Re: [PATCH/RFC][GSoC] make "git diff --no-index $directory $file" DWIM better.

From: Eric Sunshine <sunshine@sunshineco.com>
Date: Mar 15, 2015, 18:04 UTC
Message-ID: <CAPig+cQgvwd=5hLGeOpY1r_476HbpUarumsu_QDU4HwB7ynXmw@mail.gmail.com>
In-Reply-To: <CAHLaBN+93mp6PQmtfjOHSvfW7iwDXwPitGQ5W1am9KBm9EZV2Q@mail.gmail.com>

In addition to the points raised by Matthieu and Thomas...

On Sun, Mar 15, 2015 at 11:35 AM, Yurii Shevtsov <ungetch@gmail.com> wrote:

> make "git diff --no-index $directory $file" DWIM better.

Specify the area affected by the change, followed by a colon, followed by the change summary. Drop the period at the end. So, something like:

    diff: improve '--no-index <directory> <file>' DWIMing

> Changes 'git diff --no-index $directory $file' behaviour.
> Now it is transformed to 'git diff --no-index $directory/&file $file'
> instead of throwing an error.

Write in imperative mood, so "Change" rather than "Changes". By itself, the first sentence isn't saying much; it would read better if you combined the two sentences into one.

Show 14 quoted lines

> Signed-off-by: Yurii Shevtsov <ungetch <at> gmail.com>
> ---
> diff --git a/diff-no-index.c b/diff-no-index.c
> index 265709b..4e71b36 100644
> --- a/diff-no-index.c
> +++ b/diff-no-index.c
> @@ -97,8 +97,25 @@ static int queue_diff(struct diff_options *o,
>         if (get_mode(name1, &mode1) || get_mode(name2, &mode2))
>                 return -1;
>
> -       if (mode1 && mode2 && S_ISDIR(mode1) != S_ISDIR(mode2))
> -               return error("file/directory conflict: %s, %s", name1, name2);
> +       if (mode1 && mode2 && S_ISDIR(mode1) != S_ISDIR(mode2)) {
> +               struct strbuf dirnfile;

Is this name supposed to stand for "dir'n'file", a shorthand for "dir-and-file"? Perhaps a simpler more idiomatic name such as "path" would suffice. Also, you can initialize the strbuf here at point of declaration:

    struct strbuf path = STRBUF_INIT;

Show 7 quoted lines

> +               const char *dir, *file;
> +               char *filename;
> +               int ret = 0;
> +
> +               dir = S_ISDIR(mode1) ? name1 : name2;
> +               file = (dir == name1) ? name2 : name1;
> +               strbuf_init(&dirnfile, strlen(name1) + strlen(name2) + 2);

Unless this is a performance critical loop where you want to avoid multiple re-allocations, it's customary to pass 0 for the second argument. Doing so makes the code a bit easier to read and understand, and avoids questions like this one: Why are you adding 2 to the requested length? I presume that you're taking the "/" and NUL into account, however, strbuf takes NUL into account on its own as part of its contract, so you probably wanted to add only 1.

> +               strbuf_addstr(&dirnfile, dir);
> +               if (dirnfile.buf[dirnfile.len - 1] != '/')

I don't know how others feel about it, but it makes me a bit uncomfortable to see potential access of array item -1. I suppose, in this case, neither name will be zero-length, however, I'd still feel more comfortable seeing that documented formally, either via assert():

    assert(dirnfile.len > 0);
    if (dirnfile.buf[dirnfile.len - 1] != '/')

or:

    if (dirnfile.len > 0 && dirnfile.buf[dirnfile.len - 1] != '/')

Show 16 quoted lines

> +                       strbuf_addch(&dirnfile, '/');
> +               filename = strrchr(file, '/');
> +               strbuf_addstr(&dirnfile, filename ? (filename + 1) : file);
> +               ret = queue_diff(o, dirnfile.buf, file);
> +               strbuf_release(&dirnfile);
> +
> +               return ret;
> +       }
>
>         if (S_ISDIR(mode1) || S_ISDIR(mode2)) {
>                 struct strbuf buffer1 = STRBUF_INIT;
> --
>
> I hope I understood task correct. I think this patch requires writing
> additional tests, so that's what I'm going to do now.
> --