Re: [PATCH] Fix concat_path() functions to work for edge-cases too.

[Kevin J. McCarthy]

On Fri, May 08, 2026 at 12:30:01PM -0400, Derek Martin wrote:

On Fri, May 08, 2026 at 08:59:35PM +0800, Kevin J. McCarthy wrote:

This fixes the edge cases tested by Derek Martin.

BTW, I only later noticed there is also mutt_buffer_concatn_path(),
which has the same bugs, and which is used in exactly one place, and I
do not see any obvious reason why it should need to be distinguished.
Probably that function should be deleted and replaced with
mutt_buffer_concat_path().

Thanks, I'll take a look.



I'll put on my todo to make it "fixed" regardless though.

--
Kevin J. McCarthy
GPG Fingerprint: 8975 A9B3 3AA3 7910 385C  5308 ADEF 7684 8031 6BDA

I think this is a special case of expansion, with a size-constrained file part. The directory is always a DT_PATH which always has the / removed, and the front of the file is also constrained by the expansion code.

signature.asc
Description: PGP signature