[PATCH] branch: Forbid to create local branches with confusing names

Without this patch, git allows to do something like this:
  git branch remotes/origin/master
  git branch refs/remotes/origin/master
  git branch heads/master
  git branch tags/v3.4
All of these local branch names lead to severe confusion,
not only for a user but also for git itself.

This patch forbids to create local branches, with names which start with any of

  refs/
  heads/
  remotes/
  tags/

With this patch, you might still create these kind of local branches, but now you have to additionally specify the '-f' option.

Signed-off-by: Ingo Rohloff <ingo.rohloff@lauterbach.com>
---
 branch.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

 This patch refers way back to the discussion from 2014:
   From: Josef Wolf
   To: git@vger.kernel.org
   Subject: error: src refspec refs/heads/master matches more than one.
   Date: Fri, 14 Feb 2014 12:31:36 +0100
 See for example here:
   https://public-inbox.org/git/20140214113136.GA17817@raven.inka.de/

 The origin of the problem is, that git has (almost) no constraints
 what kind of names are allowed for local branches.
 There nowadays is a constraint that you are NOT allowed to create
 a branch which is called HEAD. See commit 16169285f1e6 ("Merge
 branch 'jc/branch-name-sanity'", 2017-11-28).

 In the old mail thread a lot of potential constraints for local
 branch names were discussed; in particular a lot of strategies
 were discussed what kind of local branch names might be a problem
 (the gist is: avoid ambiguities, by finding out which names
 lead to ambiguities NOW).

 I personally think it makes more sense to forbid a much
 bigger class of confusing branch names.
 In particular I think all local branch names starting with
   refs/
   heads/
   remotes/
   tags/
 should be forbidden (per default, can still be forced).
 This also avoids trouble for an unforseeable future.
 Example:
   git branch remotes/blub/master
 This might not be a problem now, because you have no 
 remote called "blub" right now.
 But if you later use
   git remote add blub <URL>
   git fetch blub
 you very likely run into trouble.

 The above approach still allows you to create local branches
 with a name of the form
    <remote name>/<something ...>
 but I cannot see how this might be avoided; remotes might
 be added later so what would you do in the case that a local
 branch already existed named like the remote or a remote
 tracking branch.

 With the proposed constraints you are at least are able to use
    heads/<remote name>/<something ...>
    remotes/<remote name>/<something ...>
 to differentiate between the two.

 This really is an issue; people seem to stumble over it
 and I guess this is particularly true if you control git
 via scripts. See for example (two years later):
   From: Duy Nguyen
   To: Junio C Hamano
   Cc: Git Mailing List <git@vger.kernel.org>,
   Subject: Re: error: src refspec refs/heads/master matches more than one.
   Date: Wed, 23 Mar 2016 18:17:05 +0700

 So with this patch I want to pick up this old discussion yet again.

 This code can probably be done a lot better I guess, but I wanted to
 send in something, to start the discussion.