On Sat, Feb 09 2013, david at tethera.net wrote:
> From: David Bremner <bremner at debian.org>
>
> In the case of large changes to the database from git, one of main
> current bottlenecks is the large number of execs of notmuch tag. This
> avoids that by using use the batch tagging facilities as of notmuch
> 0.15.
>
> We use "spawn" directly rather than inventing a "notmuch_pipe", since
> it seems the only place we need to pipe to notmuch so far.
> ---
The code LGTM. Is there any quoting differences between these 2
tagging methods ?
> This is only lightly tested; please make sure you have backups of your
> database.
Do I dare to test -- I'm lazy taking backups (I don't know how to take
one -- and especially how to restore one ;)
> I'm not 100% sure about the indentation. No doubt Tomi will let me
> know if it can be improved.
No doubt...
Tomi
>
> contrib/nmbug/nmbug | 15 ++++++++++-----
> 1 file changed, 10 insertions(+), 5 deletions(-)
>
> diff --git a/contrib/nmbug/nmbug b/contrib/nmbug/nmbug
> index f003ef9..c66c526 100755
> --- a/contrib/nmbug/nmbug
> +++ b/contrib/nmbug/nmbug
> @@ -270,17 +270,22 @@ sub do_sync {
> $D_action = '-';
> }
>
> - foreach my $pair (@{$status->{added}}) {
> + my $notmuch = spawn ({}, '|-', qw/notmuch tag --batch/)
> + or die 'notmuch tag --batch';
>
> - notmuch ('tag', $A_action.$TAGPREFIX.$pair->{tag},
> - 'id:'.$pair->{id});
> + foreach my $pair (@{$status->{added}}) {
> + print $notmuch $A_action.$TAGPREFIX.$pair->{tag}, " -- ",
> + 'id:'.$pair->{id};
> }
>
> foreach my $pair (@{$status->{deleted}}) {
> - notmuch ('tag', $D_action.$TAGPREFIX.$pair->{tag},
> - 'id:'.$pair->{id});
> + print $notmuch $D_action.$TAGPREFIX.$pair->{tag},
> + 'id:'.$pair->{id};
> }
>
> + unless (close $notmuch) {
> + die "'notmuch tag --batch' exited with nonzero value\n";
> + }
> }
>
>
> --
> 1.7.10.4
>
> _______________________________________________
> notmuch mailing list
> notmuch at notmuchmail.org
> http://notmuchmail.org/mailman/listinfo/notmuch
