Re: [PATCH] t6300: values containing ')' are broken in ref formats

From: Kousik Sanagavarapu <five231003@gmail.com>
Date: Nov 7, 2024, 02:34 UTC
Message-ID: <ZywnS4j7gxn53N+G@five231003>
In-Reply-To: <20241106185511.GB880133@coredump.intra.peff.net>

On Wed, Nov 06, 2024 at 01:55:11PM -0500, Jeff King wrote:

Show 12 quoted lines

> On Wed, Nov 06, 2024 at 09:24:08AM +0530, Kousik Sanagavarapu wrote:
> 
> > One way to fix this is refactoring the way in which we parse our format
> > string.  Although this would mean we would have to do a huge refactoring
> > as this step happens very high up in the call chain.
> > 
> > Therefore, support including parenthesis characters in "<string>" by
> > instead giving their hexcode equivalents - as a for-now hack.
> 
> So if I understand this is just expanding %<hex> and nothing else? That
> seems like the worst of both worlds. Now "%" is magic in these value
> strings, breaking compatibility, but

Yeah, I agree that this might be the worst of both worlds after I read your reply to Junio. It indeed is a hack - just trying to fix the parenthesis case and not taking into account

- backwards compatibility with regards to '%'.
- not being able to do

> we didn't buy ourselves the flexibility to do arbitrary comparisons like:
> 
>   %(if:equals=%(upstream:lstrip=3))%(refname:short)%(then)...