Re: [PATCH] Functions for updating refs.

From: Carlos Rica <jasampler@gmail.com>
Date: Sep 4, 2007, 23:32 UTC
Message-ID: <1b46aba20709041632x60ee4d3eweae9f5217d2f3b86@mail.gmail.com>
In-Reply-To: <7v642qnwr7.fsf@gitster.siamese.dyndns.org>

2007/9/4, Junio C Hamano <gitster@pobox.com>:

Show 47 quoted lines

> Carlos Rica <jasampler@gmail.com> writes:
>
> > diff --git a/refs.c b/refs.c
> > index 09a2c87..4fd5065 100644
> > --- a/refs.c
> > +++ b/refs.c
> > @@ -1455,3 +1455,35 @@ int for_each_reflog(each_ref_fn fn, void *cb_data)
> >  {
> >       return do_for_each_reflog("", fn, cb_data);
> >  }
> > +
> > +int update_ref_or_die(const char *action, const char *refname,
> > +                             const unsigned char *sha1,
> > +                             const unsigned char *oldval, int flags)
> > +{
> > +     static struct ref_lock *lock;
> > +     lock = lock_any_ref_for_update(refname, oldval, flags);
> > +     if (!lock)
> > +             die("Cannot lock the ref '%s'.", refname);
> > +     if (write_ref_sha1(lock, sha1, action) < 0)
> > +             die("Cannot update the ref '%s'.", refname);
> > +     return 0;
> > +}
> > +
> > +int update_ref_or_error(const char *action, const char *refname,
> > +                             const unsigned char *sha1,
> > +                             const unsigned char *oldval, int quiet)
> > +{
> > +     static struct ref_lock *lock;
> > +     lock = lock_any_ref_for_update(refname, oldval, 0);
> > +     if (!lock) {
> > +             if (!quiet)
> > +                     error("Cannot lock the ref '%s'.", refname);
> > +             return 1;
> > +     }
> > +     if (write_ref_sha1(lock, sha1, action) < 0) {
> > +             if (!quiet)
> > +                     error("Cannot update the ref '%s'.", refname);
> > +             return 1;
> > +     }
> > +     return 0;
> > +}
>
> This makes me wonder three things:
>
>  - Why doesn't "or_error" side allow "flags" as "or_die" one?
>    Could the 'quiet' option become part of "flags" perhaps?

I saw that the only code that needed the flags was the builtin-update-ref.c, and it also needed to die(). The others I saw only want that parameter set to 0. builtin-tag.c was doing die() also, not using flags, though.

>  - They look quite similar.  Is it a good idea to refactor them
>    further, or they are so small it does not matter?

I would like to know how to refactor it, however the code I saw sometimes needs to call die(), others to error() and others need to get only a value of success or not.

The function die() returns 128 and terminates the program, prepending "fatal: " in the message, while error() doesn't exit and prepend "error: ", so they were very different and I resolved to separate them.

Another option is returning different error codes, so the caller could decide what to output in each case, but I thought that these functions were only useful to unify those instructions with those error messages for this common operation that many builtins use, specially when they come from scripts who call to "git update-ref".

>
>  - Why isn't lock released with unlock_ref()?

I inspected this some weeks ago, and I finally came to think that it is released in the write_ref_sha1 call after the lock. But the code was complex, can someone confirm this?