inbox: add ->version method

[public-inbox.git] / scripts / import_slrnspool

diff --git a/scripts/import_slrnspool b/scripts/import_slrnspool
index 5158460bfa1bfdba325a90c28dd1dfd2e654e92c..b913cf3212d17de382ce91ba761b0c9025a58529 100755 (executable)
--- a/scripts/import_slrnspool
+++ b/scripts/import_slrnspool
@@ -1,5 +1,5 @@
 #!/usr/bin/perl -w
-# Copyright (C) 2015-2018 all contributors <meta@public-inbox.org>
+# Copyright (C) 2015-2019 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
@@ -11,7 +11,7 @@
 use strict;
 use warnings;
 use PublicInbox::Config;
-use Email::MIME;
+use PublicInbox::MIME;
 use PublicInbox::Import;
 use PublicInbox::Git;
 sub usage { "Usage:\n".join('',grep(/\t/, `head -n 10 $0`)) }
@@ -25,7 +25,18 @@ defined $recipient or die usage();
 my $config = PublicInbox::Config->new;
 my $ibx = $config->lookup($recipient);
 my $git = $ibx->git;
-my $im = PublicInbox::Import->new($git, $ibx->{name}, $ibx->{-primary_address});
+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";
@@ -36,7 +47,7 @@ sub get_min {
        my $out = $git->qx('config', "--file=$f", key($ibx));
        $out ||= 0;
        chomp $out;
-       $out =~ /\A\d+\z/ and return $out;
+       $out =~ /\A[0-9]+\z/ and return $out;
        0;
 }
 
@@ -51,34 +62,16 @@ my $n = get_min();
 my $ok;
 my $max_gap = 200000;
 my $max = $n + $max_gap;
+$spool =~ s!/+\z!!;
 
 for (; $exit == 0 && $n < $max; $n++) {
        my $fn = "$spool/$n";
-       print STDERR $fn, "\n";
        open(my $fh, '<', $fn) or next;
        $max = $n + $max_gap;
+       print STDERR $fn, "\n";
 
-       my $mime = Email::MIME->new(eval { local $/; <$fh> });
-       my $hdr = $mime->header_obj;
-
-       # gmane rewrites Received headers, which increases spamminess
-       # Some older archives set Original-To
-       foreach my $x (qw(Received To)) {
-               my @h = $hdr->header_raw("Original-$x");
-               if (@h) {
-                       $hdr->header_set($x, @h);
-                       $hdr->header_set("Original-$x");
-               }
-       }
-
-       # Approved triggers for the SA HEADER_SPAM rule,
-       # X-From is gmane specific
-       foreach my $drop (qw(Approved X-From)) {
-               $hdr->header_set($drop);
-       }
-
-       # appears to be an old gmane bug:
-       $hdr->header_set('connect()');
+       my $mime = PublicInbox::MIME->new(eval { local $/; <$fh> });
+       $filter->scrub($mime);
        $im->add($mime);
 
        $ok = $n + 1;