Felipe Contreras <felipe.contreras@xxxxxxxxx> writes: > On Tue, May 21, 2013 at 3:54 PM, Thomas Gummerer <t.gummerer@xxxxxxxxx> wrote: >> Currently the __git_ps1 git prompt gives the following error with a >> repository converted by git-svn, when used with zsh: >> >> __git_ps1_show_upstream:19: bad pattern: svn_remote[ >> >> This was introduced by 6d158cba (bash completion: Support "divergence >> from upstream" messages in __git_ps1), when the script was for bash >> only. Make it compatible with zsh. >> >> Signed-off-by: Thomas Gummerer <t.gummerer@xxxxxxxxx> > > This patch is fine by me. I would like to see an example of how to > trigger the issue with a standalone command in the commit message, but > it's not necessary. It would also make sense to address the comment > from Szeder that does raise questions about other places in the code > where 'array[ $foo ]' is used, maybe there's a caveat we are not > considering, or maybe your use-case did not execute that code. Yes, the code was not executed, it will be fixed in the re-roll. > And finally, I don't recall seen 'set -a' used elsewhere in the code. > If memory serves well, we have replaced 'local -a foo=value' with > 'local -a foo\nfoo=value' before to fix zsh issues, I think that would > be safer. Yes, thanks for the suggestion, that works. > But this patch is needed regardless, that, or the patch that broke > things should be reverted for v1.8.3. I don't think anything should or can be reverted, as this code was introduced in 2010, but nobody triggered it until now. > Thomas, if you don't have time, let me know and I can take a look. > > Cheers. > > -- > Felipe Contreras Thanks, Thomas -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html