diff options
Diffstat (limited to 'scripts/import_gmane_spool')
-rwxr-xr-x | scripts/import_gmane_spool | 51 |
1 files changed, 51 insertions, 0 deletions
diff --git a/scripts/import_gmane_spool b/scripts/import_gmane_spool new file mode 100755 index 00000000..b5573e15 --- /dev/null +++ b/scripts/import_gmane_spool @@ -0,0 +1,51 @@ +#!/usr/bin/perl -w +# Copyright (C) 2013, Eric Wong <normalperson@yhbt.net> and all contributors +# License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt) +# +# One-off script to convert an slrnpull news spool from gmane +use strict; +use warnings; +use Parallel::ForkManager; +use Email::Simple; +use PublicInbox::Filter; +use IPC::Run qw(run); +my $usage = "import_nntp_spool SLRNPULL_ROOT/news/foo/bar MAIN_REPO FAIL_REPO"; +my $spool = shift @ARGV or die "Usage: $usage\n"; +my $main_repo = shift @ARGV or die "Usage: $usage\n"; +my $fail_repo = shift @ARGV or die "Usage: $usage\n"; +my $nproc = `nproc 2>/dev/null` || 4; +my $pm = Parallel::ForkManager->new($nproc); +my @args = ('public-inbox-mda', $main_repo, $fail_repo); + +foreach my $n (<$spool/*>) { + $n =~ m{/\d+\z} or next; + $pm->start and next; + if (open my $fh, '<', $n) { + local $/; + my $s = Email::Simple->new(<$fh>); + + # gmane rewrites Received headers, which increases spamminess + my @h = $s->header("Original-Received"); + if (@h) { + $s->header_set("Received", @h); + $s->header_set("Original-Received"); + } + + # triggers for the SA HEADER_SPAM rule + foreach my $drop (qw(Approved)) { $s->header_set($drop) } + + # appears to be an old gmane bug: + $s->header_set("connect()"); + + my $orig = $s->as_string; + close $fh or die "close failed: $!\n"; + eval { run(\@args, \$orig) }; + die "fail $n: $?\n" if $?; + die "fail $n: $@\n" if $@; + } else { + warn "Failed to open $n: $!\n"; + } + $pm->finish; +} + +$pm->wait_all_children; |