git@vger.kernel.org mailing list mirror (one of many)
 help / color / mirror / code / Atom feed
From: Eric Wong <e@80x24.org>
To: Lukas Pupka-Lipinski <lukas.pupkalipinski@lpl-mind.de>
Cc: git@vger.kernel.org
Subject: Re: git-svn: Skip commit if all items of a commit are ignored by ignore configuration
Date: Sun, 29 Mar 2020 23:24:14 +0000	[thread overview]
Message-ID: <20200329232414.GA27925@dcvr> (raw)
In-Reply-To: <1467670a-a9eb-f230-4ec0-bd6d54411a69@lpl-mind.de>

Lukas Pupka-Lipinski <lukas.pupkalipinski@lpl-mind.de> wrote:
> 
> I used the ignore-paths option to ignore a lot of stuff I don’t need. The
> ignore pattern works well, but it could and up in empty commits. So just the
> message without any modifications / changes. The patch below skip a commit
> if all changes are ignored by the ignore-paths option.

Hi Lukas, this seems like an incompatible change, but it also
matches the intent of the user if they use ignore-paths (which
AFAIK is a rarely-used feature).

I guess it's OK to make it the default behavior, but maybe
others have objections...

> Signed-off-by: Lukas Pupka-Lipinski <lukas.pupkalipinski@lpl-mind.de>

Thanks :>  More comments inline...

> diff --git a/perl/Git/SVN/Ra.pm b/perl/Git/SVN/Ra.pm
> index 56ad9870bc..a5a6c0b774 100644
> --- a/perl/Git/SVN/Ra.pm
> +++ b/perl/Git/SVN/Ra.pm
> @@ -457,13 +457,22 @@ sub gs_fetch_loop_common {
>              $find_trailing_edge = 0;
>          }
>          $SVN::Error::handler = $err_handler;
> -
> +

This patch (and your other) are both white-space damaged,
but there doesn't need to be a whitespace change, there.

I'm not sure which mail client you use, but
https://git-send-email.io/ documents git-send-email
as being shipped with Git-for-Windows.

Personally, I'm fine with attachments if they can go through
without whitespace damage (not speaking for the rest of
git@vger).  You can test by emailing yourself and trying
"git am" on it.

>          my %exists = map { $_->path => $_ } @$gsv;
>          foreach my $r (sort {$a <=> $b} keys %revs) {
>              my ($paths, $logged) = @{delete $revs{$r}};
> -

Again, no extraneous whitespace changes, please.

>              foreach my $gs ($self->match_globs(\%exists, $paths,
>                                                 $globs, $r)) {
> +
> +
> +                my $fetcher=Git::SVN::Fetcher->new($gs);

Initializing Git::SVN::Fetcher here and closing it is an
expensive operation since it calls git-config(1) many times.
This is especially bad for most users who do not use
--ignore-paths at all.

I know git-svn isn't fast, but I've been hoping to find spare
time to make it use fast-import and speed it up, eventually.

Instead, I think it's possible to bail out of Git::SVN::do_fetch
and still skip the commit without extra network traffic.  I
suggest you try that route, instead.

> +
> +                my $skip=$self->is_empty_commit($paths,$fetcher);
> +                if ($skip){
> +                    print "skip commit $r\n";
> +                    next;
> +                }
> +                $fetcher->close_edit();
>                  if ($gs->rev_map_max >= $r) {
>                      next;
>                  }

Style: put whitespace between operators for readability
and consistency with the rest of our code:

	my $skip = $self->is_empty_commit($paths, $fetcher);
        if ($skip) {
		print "skip commit $r\n";
		...

We also use hard tabs for indentation.

> @@ -506,6 +517,21 @@ sub gs_fetch_loop_common {
> Git::SVN::gc();
>  }
> 
> +sub is_empty_commit{
> +    my ($self, $paths,$fetcher) = @_;
> +    my $path="";
> +    foreach $path (keys %$paths){
> +        unless (defined $path && -d $path ){

The `-d' check is invalid.  "git-svn fetch" never touches the
working tree.

> +            my $ignored=$fetcher->is_path_ignored($path);
> +            if (!$ignored){
> +                return 0;
> +            }
> +        }
> +    }
> +    return 1;
> +}
> +
> +
>  sub get_dir_globbed {
>      my ($self, $left, $depth, $r) = @_;

Same style comments as before, and a single empty line
in-between subs is enough.  Thanks.

      reply	other threads:[~2020-03-29 23:24 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-03-27  8:11 git-svn: Skip commit if all items of a commit are ignored by ignore configuration Lukas Pupka-Lipinski
2020-03-27  8:13 ` Lukas Pupka-Lipinski
2020-03-29 23:24   ` Eric Wong [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: http://vger.kernel.org/majordomo-info.html

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200329232414.GA27925@dcvr \
    --to=e@80x24.org \
    --cc=git@vger.kernel.org \
    --cc=lukas.pupkalipinski@lpl-mind.de \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://80x24.org/mirrors/git.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).