+ my $id = $msgid;
+ if ($id !~ /\A[a-f0-9]{40}\z/) {
+ $id = mid_compress(mid_clean($id), 1);
+ }
+ 'm' . $id;
+}
+
+sub thread_html_head {
+ my ($cb, $mime) = @_;
+ $$cb = $$cb->([200, ['Content-Type'=> 'text/html; charset=UTF-8']]);
+
+ my $s = PublicInbox::Hval->new_oneline($mime->header('Subject'));
+ $s = $s->as_html;
+ $$cb->write("<html><head><title>$s</title></head><body>");
+}
+
+sub pre_anchor_entry {
+ my ($seen, $mime) = @_;
+ my $id = anchor_for($mime->header('Message-ID'));
+ $seen->{$id} = "#$id"; # save the anchor for children, later
+}
+
+sub ghost_parent {
+ my ($upfx, $mid) = @_;
+ # 'subject dummy' is used internally by Mail::Thread
+ return '[no common parent]' if ($mid eq 'subject dummy');
+
+ $mid = PublicInbox::Hval->new_msgid($mid);
+ my $href = $mid->as_href;
+ my $html = $mid->as_html;
+ qq{[parent not found: <<a\nhref="$upfx../$href/">$html</a>>]};
+}
+
+sub __thread_entry {
+ my ($cb, $git, $state, $mime, $level) = @_;
+
+ # lazy load the full message from mini_mime:
+ $mime = eval {
+ my $path = mid2path(mid_clean($mime->header('Message-ID')));
+ Email::MIME->new($git->cat_file('HEAD:'.$path));
+ } or return;
+
+ if ($state->{anchor_idx} == 0) {
+ thread_html_head($cb, $mime, $state);
+ }
+
+ if (my $ghost = delete $state->{ghost}) {
+ # n.b. ghost messages may only be parents, not children
+ foreach my $g (@$ghost) {
+ $$cb->write("<table\nsummary=ghost><tr><td>" .
+ (INDENT x $g->[1]) . "</td><td>" .
+ PRE_WRAP . ghost_parent('../', $g->[0]) .
+ '</pre></td></table>');
+ }
+ }
+ index_entry($$cb, $mime, $level, $state);
+ 1;
+}
+
+sub __ghost_entry {
+ my ($state, $node, $level) = @_;
+ my $ghost = $state->{ghost} ||= [];
+ push @$ghost, [ $node->messageid, $level ];
+}
+
+sub thread_entry {
+ my ($cb, $git, $state, $node, $level) = @_;
+ return unless $node;
+ if (my $mime = $node->message) {
+ unless (__thread_entry($cb, $git, $state, $mime, $level)) {
+ __ghost_entry($state, $node, $level);
+ }
+ } else {
+ __ghost_entry($state, $node, $level);
+ }
+
+ thread_entry($cb, $git, $state, $node->child, $level + 1);
+ thread_entry($cb, $git, $state, $node->next, $level);
+}
+
+sub load_results {
+ my ($res) = @_;
+
+ [ map { $_->mini_mime } @{delete $res->{msgs}} ];
+}
+
+sub msg_timestamp {
+ my ($mime) = @_;
+ my $ts = eval { str2time($mime->header('Date')) };
+ defined($ts) ? $ts : 0;
+}
+
+sub thread_results {
+ my ($msgs, $nosubject) = @_;
+ require PublicInbox::Thread;
+ my $th = PublicInbox::Thread->new(@$msgs);
+ no warnings 'once';
+ $Mail::Thread::nosubject = $nosubject;
+ $th->thread;
+ $th->order(*sort_ts);
+ $th
+}
+
+sub missing_thread {
+ my ($cb, $ctx) = @_;
+ require PublicInbox::ExtMsg;
+
+ $cb->(PublicInbox::ExtMsg::ext_msg($ctx))
+}
+
+sub _msg_date {
+ my ($mime) = @_;
+ my $ts = $mime->header('X-PI-TS') || msg_timestamp($mime);
+ POSIX::strftime('%Y-%m-%d %H:%M', gmtime($ts));
+}
+
+sub _inline_header {
+ my ($dst, $state, $upfx, $mime, $level) = @_;
+ my $pfx = INDENT x ($level - 1);
+ my $dot = $level == 0 ? '' : '` ';
+
+ my $cur = $state->{cur};
+ my $mid = mid_clean($mime->header('Message-ID'));
+ my $f = $mime->header('X-PI-From');
+ my $d = _msg_date($mime);
+ $f = PublicInbox::Hval->new($f)->as_html;
+ $d = PublicInbox::Hval->new($d)->as_html;
+ my $attr = "$f @ $d";
+ $state->{first_level} ||= $level;
+ if ($attr ne $state->{prev_attr} || $state->{prev_level} > $level) {
+ $state->{prev_attr} = $attr;
+ $attr = ' - ' . $attr;
+ $attr .= ' UTC' if $level >= $state->{first_level};
+ } else {
+ $attr = '';
+ }
+ $state->{prev_level} = $level;
+
+ if ($cur) {
+ if ($cur eq $mid) {
+ delete $state->{cur};
+ $$dst .= "$pfx$dot<b><a\nid=\"r\"\nhref=\"#t\">".
+ "[this message]</a></b>$attr\n";
+
+ return;
+ }
+ } else {
+ $state->{next_msg} ||= $mid;
+ }
+
+ # Subject is never undef, this mail was loaded from
+ # our Xapian which would've resulted in '' if it were
+ # really missing (and Filter rejects empty subjects)
+ my $s = $mime->header('Subject');
+ my $h = $state->{srch}->subject_path($s);
+ if ($state->{seen}->{$h}) {
+ $s = undef;
+ } else {
+ $state->{seen}->{$h} = 1;
+ $s = PublicInbox::Hval->new($s);
+ $s = $s->as_html;
+ }
+ my $m = PublicInbox::Hval->new_msgid($mid);
+ $m = $upfx . '../' . $m->as_href . '/';
+ if (defined $s) {
+ $$dst .= "$pfx$dot<a\nhref=\"$m\">$s</a>$attr\n";
+ } else {
+ $$dst .= "$pfx$dot<a\nhref=\"$m\">$f @ $d</a>\n";
+ }
+}
+
+sub inline_dump {
+ my ($dst, $state, $upfx, $node, $level) = @_;
+ return unless $node;
+ if (my $mime = $node->message) {
+ my $mid = mid_clean($mime->header('Message-ID'));
+ if ($mid eq $state->{parent_cmp}) {
+ $state->{parent} = $mid;
+ }
+ _inline_header($dst, $state, $upfx, $mime, $level);
+ } else {
+ my $dot = $level == 0 ? '' : '` ';
+ my $pfx = (INDENT x $level) . $dot;
+ $$dst .= $pfx . ghost_parent($upfx, $node->messageid) . "\n";
+ }
+ inline_dump($dst, $state, $upfx, $node->child, $level+1);
+ inline_dump($dst, $state, $upfx, $node->next, $level);
+}
+
+sub sort_ts {
+ sort {
+ (eval { $a->topmost->message->header('X-PI-TS') } || 0) <=>
+ (eval { $b->topmost->message->header('X-PI-TS') } || 0)
+ } @_;
+}
+
+sub rsort_ts {
+ sort {
+ (eval { $b->topmost->message->header('X-PI-TS') } || 0) <=>
+ (eval { $a->topmost->message->header('X-PI-TS') } || 0)
+ } @_;
+}
+
+# accumulate recent topics if search is supported
+# returns 1 if done, undef if not
+sub add_topic {
+ my ($state, $node, $level) = @_;
+ return unless $node;
+ my $child_adjust = 1;
+
+ if (my $x = $node->message) {
+ $x = $x->header_obj;
+ my ($topic, $subj);
+
+ $subj = $x->header('Subject');
+ $subj = $state->{srch}->subject_normalized($subj);
+ $topic = $subj;
+
+ # kill "[PATCH v2]" etc. for summarization
+ $topic =~ s/\A\s*\[[^\]]+\]\s*//g;
+ $topic = substr($topic, 0, 30);
+
+ if (++$state->{subjs}->{$topic} == 1) {
+ push @{$state->{order}}, [ $level, $subj, $topic ];
+ }
+
+ my $mid = mid_clean($x->header('Message-ID'));
+
+ my $u = $x->header('X-PI-From');
+ my $ts = $x->header('X-PI-TS');
+ $state->{latest}->{$topic} = [ $mid, $u, $ts ];
+ } else {
+ # ghost message, do not bump level
+ $child_adjust = 0;
+ }
+
+ add_topic($state, $node->child, $level + $child_adjust);
+ add_topic($state, $node->next, $level);
+}
+
+sub dump_topics {
+ my ($state) = @_;
+ my $order = $state->{order};
+ my $subjs = $state->{subjs};
+ my $latest = $state->{latest};
+ return "\n[No recent topics]</pre>" unless (scalar @$order);
+ my $dst = '';
+ my $pfx;
+ my $prev = 0;
+ my $prev_attr = '';
+ while (defined(my $info = shift @$order)) {
+ my ($level, $subj, $topic) = @$info;
+ my $n = delete $subjs->{$topic};
+ my ($mid, $u, $ts) = @{delete $latest->{$topic}};
+ $mid = PublicInbox::Hval->new($mid)->as_href;
+ $subj = PublicInbox::Hval->new($subj)->as_html;
+ $u = PublicInbox::Hval->new($u)->as_html;
+ $pfx = INDENT x ($level - 1);
+ my $nl = $level == $prev ? "\n" : '';
+ my $dot = $level == 0 ? '' : '` ';
+ $dst .= "$nl$pfx$dot<a\nhref=\"$mid/t/#u\"><b>$subj</b></a>\n";
+
+ my $attr;
+ $ts = POSIX::strftime('%Y-%m-%d %H:%M', gmtime($ts));
+ if ($n == 1) {
+ $attr = "created by $u @ $ts UTC";
+ $n = "\n";
+ } else {
+ # $n isn't the total number of posts on the topic,
+ # just the number of posts in the current results
+ # window, so leave it unlabeled
+ $attr = "updated by $u @ $ts UTC";
+ $n = " ($n)\n";
+ }
+ if ($level == 0 || $attr ne $prev_attr) {
+ $pfx .= INDENT if $level > 0;
+ $dst .= "$pfx- ". $attr . $n;
+ $prev_attr = $attr;
+ }
+ }
+ $dst .= '</pre>';
+}
+
+sub emit_index_topics {
+ my ($state, $fh) = @_;
+ my $off = $state->{ctx}->{cgi}->param('o');
+ $off = 0 unless defined $off;
+ $state->{order} = [];
+ $state->{subjs} = {};
+ $state->{latest} = {};
+ my $max = 25;
+ my %opts = ( offset => int $off, limit => $max * 4 );
+ while (scalar @{$state->{order}} < $max) {
+ my $res = $state->{srch}->query('', \%opts);
+ my $nr = scalar @{$res->{msgs}} or last;
+
+ for (rsort_ts(thread_results(load_results($res), 1)->rootset)) {
+ add_topic($state, $_, 0);
+ }
+ $opts{offset} += $nr;
+ }
+
+ $fh->write(dump_topics($state));
+ $opts{offset};