watchmaildir: do not reject lowercase flags on Maildir files

[public-inbox.git] / lib / PublicInbox / WatchMaildir.pm

diff --git a/lib/PublicInbox/WatchMaildir.pm b/lib/PublicInbox/WatchMaildir.pm
index b25704e628bad30be39b613e4dd9222d3beccb51..985f9192078f3497de62e6da5691f706579ef90a 100644 (file)
--- a/lib/PublicInbox/WatchMaildir.pm
+++ b/lib/PublicInbox/WatchMaildir.pm
@@ -6,7 +6,7 @@
 package PublicInbox::WatchMaildir;
 use strict;
 use warnings;
-use Email::MIME;
+use PublicInbox::MIME;
 use Email::MIME::ContentType;
 $Email::MIME::ContentType::STRICT_PARAMS = 0; # user input is imperfect
 use PublicInbox::Git;
@@ -18,21 +18,25 @@ sub new {
        my ($class, $config) = @_;
        my (%mdmap, @mdir, $spamc);
 
-       # XXX is "publicinboxlearn" really a good namespace for this?
-       my $k = 'publicinboxlearn.watchspam';
-       if (my $spamdir = $config->{$k}) {
-               if ($spamdir =~ s/\Amaildir://) {
-                       $spamdir =~ s!/+\z!!;
-                       # skip "new", no MUA has seen it, yet.
-                       my $cur = "$spamdir/cur";
-                       push @mdir, $cur;
-                       $mdmap{$cur} = 'watchspam';
-               } else {
-                       warn "unsupported $k=$spamdir\n";
+       # "publicinboxwatch" is the documented namespace
+       # "publicinboxlearn" is legacy but may be supported
+       # indefinitely...
+       foreach my $pfx (qw(publicinboxwatch publicinboxlearn)) {
+               my $k = "$pfx.watchspam";
+               if (my $spamdir = $config->{$k}) {
+                       if ($spamdir =~ s/\Amaildir://) {
+                               $spamdir =~ s!/+\z!!;
+                               # skip "new", no MUA has seen it, yet.
+                               my $cur = "$spamdir/cur";
+                               push @mdir, $cur;
+                               $mdmap{$cur} = 'watchspam';
+                       } else {
+                               warn "unsupported $k=$spamdir\n";
+                       }
                }
        }
 
-       $k = 'publicinboxwatch.spamcheck';
+       my $k = 'publicinboxwatch.spamcheck';
        my $spamcheck = $config->{$k};
        if ($spamcheck) {
                if ($spamcheck eq 'spamc') {
@@ -76,6 +80,7 @@ sub new {
                mdmap => \%mdmap,
                mdir => \@mdir,
                mdre => $mdre,
+               config => $config,
                importers => {},
        }, $class;
 }
@@ -92,18 +97,22 @@ sub _try_fsn_paths {
 
 sub _remove_spam {
        my ($self, $path) = @_;
-       $path =~ /:2,[A-R]*S[T-Z]*\z/i or return;
+       # path must be marked as (S)een
+       $path =~ /:2,[A-R]*S[T-Za-z]*\z/ or return;
        my $mime = _path_to_mime($path) or return;
        _force_mid($mime);
-       foreach my $inbox (values %{$self->{mdmap}}) {
-               next unless ref $inbox;
-               my $im = _importer_for($self, $inbox);
-               $im->remove($mime);
-               if (my $scrub = _scrubber_for($inbox)) {
-                       my $scrubbed = $scrub->scrub($mime) or next;
-                       $im->remove($scrubbed);
-               }
-       }
+       $self->{config}->each_inbox(sub {
+               my ($ibx) = @_;
+               eval {
+                       my $im = _importer_for($self, $ibx);
+                       $im->remove($mime);
+                       if (my $scrub = _scrubber_for($ibx)) {
+                               my $scrubbed = $scrub->scrub($mime) or return;
+                               $im->remove($scrubbed);
+                       }
+               };
+               warn "error removing spam at $path from $ibx->{name}\n" if $@;
+       })
 }
 
 # used to hash the relevant portions of a message when there are conflicts
@@ -199,7 +208,7 @@ sub _path_to_mime {
                local $/;
                my $str = <$fh>;
                $str or return;
-               return Email::MIME->new(\$str);
+               return PublicInbox::MIME->new(\$str);
        } elsif ($!{ENOENT}) {
                return;
        } else {
@@ -214,20 +223,33 @@ sub _importer_for {
                my $git = $inbox->git;
                my $name = $inbox->{name};
                my $addr = $inbox->{-primary_address};
-               PublicInbox::Import->new($git, $name, $addr);
+               PublicInbox::Import->new($git, $name, $addr, $inbox);
        };
-       $self->{importers}->{"$im"} = $im;
+
+       my $importers = $self->{importers};
+       if (scalar(keys(%$importers)) > 2) {
+               delete $importers->{"$im"};
+               _done_for_now($self);
+       }
+
+       $importers->{"$im"} = $im;
 }
 
 sub _scrubber_for {
        my ($inbox) = @_;
        my $f = $inbox->{filter};
        if ($f && $f =~ /::/) {
+               my @args;
+               # basic line splitting, only
+               # Perhaps we can have proper quote splitting one day...
+               ($f, @args) = split(/\s+/, $f) if $f =~ /\s+/;
+
                eval "require $f";
                if ($@) {
                        warn $@;
                } else {
-                       return $f->new;
+                       # e.g: PublicInbox::Filter::Vger->new(@args)
+                       return $f->new(@args);
                }
        }
        undef;
@@ -239,8 +261,9 @@ sub _spamcheck_cb {
                my ($mime) = @_;
                my $tmp = '';
                if ($sc->spamcheck($mime, \$tmp)) {
-                       return Email::MIME->new(\$tmp);
+                       return PublicInbox::MIME->new(\$tmp);
                }
+               warn $mime->header('Message-ID')." failed spam check\n";
                undef;
        }
 }