Re: [PATCH v2 1/3] avoid unnecessary worktrees traversing

From: Rubén Justo <rjusto@gmail.com>
Date: Jan 19, 2023, 23:26 UTC
Message-ID: <f369f68e-5921-095f-e3e8-3d324d1688f4@gmail.com>
In-Reply-To: <xmqqilh21ale.fsf@gitster.g>

On 19-ene-2023 13:24:45, Junio C Hamano wrote:

Show 15 quoted lines

> Rubén Justo <rjusto@gmail.com> writes:
> 
> > -static void reject_rebase_or_bisect_branch(const char *target)
> > +static int ishead_and_reject_rebase_or_bisect_branch(const char *target)
> 
> The original name was already horrible but it became much worse by
> acquiring a non-word "ishead" as part of it X-<.
> 
> At least let's replace "rebase or bisect" with something a bit more
> generic, extensible, and shorter phrase.  For example, isn't the
> point of having the function was to give us a mechansim to see if
> the branch with the given name is not to be modified because it is
> being worked on elsewhere?  "The branch is in use" would be a good
> phrase to express such a concept, so die_if_branch_is_in_use() or
> something along that line may be easier to grok.

I agree, the naming is ugly.

The idea is to return, if not die(), from the iteration we are doing in that function, whether the branch is checked out in any worktree. That information allows us later, if we know in advance that no HEAD needs to be adjusted, to avoid calling replace_each_worktree_head_symref(), saving us a new and unnecessary traversal of the worktrees.

There is a second idea to, in next commits, return also if the branch is an unborn branch.

die_if_branch_is_in_use() is a better name for reject_rebase_or_... but don't know how it fits with these ideas. I'm open to suggestions.

I'll reroll with a better approach.

Thank you.