Re: [OUTREACHY][PATCH] diff: do not show submodule with untracked files as "-dirty"

[Phillip Wood <phillip.wood123@gmail.com>]

Hi Sangeeta

On 20/10/2020 19:10, Sangeeta NB wrote:
> Hey Phillip,
> 
>> I think we require your full name on the From line to match the
>> Signed-off-by line below (c.f.
>> https://lore.kernel.org/git/xmqqpn5dd5dv.fsf@gitster.c.googlers.com)
> 
>> That is a good summary of the issue that this change addresses, we
>> normally wrap lines at 72 characters though.
> 
> This was sent by gitgitgadget so I am unable to find how I can
> customize it. Can you help me with this? Or else I have set the Travis
> so should I send another patch using send-email?

When you create the commit message you need to get your editor to wrap 
the lines with line breaks, how you do this depends on your editor - for 
emacs you can use fill-paragraph, in vim 'gq' you should be able to 
google it for your editor.

It would be a good idea to add [Outreachy] to the beginning of the 
commit subject line as well so people can easily find outreachy related 
patches on the list (anything inside [] is removed by `git am` when the 
patch is applied)

>> This unconditionally overrides diff.ignoreSubmodules, grepping seems to
>> show that we don't have any tests that test a config value of "none".
>> There are a few that check "dirty" which we should look at to consider
>> if they still add value now we've changed the default.
> 
> Yes, sorry, noticed it now. I have now added a flag in diff_flags.
> Also, do I have to add tests that check whether diff.ignoreSubmodules
> is not being overwritten?

I think adding a test that checks diff.ignoreSubmodules=none is 
respected would be a good idea

> 
>>        if (diff_no_prefix) {
>>                options->a_prefix = options->b_prefix = "";
>> diff --git a/submodule.c b/submodule.c
>> index b3bb59f066..762298c010 100644
>> --- a/submodule.c
>> +++ b/submodule.c
>> @@ -1678,6 +1678,8 @@ unsigned is_submodule_modified(const char *path, int ignore_untracked)
>>        strvec_pushl(&cp.args, "status", "--porcelain=2", NULL);
>>        if (ignore_untracked)
>>                strvec_push(&cp.args, "-uno");
>> +     else
>> +        strvec_push (&cp.args, "--ignore-submodules=none");
>>
>> We need to do this as a consequence of changing the default for
>> '--ignore-submodules`, we should think what the consequences are for
>> other users of `git status` and whether we need to do something similar
>> there if diff.ignoreSubmodules is not set.
>>
> 
> Oh okay. I understood what you said. But I couldn't figure out how to
> do that.

I'm not sure if we need to do this or not, I was hoping to get some 
input from the list

> As all the tests of status were passing so I don't think
> there might be any issue with this.

Possibly, though I think it is more likely that we're not testing 
anything that gets broken by this change.


Best Wishes

Phillip

>> I think we want to keep testing that we get the correct output when
>> --ignore-submodules=none as well as maybe adding a couple of new tests
>> that check the default in this file, rather than changing the expected
>> output.
>>
> 
> Thanks, I have added more tests.
> 
> Thanks and Regards,
> Sangeeta
>