X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FMID.pm;h=5aeffb8c675bb2882a1f50b92505206e386191e9;hb=c6ca576baf1700a8341ae7b6baa19257ac7deaa7;hp=c82e84013ec0d68730423ab4b691074c5b690dc5;hpb=cfb8d16578e7f2f2e300f9f436205e4a8fc7f322;p=public-inbox.git

diff --git a/lib/PublicInbox/MID.pm b/lib/PublicInbox/MID.pm
index c82e8401..5aeffb8c 100644
--- a/lib/PublicInbox/MID.pm
+++ b/lib/PublicInbox/MID.pm
@@ -1,4 +1,4 @@
-# Copyright (C) 2015-2018 all contributors <meta@public-inbox.org>
+# Copyright (C) 2015-2020 all contributors <meta@public-inbox.org>
 # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
 #
 # Various Message-ID-related functions.
@@ -6,30 +6,33 @@ package PublicInbox::MID;
 use strict;
 use warnings;
 use base qw/Exporter/;
-our @EXPORT_OK = qw/mid_clean id_compress mid2path mid_mime mid_escape MID_ESC
-	mids references/;
+our @EXPORT_OK = qw(mid_clean id_compress mid2path mid_escape MID_ESC
+	mids references mids_for_index $MID_EXTRACT);
 use URI::Escape qw(uri_escape_utf8);
 use Digest::SHA qw/sha1_hex/;
+require PublicInbox::Address;
 use constant {
-	MID_MAX => 40, # SHA-1 hex length # TODO: get rid of this
+	ID_MAX => 40, # SHA-1 hex length for HTML id anchors
 	MAX_MID_SIZE => 244, # max term size (Xapian limitation) - length('Q')
 };
 
+our $MID_EXTRACT = qr/<([^>]+)>/s;
+
 sub mid_clean {
 	my ($mid) = @_;
 	defined($mid) or die "no Message-ID";
 	# MDA->precheck did more checking for us
-	if ($mid =~ /<([^>]+)>/) {
+	if ($mid =~ $MID_EXTRACT) {
 		$mid = $1;
 	}
 	$mid;
 }
 
-# this is idempotent
+# this is idempotent, used for HTML anchor/ids and such
 sub id_compress {
 	my ($id, $force) = @_;
 
-	if ($force || $id =~ /[^\w\-]/ || length($id) > MID_MAX) {
+	if ($force || $id =~ /[^a-zA-Z0-9_\-]/ || length($id) > ID_MAX) {
 		utf8::encode($id);
 		return sha1_hex($id);
 	}
@@ -50,22 +53,34 @@ sub mid2path {
 	"$x2/$x38";
 }
 
-# Only for v1 code paths:
-sub mid_mime ($) { mids($_[0]->header_obj)->[0] }
-
-sub mids ($) {
-	my ($hdr) = @_;
+# only intended for Message-ID and X-Alt-Message-ID
+sub extract_mids {
 	my @mids;
-	my @v = $hdr->header_raw('Message-Id');
-	foreach my $v (@v) {
-		my @cur = ($v =~ /<([^>]+)>/sg);
+	for my $v (@_) {
+		my @cur = ($v =~ /$MID_EXTRACT/g);
 		if (@cur) {
 			push(@mids, @cur);
 		} else {
 			push(@mids, $v);
 		}
 	}
-	uniq_mids(\@mids);
+	\@mids;
+}
+
+sub mids ($) {
+	my ($hdr) = @_;
+	my @mids = $hdr->header_raw('Message-ID');
+	uniq_mids(extract_mids(@mids));
+}
+
+# we allow searching on X-Alt-Message-ID since PublicInbox::NNTP uses them
+# to placate some clients, and we want to ensure NNTP-only clients can
+# import and index without relying on HTTP endpoints
+sub mids_for_index ($) {
+	my ($hdr) = @_;
+	my @mids = $hdr->header_raw('Message-ID');
+	my @alts = $hdr->header_raw('X-Alt-Message-ID');
+	uniq_mids(extract_mids(@mids, @alts));
 }
 
 # last References should be IRT, but some mail clients do things
@@ -76,30 +91,41 @@ sub references ($) {
 	foreach my $f (qw(References In-Reply-To)) {
 		my @v = $hdr->header_raw($f);
 		foreach my $v (@v) {
-			push(@mids, ($v =~ /<([^>]+)>/sg));
+			push(@mids, ($v =~ /$MID_EXTRACT/g));
+		}
+	}
+
+	# old versions of git-send-email would prompt users for
+	# In-Reply-To and users' muscle memory would use 'y' or 'n'
+	# as responses:
+	my %addr = ( y => 1, n => 1 );
+
+	foreach my $f (qw(To From Cc)) {
+		my @v = $hdr->header_raw($f);
+		foreach my $v (@v) {
+			$addr{$_} = 1 for (PublicInbox::Address::emails($v));
 		}
 	}
-	uniq_mids(\@mids);
+	uniq_mids(\@mids, \%addr);
 }
 
-sub uniq_mids ($) {
-	my ($mids) = @_;
+sub uniq_mids ($;$) {
+	my ($mids, $seen) = @_;
 	my @ret;
-	my %seen;
+	$seen ||= {};
 	foreach my $mid (@$mids) {
+		$mid =~ tr/\n\t\r//d;
 		if (length($mid) > MAX_MID_SIZE) {
 			warn "Message-ID: <$mid> too long, truncating\n";
 			$mid = substr($mid, 0, MAX_MID_SIZE);
 		}
-		next if $seen{$mid};
-		push @ret, $mid;
-		$seen{$mid} = 1;
+		push(@ret, $mid) unless $seen->{$mid}++;
 	}
 	\@ret;
 }
 
 # RFC3986, section 3.3:
-sub MID_ESC () { '^A-Za-z0-9\-\._~!\$\&\';\(\)\*\+,;=:@' }
+sub MID_ESC () { '^A-Za-z0-9\-\._~!\$\&\'\(\)\*\+,;=:@' }
 sub mid_escape ($) { uri_escape_utf8($_[0], MID_ESC) }
 
 1;