-ID: <7vbp0ihnou.fsf@alter.siamese.dyndns.org>
In-Reply-To: <20110407005750.GC28813@sigill.intra.peff.net>

Jeff King <peff@peff.net> writes:

Show 14 quoted lines

>>   mkdir settings
>>   cd settings
>>   git init
>>   touch x
>> ...
>>   cd ..
>>   git init
>>   git add settings/
>>    (should complain)
>
> If you do "git add settings" (without the slash) it will add the
> repository as a submodule.  Which is not the behavior you asked for, but
> is at least reasonable. So the real bug seems to me the fact that "git
> add settings/" and "git add settings" behave differently.

Also if "git add settings/x" does not complain, that would be a bigger issue, whose solution would probably be in the same area.

It may finally be the time to rename has_symlink_leading_path() function and enhance its feature. It happens to check leading symlinks, but its real purpose is to check if the given path is outside the working tree, and a path that is in a subdirectory with its own .git repository certainly is outside the working tree. The callers should be able to say what check is being asked for by naming the _purpose_ of the test ("is this path outside the working tree?"), not the mechanics ("does the path have a symlink that points outside?").