Re: [Question] How to force relative path output when using `git status --porcelain=2 -z`

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Junio,

Thank you for the quick response! That is exactly what I have encountered.

I understand that it is highly probable that the result of `git status
--porcelain=2` is unexpected, and it may not be fixed due to
consistency.

Thanks a lot!


Junio C Hamano <gitster@xxxxxxxxx> 于2024年11月6日周三 08:59写道:
>
> 叶志晟 <yzs981130@xxxxxxxxx> writes:
>
> > I am experiencing a problem when parsing the results of `git status`.
> > My primary goal is to parse the output of `git status`
> > programmatically, so I am trying to parse the result of `git status
> > --porcelain=2 -z`.
>
> Interesting finding.  Here is what I tried in my working tree, where
> I have a few cruft files left at the top, and ran these commands
> inside the 't/' subdirectory:
>
>     $ git status --porcelain=1
>     ?? +ps8-v2-reftable-iterator-reuse
>     ?? :runme.sh
>     : git t/master; git status --porcelain=2
>     ? ../+ps8-v2-reftable-iterator-reuse
>     ? ../:runme.sh
>     : git t/master; git status --porcelain=1 -z | tr '\0' '\n'
>     ?? +ps8-v2-reftable-iterator-reuse
>     ?? :runme.sh
>     : git t/master; git status --porcelain=2 -z | tr '\0' '\n'
>     ? +ps8-v2-reftable-iterator-reuse
>     ? :runme.sh
>
> I haven't dug into the history to see if these inconsistencies in
> behaviour are deliberate, but I somehow suspect that the output from
> "--porcelain=2" without "-z" is mistaken.  As the porcelain output
> is designed to be machine parseable with or without "-z" (and the
> reader of the output, the implementation of a Porcelain command, is
> responsible for turning it into relative if it prefers), it would be
> a mistake to turn relative output on based on the lack of "-z".
>
> Not that we can "fix" the output from "--porcelain=2" at this point.
> To changge them to uniformly show absolute (or relative for that
> matter), we'd need "--porcelain=3" to preserve these inconsistency
> output because it is very certain that existing users and scripts
> already rely on it.
>
> Thanks.



-- 


志晟 叶
yezhisheng.com.cn





[Index of Archives]     [Linux Kernel Development]     [Gcc Help]     [IETF Annouce]     [DCCP]     [Netdev]     [Networking]     [Security]     [V4L]     [Bugtraq]     [Yosemite]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux SCSI]     [Fedora Users]

  Powered by Linux