X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=scripts%2Fimport_slrnspool;h=d9a35dfdfa86b817018540bf3450ffbbbf289f94;hb=af0b0fb7a454470a32c452119d0392e0dedb3fe1;hp=ea7a97bf69689bdf85950264d9919cd1fccd5b17;hpb=4438f547ecbb52d900dcc73fe4c03cb2e7d5562c;p=public-inbox.git

diff --git a/scripts/import_slrnspool b/scripts/import_slrnspool
index ea7a97bf..d9a35dfd 100755
--- a/scripts/import_slrnspool
+++ b/scripts/import_slrnspool
@@ -1,72 +1,81 @@
 #!/usr/bin/perl -w
-# Copyright (C) 2015, all contributors <meta@public-inbox.org>
-# License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt)
+# Copyright (C) 2015-2021 all contributors <meta@public-inbox.org>
+# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
 #
 # Incremental (or one-shot) importer of a slrnpull news spool
 =begin usage
 	export ORIGINAL_RECIPIENT=address@example.com
-	public-inbox-init $LISTNAME $GIT_DIR $HTTP_URL $ORIGINAL_RECIPIENT
+	public-inbox-init $INBOX $GIT_DIR $HTTP_URL $ORIGINAL_RECIPIENT
 	./import_slrnspool SLRNPULL_ROOT/news/foo/bar
 =cut
 use strict;
 use warnings;
 use PublicInbox::Config;
-use Email::Filter;
-use Email::LocalDelivery;
+use PublicInbox::Eml;
+use PublicInbox::Import;
+use PublicInbox::Git;
 sub usage { "Usage:\n".join('',grep(/\t/, `head -n 10 $0`)) }
+my $exit = 0;
+my $sighandler = sub { $exit = 1 };
+$SIG{INT} = $sighandler;
+$SIG{TERM} = $sighandler;
 my $spool = shift @ARGV or die usage();
 my $recipient = $ENV{ORIGINAL_RECIPIENT};
 defined $recipient or die usage();
-my @mda = qw(public-inbox-mda);
-my $config = PublicInbox::Config->new;
-my $cfg = $config->lookup($recipient);
-defined $cfg or exit(1);
-use Data::Dumper; print STDERR Dumper($cfg);
+my $cfg = PublicInbox::Config->new;
+my $ibx = $cfg->lookup($recipient);
+my $git = $ibx->git;
+my $im;
+if ($ibx->version == 2) {
+	require PublicInbox::V2Writable;
+	$im = PublicInbox::V2Writable->new($ibx);
+	$im->{parallel} = 0; # pointless to be parallel for a single message
+} else {
+	$im = PublicInbox::Import->new($git, $ibx->{name},
+					$ibx->{-primary_address});
+}
+
+$ibx->{filter} ||= 'PublicInbox::Filter::Gmane';
+my $filter = $ibx->filter;
+
+sub key {
+	"publicinbox.$ibx->{name}.importslrnspoolstate";
+}
 
 sub get_min {
-	my ($cfg) = @_;
-	$cfg->{importslrnspoolstate} || 0;
+	my $f = PublicInbox::Config->default_file;
+	my $out = $git->qx('config', "--file=$f", key($ibx));
+	$out ||= 0;
+	chomp $out;
+	$out =~ /\A[0-9]+\z/ and return $out;
+	0;
 }
 
 sub set_min {
-	my ($cfg, $num) = @_;
+	my ($num) = @_;
 	my $f = PublicInbox::Config->default_file;
-	my @cmd = (qw/git config/, "--file=$f",
-		   "publicinbox.$cfg->{listname}.importslrnspoolstate", $num);
+	my @cmd = (qw/git config/, "--file=$f", key($ibx), $num);
 	system(@cmd) == 0 or die join(' ', @cmd). " failed: $?\n";
 }
 
 my $n = get_min();
 my $ok;
-my $max_gap = 10000;
+my $max_gap = 200000;
 my $max = $n + $max_gap;
+$spool =~ s!/+\z!!;
 
-for (; $n < $max; $n++) {
+for (; $exit == 0 && $n < $max; $n++) {
 	my $fn = "$spool/$n";
-	print STDERR $fn, "\n";
 	open(my $fh, '<', $fn) or next;
 	$max = $n + $max_gap;
-	my $f = Email::Filter->new(data => eval { local $/; <$fh> });
-	my $s = $f->simple;
-
-	# gmane rewrites Received headers, which increases spamminess
-	# Some older archives set Original-To
-	foreach my $x (qw(Received To)) {
-		my @h = $s->header("Original-$x");
-		if (@h) {
-			$s->header_set($x, @h);
-			$s->header_set("Original-$x");
-		}
-	}
-
-	# triggers for the SA HEADER_SPAM rule
-	foreach my $drop (qw(Approved)) { $s->header_set($drop) }
+	print STDERR $fn, "\n";
 
-	# appears to be an old gmane bug:
-	$s->header_set('connect()');
+	my $mime = PublicInbox::Eml->new(do { local $/; <$fh> });
+	$filter->scrub($mime);
+	$im->add($mime);
 
-	$f->exit(0);
-	$f->pipe(@mda);
 	$ok = $n + 1;
-	set_min($cfg, $ok);
+	set_min($ok);
 }
+
+$im->done;