about summary refs log tree commit homepage
path: root/lib/PublicInbox/Filter.pm
diff options
context:
space:
mode:
Diffstat (limited to 'lib/PublicInbox/Filter.pm')
-rw-r--r--lib/PublicInbox/Filter.pm57
1 files changed, 28 insertions, 29 deletions
diff --git a/lib/PublicInbox/Filter.pm b/lib/PublicInbox/Filter.pm
index bd33130c..a1587682 100644
--- a/lib/PublicInbox/Filter.pm
+++ b/lib/PublicInbox/Filter.pm
@@ -21,9 +21,9 @@ our $MIME_TEXT_ANY = qr!\btext/[a-z0-9\+\._-]+\b!i;
 # this is highly opinionated delivery
 # returns 0 only if there is nothing to deliver
 sub run {
-        my ($class, $simple) = @_;
+        my ($class, $mime) = @_;
 
-        my $content_type = $simple->header("Content-Type") || "text/plain";
+        my $content_type = $mime->header('Content-Type') || 'text/plain';
 
         # kill potentially bad/confusing headers
         # Note: ssoma already does this, but since we mangle the message,
@@ -32,35 +32,35 @@ sub run {
         # the nature of public-inbox having no real subscribers.
         foreach my $d (qw(status lines content-length
                         mail-followup-to mail-reply-to reply-to)) {
-                $simple->header_set($d);
+                $mime->header_set($d);
         }
 
         if ($content_type =~ m!\btext/plain\b!i) {
                 return 1; # yay, nothing to do
         } elsif ($content_type =~ $MIME_HTML) {
                 # HTML-only, non-multipart
-                my $body = $simple->body;
+                my $body = $mime->body;
                 my $ct_parsed = parse_content_type($content_type);
                 dump_html(\$body, $ct_parsed->{attributes}->{charset});
-                replace_body($simple, $body);
+                replace_body($mime, $body);
                 return 1;
         } elsif ($content_type =~ m!\bmultipart/!i) {
-                return strip_multipart($simple, $content_type);
+                return strip_multipart($mime, $content_type);
         } else {
-                replace_body($simple, "$content_type message scrubbed");
+                replace_body($mime, "$content_type message scrubbed");
                 return 0;
         }
 }
 
 sub replace_part {
-        my ($simple, $part, $type) = ($_[0], $_[1], $_[3]);
+        my ($mime, $part, $type) = ($_[0], $_[1], $_[3]);
         # don't copy $_[2], that's the body (it may be huge)
 
         # Email::MIME insists on setting Date:, so just set it consistently
         # to avoid conflicts to avoid git merge conflicts in a split brain
         # situation.
         unless (defined $part->header('Date')) {
-                my $date = $simple->header('Date') ||
+                my $date = $mime->header('Date') ||
                            'Thu, 01 Jan 1970 00:00:00 +0000';
                 $part->header_set('Date', $date);
         }
@@ -77,11 +77,11 @@ sub replace_part {
 
 # converts one part of a multipart message to text
 sub html_part_to_text {
-        my ($simple, $part) = @_;
+        my ($mime, $part) = @_;
         my $body = $part->body;
         my $ct_parsed = parse_content_type($part->content_type);
         dump_html(\$body, $ct_parsed->{attributes}->{charset});
-        replace_part($simple, $part, $body, 'text/plain');
+        replace_part($mime, $part, $body, 'text/plain');
 }
 
 # modifies $_[0] in place
@@ -110,8 +110,7 @@ sub dump_html {
 # unfortunately, too many people send HTML mail and we'll attempt to convert
 # it to something safer, smaller and harder-to-track.
 sub strip_multipart {
-        my ($simple, $content_type) = @_;
-        my $mime = Email::MIME->new($simple->as_string);
+        my ($mime, $content_type) = @_;
 
         my (@html, @keep);
         my $rejected = 0;
@@ -164,11 +163,11 @@ sub strip_multipart {
 
         if ($content_type =~ m!\bmultipart/alternative\b!i) {
                 if (scalar @keep == 1) {
-                        return collapse($simple, $keep[0]);
+                        return collapse($mime, $keep[0]);
                 }
         } else { # convert HTML parts to plain text
                 foreach my $part (@html) {
-                        html_part_to_text($simple, $part);
+                        html_part_to_text($mime, $part);
                         push @keep, $part;
                 }
         }
@@ -186,34 +185,34 @@ sub strip_multipart {
         }
         if (scalar(@html) || $rejected) {
                 $mime->parts_set(\@keep);
-                $simple->body_set($mime->body_raw);
-                mark_changed($simple);
+                $mime->body_set($mime->body_raw);
+                mark_changed($mime);
         } # else: no changes
 
         return $ok;
 }
 
 sub mark_changed {
-        my ($simple) = @_;
-        $simple->header_set("X-Content-Filtered-By", __PACKAGE__ ." $VERSION");
+        my ($mime) = @_;
+        $mime->header_set('X-Content-Filtered-By', __PACKAGE__ ." $VERSION");
 }
 
 sub collapse {
-        my ($simple, $part) = @_;
-        $simple->header_set("Content-Type", $part->content_type);
-        $simple->body_set($part->body_raw);
-        mark_changed($simple);
+        my ($mime, $part) = @_;
+        $mime->header_set('Content-Type', $part->content_type);
+        $mime->body_set($part->body_raw);
+        mark_changed($mime);
         return 1;
 }
 
 sub replace_body {
-        my $simple = $_[0];
-        $simple->body_set($_[1]);
-        $simple->header_set("Content-Type", "text/plain");
-        if ($simple->header("Content-Transfer-Encoding")) {
-                $simple->header_set("Content-Transfer-Encoding", undef);
+        my $mime = $_[0];
+        $mime->body_set($_[1]);
+        $mime->header_set('Content-Type', 'text/plain');
+        if ($mime->header('Content-Transfer-Encoding')) {
+                $mime->header_set('Content-Transfer-Encoding', undef);
         }
-        mark_changed($simple);
+        mark_changed($mime);
 }
 
 # Check for display-able text, no messed up binaries