Re: [PATCH 2/2] clean: avoid to differentiate untracked and ignored when appropriate

From: Elijah Newren <newren@gmail.com>
Date: Feb 16, 2022, 04:16 UTC
Message-ID: <CABPp-BFnrd4=DM-xQtw+j=LRkA4fwYp1EZ2j3cBVjZRsYPU-9Q@mail.gmail.com>
In-Reply-To: <20220215221615.20683-2-patrick.marlier@gmail.com>

On Tue, Feb 15, 2022 at 2:16 PM Patrick Marlier <patrick.marlier@gmail.com> wrote:

Could I suggest a bit longer commit message?  Perhaps:

""" clean: avoid traversing into untracked dirs when unnecessary

When deleting all untracked and ignored files and any nested repositories (such as with `git clean -ffdx`), we do not need to recurse into an untracked directory to see if any of the entries under it are ignored or a nested repository. Special case this condition to avoid unnecessary recursion. """

Show 31 quoted lines

> ---
>  builtin/clean.c  |  4 +++-
>  t/t7300-clean.sh | 24 ++++++++++++++++++++++++
>  2 files changed, 27 insertions(+), 1 deletion(-)
>
> diff --git a/builtin/clean.c b/builtin/clean.c
> index 18b37e3fd9..1b1454d052 100644
> --- a/builtin/clean.c
> +++ b/builtin/clean.c
> @@ -978,7 +978,9 @@ int cmd_clean(int argc, const char **argv, const char *prefix)
>                 remove_directories = 1;
>         }
>
> -       if (remove_directories && !ignored_only) {
> +       if (remove_directories && ignored && !exclude_list.nr && force > 1)
> +               ; /* No need to recurse to look for ignored files */
> +       else if (remove_directories && !ignored_only) {
>                 /*
>                  * We need to know about ignored files too:
>                  *
> diff --git a/t/t7300-clean.sh b/t/t7300-clean.sh
> index 0399701e62..684eba914b 100755
> --- a/t/t7300-clean.sh
> +++ b/t/t7300-clean.sh
> @@ -788,4 +788,28 @@ test_expect_success 'traverse into directories that may have ignored entries' '
>         )
>  '
>
> +test_expect_success 'avoid traversing into untracked directories' '
> +       test_when_finished rm -f output error trace.* &&
> +       test_create_repo avoid-traversing-untracked-hierarchy &&

I think you may have been copying from an example earlier in the file that I like wrote. As someone else recently pointed out to me, t/test-lib-functions.sh points out that test_create_repo is deprecated and we should just use "git init <dirname>" instead.

Show 24 quoted lines

> +       (
> +               cd avoid-traversing-untracked-hierarchy &&
> +
> +               mkdir -p untracked/subdir/with/b &&
> +               mkdir -p untracked/subdir/with/a &&
> +               >untracked/subdir/with/a/random-file.txt &&
> +
> +               GIT_TRACE2_PERF="$TRASH_DIRECTORY/trace.output" \
> +               git clean -ffdx
> +       ) &&
> +
> +       # Make sure we only visited into the top-level directory, and did
> +       # not traverse into the "untracked" subdirectory since it was excluded
> +       grep data.*read_directo.*directories-visited trace.output |
> +               cut -d "|" -f 9 >trace.relevant &&
> +       cat >trace.expect <<-EOF &&
> +        ..directories-visited:1
> +       EOF
> +       test_cmp trace.expect trace.relevant
> +'
> +
>  test_done
> --
> 2.35.1

Looks good, other than those minor items.  Thanks for working on this!