Volume XXII, number 280Wednesday, October 7, 2026Latest message 1 hour ago

The Git List

News and archive of git@vger.kernel.org, since April 2005

[GSoC] microproject idea: improve corrupt patch location reporting in git am

2 messages between Mar 15, 2026 and Mar 16, 2026, from Jerry Wang, Karthik Nayak.

Plain Markdown or JSON for tools and agents.

Jerry WangMar 15, 2026, 22:44 UTC on lore
Hi,
My name is Jialong Wang, and I plan to apply to Git for GSoC 2026.

I have been reading the Git contribution documentation, including Documentation/CodingGuidelines, Documentation/SubmittingPatches, and Documentation/MyFirstContribution.txt. I have also built Git from source locally and started looking for a small microproject.

I found an older open issue that still seems reproducible on my setup:

when git am encounters a corrupt patch, it reports "corrupt patch at line <n>", but that line number does not correspond to the original mail/mbox as the user sees it.

Related discussion I found:
  - bug report from 2019-10-02:
    https://public-inbox.org/git/20191002184546.GA22174@generichostname/
  - follow-up patch from 2019-10-04:
    https://public-inbox.org/git/ec38908d05f0d40190173158ef3f0753fa9f1184.1570226253.git.liu.denton@gmail.com/
  - GitGitGadget issue #374:
    https://github.com/gitgitgadget/git/issues/374

I also saw review feedback mentioning additional test coverage for cases such as stdin input and running from a subdirectory.

I reproduced the issue locally with Git 2.51.0 on 2026-03-15, and I am considering working on a small patch in this area.

Before I start, I wanted to ask whether this still looks like an appropriate microproject for a GSoC applicant, and whether improving the error location/reporting plus adding the missing tests would be a good direction.

Thanks, Jialong

Karthik NayakMar 16, 2026, 11:17 UTC in reply to Jerry Wang on lore

Re: [GSoC] microproject idea: improve corrupt patch location reporting in git am

Jerry Wang <jerrywang1201@gmail.com> writes:
Show 36 quoted lines
> Hi,
>
> My name is Jialong Wang, and I plan to apply to Git for GSoC 2026.
>
> I have been reading the Git contribution documentation, including
> Documentation/CodingGuidelines, Documentation/SubmittingPatches, and
> Documentation/MyFirstContribution.txt. I have also built Git from
> source locally and started looking for a small microproject.
>
> I found an older open issue that still seems reproducible on my setup:
>
> when git am encounters a corrupt patch, it reports "corrupt patch at
> line <n>", but that line number does not correspond to the original
> mail/mbox as the user sees it.
>
> Related discussion I found:
>   - bug report from 2019-10-02:
>     https://public-inbox.org/git/20191002184546.GA22174@generichostname/
>   - follow-up patch from 2019-10-04:
>     https://public-inbox.org/git/ec38908d05f0d40190173158ef3f0753fa9f1184.1570226253.git.liu.denton@gmail.com/
>   - GitGitGadget issue #374:
>     https://github.com/gitgitgadget/git/issues/374
>
> I also saw review feedback mentioning additional test coverage for
> cases such as stdin input and running from a subdirectory.
>
> I reproduced the issue locally with Git 2.51.0 on 2026-03-15, and I am
> considering working on a small patch in this area.
>
> Before I start, I wanted to ask whether this still looks like an
> appropriate microproject for a GSoC applicant, and whether improving
> the error location/reporting plus adding the missing tests would be a
> good direction.
>
> Thanks,
> Jialong

I see that you already sent in a patch. For what its worth, the General information page also covers how to approach microprojects [1].

That said, this is a good microproject. It doesn't touch too many files, is not too complex and gives you a good starting point with the mailing list.

[1]: https://git.github.io/General-Microproject-Information/

Regards, Karthik

Back to recent threads