Re: [PATCH v5] diff: add config option relative

From: Shourya Shukla <shouryashukla.oo@gmail.com>
To: laurent@spkdev.net
Cc: congdanhqx@gmail.com, git@vger.kernel.org, gitster@pobox.com,
	phillip.wood@dunelm.org.uk, sandals@crustytoothpaste.net
Subject: Re: [PATCH v5] diff: add config option relative
Date: Mon, 18 May 2020 17:16:30 +0530	[thread overview]
Message-ID: <20200518114630.GA30622@konoha> (raw)
In-Reply-To: <20200518094021.GA2069@spk-laptop>

Hello Laurent,

I have not worked in this part of Git before, please pardon my
ignorance. I went through the conversations held in this thread
Here are my inferences:

> The `diff.relative` boolean option set to `true` show only changes on
> the current directory and show relative pathnames to the current
> directory.

> Teach --no-relative to override earlier --relative

I think the message can be written better, maybe something like:

    The `diff.relative` boolean option set to `true` shows only changes
    in the current directory/value specified by the `path` argument of
    the `relative` option and shows pathnames relative to the
    aforementioned directory.

    Teach `--no-relative` to override the earlier `--relative`

> --- a/Documentation/config/diff.txt
> +++ b/Documentation/config/diff.txt
> @@ -105,6 +105,10 @@ diff.mnemonicPrefix::
>  diff.noprefix::
>  	If set, 'git diff' does not show any source or destination prefix.
> 
> +diff.relative::
> +	If set to 'true', 'git diff' does not show changes outside of the directory
> +	and show pathnames relative to the current directory.

In the second last line, it would be better to write is as:

    ... does not show changes outside of the current directory/the path
    provided as the argument in `relative`?

Even my version of the above line seems very crude and innaccurate, but
I think that we should take into account the `path` passed down to us
by the `--relative[=<path>]` option.

Moving on, what I wondered was that maybe making it a `true/false`
option would be better? Something like:

    git diff --relative=false //override the diff.relative setting
    git diff --relative=true <path> //works like the usual `relative`

And by the usual, I mean :
    git diff --relative[=<path>]

What do you think? I think it would be better than adding a new option.

Also, if I am not mistaken:

> @@ -5195,8 +5202,7 @@ static int diff_opt_relative(const struct option *opt,
>  {
>  	struct diff_options *options = opt->value;
> 
> -	BUG_ON_OPT_NEG(unset);
> -	options->flags.relative_name = 1;
> +	options->flags.relative_name = !unset;

This is the overriding part right?

BTW you mention the `no-relative` setting which will override the
`relative` option. I am not able to see where exactly is the occurence
of the `no-relative` option in the code? What I mean is that I am not
able to observe where exactly does the C code identify a `no-relative`
option from the command line input? Did I miss something out here?

All nits aside, it looks like a very good concept to me. Nice work! :)

Regards,
Shourya Shukla