Re: [PATCH/RFC] blame: respect "core.ignorecase"

From: Jeff King <peff@peff.net>
Date: Sep 10, 2012, 21:41 UTC
Message-ID: <20120910214152.GA2139@sigill.intra.peff.net>
In-Reply-To: <7vtxv5lfsv.fsf@alter.siamese.dyndns.org>

On Mon, Sep 10, 2012 at 02:38:08PM -0700, Junio C Hamano wrote:

Show 14 quoted lines

> > Thanks, that helped. I got excited when I saw the "icase" in the
> > comments and thought it might already be implemented. But it looks like
> > it is still to be done. :)
> 
> Yeah, some are tongue-in-cheek (e.g. I do not know what "recursive
> pathspec" even means), but "noglob" probably is an urgent need from
> correctness point of view for people who are writing Porcelain and
> want to interact with a history that records funny filenames.
> Currently you can "git <cmd> 'foo\*'" to match a path that is
> exactly 'foo*' (because it matches) but you also have to hope there
> is no other paths that happens to match that pattern.  A script that
> grabs paths out of ls-files output and then tries to use them as
> pathspec would want to have a way to say "This is literal. Do not
> honor globs in it".

I agree that the automatic globbing is currently a problem (although one that comes up quite infrequently; I guess people use sane pathnames). But I would think for that particular use case, you would not want to do a per-glob prefix for that, but would rather use a command-line switch. IOW, would you rather do:

  git ls-files |
  while read fn; do
    echo ":(noglob)$fn"
  done |
  xargs git log --stdin --

or:

  git ls-files |
  xargs git log --stdin --no-glob-pathspec --

-Peff