No ext_urls

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

diff --git a/lib/PublicInbox/Smsg.pm b/lib/PublicInbox/Smsg.pm
index f22cd43e7d4b69ce97eaf248546dd1598715ebd4..b132381b4ab2c3a93fd78b1fe042d1710a2c3fe8 100644 (file)
--- a/lib/PublicInbox/Smsg.pm
+++ b/lib/PublicInbox/Smsg.pm
@@ -1,4 +1,4 @@
-# Copyright (C) 2015-2020 all contributors <meta@public-inbox.org>
+# Copyright (C) 2015-2021 all contributors <meta@public-inbox.org>
 # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
 #
 # A small/skeleton/slim representation of a message.
@@ -9,19 +9,14 @@
 # large threads in our WWW UI and the NNTP range responses.
 package PublicInbox::Smsg;
 use strict;
-use warnings;
-use base qw(Exporter);
+use v5.10.1;
+use parent qw(Exporter);
 our @EXPORT_OK = qw(subject_normalized);
-use PublicInbox::MID qw(mids);
+use PublicInbox::MID qw(mids references);
 use PublicInbox::Address;
 use PublicInbox::MsgTime qw(msg_timestamp msg_datestamp);
-use Time::Local qw(timegm);
 
-sub get_val ($$) {
-       my ($doc, $col) = @_;
-       # sortable_unserialise is defined by PublicInbox::Search::load_xapian()
-       sortable_unserialise($doc->get_value($col));
-}
+sub oidbin { pack('H*', $_[0]->{blob}) }
 
 sub to_doc_data {
        my ($self) = @_;
@@ -61,40 +56,39 @@ sub load_from_data ($$) {
        ) = split(/\n/, $_[1]);
 }
 
-sub load_expand {
-       my ($self, $doc) = @_;
-       my $data = $doc->get_data or return;
-       $self->{ts} = get_val($doc, PublicInbox::Search::TS());
-       my $dt = get_val($doc, PublicInbox::Search::DT());
-       my ($yyyy, $mon, $dd, $hh, $mm, $ss) = unpack('A4A2A2A2A2A2', $dt);
-       $self->{ds} = timegm($ss, $mm, $hh, $dd, $mon - 1, $yyyy);
-       load_from_data($self, $data);
-       $self;
-}
-
 sub psgi_cull ($) {
        my ($self) = @_;
 
-       # ghosts don't have ->{from}
-       my $from = delete($self->{from}) // '';
-       my @n = PublicInbox::Address::names($from);
-       $self->{from_name} = join(', ', @n);
-
        # drop NNTP-only fields which aren't relevant to PSGI results:
        # saves ~80K on a 200 item search result:
-       delete @$self{qw(ts to cc bytes lines)};
+       # TODO: we may need to keep some of these for JMAP...
+       my ($f) = delete @$self{qw(from tid to cc bytes lines)};
+       # ghosts don't have ->{from}
+       $self->{from_name} = join(', ', PublicInbox::Address::names($f // ''));
        $self;
 }
 
-# Only called by PSGI interface, not NNTP
-sub from_mitem {
-       my ($mitem, $srch) = @_;
-       return $srch->retry_reopen(\&from_mitem, $mitem) if $srch;
-       my $self = bless {}, __PACKAGE__;
-       psgi_cull(load_expand($self, $mitem->get_document));
+sub parse_references ($$$) {
+       my ($smsg, $hdr, $mids) = @_;
+       my $refs = references($hdr);
+       push(@$refs, @$mids) if scalar(@$mids) > 1;
+       return $refs if scalar(@$refs) == 0;
+
+       # prevent circular references here:
+       my %seen = ( ($smsg->{mid} // '') => 1 );
+       my @keep;
+       foreach my $ref (@$refs) {
+               if (length($ref) > PublicInbox::MID::MAX_MID_SIZE) {
+                       warn "References: <$ref> too long, ignoring\n";
+                       next;
+               }
+               $seen{$ref} //= push(@keep, $ref);
+       }
+       $smsg->{references} = '<'.join('> <', @keep).'>' if @keep;
+       \@keep;
 }
 
-# for Import and v1 non-SQLite WWW code paths
+# used for v2, Import and v1 non-SQLite WWW code paths
 sub populate {
        my ($self, $hdr, $sync) = @_;
        for my $f (qw(From To Cc Subject)) {
@@ -105,9 +99,6 @@ sub populate {
                # to protect git and NNTP clients
                $val =~ tr/\0\t\n/   /;
 
-               # rare: in case headers have wide chars (not RFC2047-encoded)
-               utf8::decode($val);
-
                # lower-case fields for read-only stuff
                $self->{lc($f)} = $val;
 
@@ -121,13 +112,13 @@ sub populate {
                $self->{$f} = $val if $val ne '';
        }
        $sync //= {};
-       $self->{-ds} = [ my @ds = msg_datestamp($hdr, $sync->{autime}) ];
-       $self->{-ts} = [ my @ts = msg_timestamp($hdr, $sync->{cotime}) ];
+       my @ds = msg_datestamp($hdr, $sync->{autime} // $self->{ds});
+       my @ts = msg_timestamp($hdr, $sync->{cotime} // $self->{ts});
+       $self->{-ds} = \@ds;
+       $self->{-ts} = \@ts;
        $self->{ds} //= $ds[0]; # no zone
        $self->{ts} //= $ts[0];
-
-       # for v1 users w/o SQLite
-       $self->{mid} //= eval { mids($hdr)->[0] } // '';
+       $self->{mid} //= mids($hdr)->[0];
 }
 
 # no strftime, that is locale-dependent and not for RFC822
@@ -152,6 +143,8 @@ sub internaldate { # for IMAP
 
 our $REPLY_RE = qr/^re:\s+/i;
 
+# TODO: see RFC 5256 sec 2.1 "Base Subject" and evaluate compatibility
+# w/ existing indices...
 sub subject_normalized ($) {
        my ($subj) = @_;
        $subj =~ s/\A\s+//s; # no leading space
@@ -162,4 +155,17 @@ sub subject_normalized ($) {
        $subj;
 }
 
+# returns the number of bytes to add if given a non-CRLF arg
+sub crlf_adjust ($) {
+       if (index($_[0], "\r\n") < 0) {
+               # common case is LF-only, every \n needs an \r;
+               # so favor a cheap tr// over an expensive m//g
+               $_[0] =~ tr/\n/\n/;
+       } else { # count number of '\n' w/o '\r', expensive:
+               scalar(my @n = ($_[0] =~ m/(?<!\r)\n/g));
+       }
+}
+
+sub set_bytes { $_[0]->{bytes} = $_[2] + crlf_adjust($_[1]) }
+
 1;