-# 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>
# based on notmuch, but with no concept of folders, files or flags
#
# Wraps a document inside our Xapian search index.
+# There may be many of these objects loaded in memory at once
+# for large threads in our WWW UI.
package PublicInbox::SearchMsg;
use strict;
use warnings;
-use Search::Xapian;
+use base qw(Exporter);
+our @EXPORT_OK = qw(subject_normalized);
use PublicInbox::MID qw/mid_clean mid_mime/;
use PublicInbox::Address;
-use PublicInbox::MsgTime qw(msg_timestamp);
+use PublicInbox::MsgTime qw(msg_timestamp msg_datestamp);
+use Time::Local qw(timegm);
sub new {
my ($class, $mime) = @_;
- my $doc = Search::Xapian::Document->new;
- $doc->add_boolean_term('T' . 'mail');
-
- bless { type => 'mail', doc => $doc, mime => $mime }, $class;
+ bless { mime => $mime }, $class;
}
sub wrap {
- my ($class, $doc, $mid) = @_;
- bless { doc => $doc, mime => undef, mid => $mid }, $class;
+ my ($class, $mid) = @_;
+ bless { mid => $mid }, $class;
}
sub get_val ($$) {
my ($doc, $col) = @_;
- Search::Xapian::sortable_unserialise($doc->get_value($col));
+ # sortable_unserialise is defined by PublicInbox::Search::load_xapian()
+ sortable_unserialise($doc->get_value($col));
+}
+
+sub to_doc_data {
+ my ($self, $oid, $mid0) = @_;
+ $oid = '' unless defined $oid;
+ join("\n",
+ $self->subject,
+ $self->from,
+ $self->references,
+ $self->to,
+ $self->cc,
+ $oid,
+ $mid0,
+ $self->{bytes} || '',
+ $self->{lines} || ''
+ );
}
sub load_from_data ($$) {
my ($self) = $_[0]; # data = $_[1]
- my ($subj, $from, $refs, $to, $cc, $blob, $mid0) = split(/\n/, $_[1]);
- $self->{subject} = $subj;
- $self->{from} = $from;
- $self->{references} = $refs;
- $self->{to} = $to;
- $self->{cc} = $cc;
- $self->{blob} = $blob;
- $self->{mid} = $mid0;
+ (
+ $self->{subject},
+ $self->{from},
+ $self->{references},
+
+ # To: and Cc: are stored to optimize HDR/XHDR in NNTP since
+ # some NNTP clients will use that for message displays.
+ # NNTP only, and only stored in Over(view), not Xapian
+ $self->{to},
+ $self->{cc},
+
+ $self->{blob},
+ $self->{mid},
+
+ # NNTP only
+ $self->{bytes},
+ $self->{lines}
+ ) = split(/\n/, $_[1]);
}
sub load_expand {
- my ($self) = @_;
- my $doc = $self->{doc};
+ my ($self, $doc) = @_;
my $data = $doc->get_data or return;
- $self->{ts} = get_val($doc, &PublicInbox::Search::TS);
+ $self->{ts} = get_val($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, $yyyy);
utf8::decode($data);
load_from_data($self, $data);
$self;
}
-sub load_doc {
- my ($class, $doc) = @_;
- my $data = $doc->get_data or return;
- my $ts = get_val($doc, &PublicInbox::Search::TS);
- utf8::decode($data);
- my $self = bless { doc => $doc, ts => $ts }, $class;
- load_from_data($self, $data);
- $self
+sub psgi_cull ($) {
+ my ($self) = @_;
+ from_name($self); # fill in {from_name} so we can delete {from}
+
+ # drop NNTP-only fields which aren't relevant to PSGI results:
+ # saves ~80K on a 200 item search result:
+ delete @$self{qw(from ts to cc bytes lines)};
+ $self;
}
-# :bytes and :lines metadata in RFC 3977
-sub bytes ($) { get_val($_[0]->{doc}, &PublicInbox::Search::BYTES) }
-sub lines ($) { get_val($_[0]->{doc}, &PublicInbox::Search::LINES) }
-sub num ($) {
- $_[0]->{num} ||= get_val($_[0]->{doc}, PublicInbox::Search::NUM)
+# 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, $mitem->get_document));
}
+# :bytes and :lines metadata in RFC 3977
+sub bytes ($) { $_[0]->{bytes} }
+sub lines ($) { $_[0]->{lines} }
+
sub __hdr ($$) {
my ($self, $field) = @_;
my $val = $self->{$field};
return $val if defined $val;
my $mime = $self->{mime} or return;
- $val = $mime->header($field);
- $val = '' unless defined $val;
- $val =~ tr/\n/ /;
+ my @raw = $mime->header($field);
+ $val = join(', ', @raw);
+ $val =~ tr/\t\n/ /;
$val =~ tr/\r//d;
$self->{$field} = $val;
}
sub date ($) {
my ($self) = @_;
- my $ts = $self->{ts};
- return unless defined $ts;
- my ($sec, $min, $hour, $mday, $mon, $year, $wday) = gmtime($ts);
+ my $ds = $self->{ds};
+ return unless defined $ds;
+ my ($sec, $min, $hour, $mday, $mon, $year, $wday) = gmtime($ds);
"$DoW[$wday], " . sprintf("%02d $MoY[$mon] %04d %02d:%02d:%02d +0000",
$mday, $year+1900, $hour, $min, $sec);
sub ts {
my ($self) = @_;
- $self->{ts} ||= eval {
- msg_timestamp($self->{mime}->header_obj);
- } || 0;
+ $self->{ts} ||= eval { msg_timestamp($self->{mime}->header_obj) } || 0;
}
-sub to_doc_data {
- my ($self, $oid, $mid0) = @_;
- my @rows = ($self->subject, $self->from, $self->references,
- $self->to, $self->cc);
- $oid = '' unless defined $oid;
- join("\n", @rows, $oid, $mid0);
+sub ds {
+ my ($self) = @_;
+ $self->{ds} ||= eval { msg_datestamp($self->{mime}->header_obj); } || 0;
}
sub references {
defined $x ? $x : '';
}
-sub _get_term_val ($$$) {
- my ($self, $pfx, $re) = @_;
- my $doc = $self->{doc};
- my $end = $doc->termlist_end;
- my $i = $doc->termlist_begin;
- $i->skip_to($pfx);
- if ($i != $end) {
- my $val = $i->get_termname;
- $val =~ s/$re// and return $val;
- }
- undef;
-}
-
sub mid ($;$) {
my ($self, $mid) = @_;
if (defined $mid) {
$self->{mid} = $mid;
- } elsif (my $rv = $self->{mid}) {
+ } elsif (defined(my $rv = $self->{mid})) {
$rv;
} else {
- $self->{mid} = _get_term_val($self, 'Q', qr/\AQ/) ||
- $self->_extract_mid;
+ die "NO {mime} for mid\n" unless $self->{mime};
+ $self->_extract_mid; # v1 w/o Xapian
}
}
sub _extract_mid { mid_clean(mid_mime($_[0]->{mime})) }
-sub thread_id {
- my ($self) = @_;
- my $tid = $self->{thread};
- return $tid if defined $tid;
- $self->{thread} = _get_term_val($self, 'G', qr/\AG/); # *G*roup
-}
-
-# XXX: consider removing this, we can phrase match subject
-sub path {
- my ($self) = @_;
- my $path = $self->{path};
- return $path if defined $path;
- $self->{path} = _get_term_val($self, 'XPATH', qr/\AXPATH/); # path
-}
+our $REPLY_RE = qr/^re:\s+/i;
-sub type {
- my ($self) = @_;
- my $type = $self->{type};
- return $type if defined $type;
- $self->{type} = _get_term_val($self, 'T', qr/\AT/);
+sub subject_normalized ($) {
+ my ($subj) = @_;
+ $subj =~ s/\A\s+//s; # no leading space
+ $subj =~ s/\s+\z//s; # no trailing space
+ $subj =~ s/\s+/ /gs; # no redundant spaces
+ $subj =~ s/\.+\z//; # no trailing '.'
+ $subj =~ s/$REPLY_RE//igo; # remove reply prefix
+ $subj;
}
1;