X-Git-Url: http://www.git.stargrave.org/?p=public-inbox.git;a=blobdiff_plain;f=lib%2FPublicInbox%2FView.pm;h=33b323dce248670dc1da248b795419470b4c0019;hp=544ded7dd029f8f67e8c47e50e24735ac1c26823;hb=78792fee4031aee0a8bd532bff98efc2b9409d68;hpb=256d049b7ac24f40bb1df691fe014853b75bf491

diff --git a/lib/PublicInbox/View.pm b/lib/PublicInbox/View.pm
index 544ded7d..33b323dc 100644
--- a/lib/PublicInbox/View.pm
+++ b/lib/PublicInbox/View.pm
@@ -1,4 +1,4 @@
-# Copyright (C) 2014-2019 all contributors <meta@public-inbox.org>
+# Copyright (C) 2014-2020 all contributors <meta@public-inbox.org>
 # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
 #
 # Used for displaying the HTML web interface.
@@ -8,9 +8,10 @@ use strict;
 use warnings;
 use bytes (); # only for bytes::length
 use PublicInbox::MsgTime qw(msg_datestamp);
-use PublicInbox::Hval qw(ascii_html obfuscate_addrs prurl);
+use PublicInbox::Hval qw(ascii_html obfuscate_addrs prurl mid_href);
 use PublicInbox::Linkify;
-use PublicInbox::MID qw/id_compress mid_escape mids mids_for_index references/;
+use PublicInbox::MID qw(id_compress mids mids_for_index references
+			$MID_EXTRACT);
 use PublicInbox::MsgIter;
 use PublicInbox::Address;
 use PublicInbox::WwwStream;
@@ -18,29 +19,22 @@ use PublicInbox::Reply;
 use PublicInbox::ViewDiff qw(flush_diff);
 use POSIX qw(strftime);
 use Time::Local qw(timegm);
-use PublicInbox::SearchMsg qw(subject_normalized);
+use PublicInbox::Smsg qw(subject_normalized);
 use constant COLS => 72;
 use constant INDENT => '  ';
 use constant TCHILD => '` ';
 sub th_pfx ($) { $_[0] == 0 ? '' : TCHILD };
 
-sub msg_html_i {
+sub msg_page_i {
 	my ($nr, $ctx) = @_;
-	my $more = $ctx->{more};
-	if ($nr == 1) {
-		# $more cannot be true w/o $smsg being defined:
-		my $upfx = $more ? '../'.mid_escape($ctx->{smsg}->mid).'/' : '';
-		multipart_text_as_html(delete $ctx->{mime}, $upfx, $ctx);
-		${delete $ctx->{obuf}} .= '</pre><hr>';
-	} elsif ($more) {
-		++$ctx->{end_nr};
-		# fake an EOF if {more} retrieval fails fails;
-		eval { msg_html_more($ctx, $nr) };
-	} elsif ($nr == $ctx->{end_nr}) {
+	if (my $more = delete $ctx->{more}) { # unlikely
+		# fake an EOF if $more retrieval fails;
+		eval { msg_page_more($ctx, $nr, @$more) };
+	} elsif (my $hdr = delete $ctx->{hdr}) {
 		# fake an EOF if generating the footer fails;
 		# we want to at least show the message if something
 		# here crashes:
-		eval { html_footer($ctx) };
+		eval { html_footer($ctx, $hdr) };
 	} else {
 		undef
 	}
@@ -48,51 +42,42 @@ sub msg_html_i {
 
 # public functions: (unstable)
 
-sub msg_html {
-	my ($ctx, $mime, $smsg) = @_;
-	my $ibx = $ctx->{-inbox};
-	$ctx->{-obfs_ibx} = $ibx->{obfuscate} ? $ibx : undef;
-	my $hdr = $ctx->{hdr} = $mime->header_obj;
-	_msg_html_prepare_obuf($hdr, $ctx, 0);
-	$ctx->{end_nr} = 2;
-	$ctx->{smsg} = $smsg;
-	$ctx->{mime} = $mime;
-	PublicInbox::WwwStream->response($ctx, 200, \&msg_html_i);
-}
-
 sub msg_page {
 	my ($ctx) = @_;
 	my $mid = $ctx->{mid};
 	my $ibx = $ctx->{-inbox};
-	my ($first);
-	my $smsg;
+	my ($smsg, $first, $next);
 	if (my $over = $ibx->over) {
 		my ($id, $prev);
-		$smsg = $over->next_by_mid($mid, \$id, \$prev);
-		$first = $ibx->msg_by_smsg($smsg) if $smsg;
-		if ($first) {
-			my $next = $over->next_by_mid($mid, \$id, \$prev);
-			$ctx->{more} = [ $id, $prev, $next ] if $next;
-		}
-		return unless $first;
+		$smsg = $over->next_by_mid($mid, \$id, \$prev) or return;
+		$first = $ibx->msg_by_smsg($smsg) or return;
+		$next = $over->next_by_mid($mid, \$id, \$prev);
+		$ctx->{more} = [ $id, $prev, $next ] if $next;
 	} else {
 		$first = $ibx->msg_by_mid($mid) or return;
 	}
-	msg_html($ctx, PublicInbox::MIME->new($first), $smsg);
+	my $mime = PublicInbox::MIME->new($first);
+	$ctx->{-obfs_ibx} = $ibx->{obfuscate} ? $ibx : undef;
+	my $hdr = $ctx->{hdr} = $mime->header_obj;
+	$ctx->{obuf} = _msg_page_prepare_obuf($hdr, $ctx, 0);
+	$ctx->{smsg} = $smsg;
+	# $next cannot be true w/o $smsg being defined:
+	$ctx->{mhref} = $next ? '../'.mid_href($smsg->{mid}).'/' : '';
+	multipart_text_as_html($mime, $ctx);
+	$ctx->{-html_tip} = (${delete $ctx->{obuf}} .= '</pre><hr>');
+	PublicInbox::WwwStream->response($ctx, 200, \&msg_page_i);
 }
 
-sub msg_html_more {
-	my ($ctx, $nr) = @_;
-	my ($id, $prev, $smsg) = @{$ctx->{more}};
+sub msg_page_more { # cold
+	my ($ctx, $nr, $id, $prev, $smsg) = @_;
 	my $ibx = $ctx->{-inbox};
-	$smsg = $ibx->smsg_mime($smsg);
 	my $next = $ibx->over->next_by_mid($ctx->{mid}, \$id, \$prev);
-	$ctx->{more} = $next ? [ $id, $prev, $next ] : undef;
-	return '' unless $smsg;
-	my $upfx = '../' . mid_escape($smsg->mid) . '/';
+	$ctx->{more} = [ $id, $prev, $next ] if $next;
+	$smsg = $ibx->smsg_mime($smsg) or return '';
+	$ctx->{mhref} = '../' . mid_href($smsg->{mid}) . '/';
 	my $mime = delete $smsg->{mime};
-	_msg_html_prepare_obuf($mime->header_obj, $ctx, $nr);
-	multipart_text_as_html($mime, $upfx, $ctx);
+	$ctx->{obuf} = _msg_page_prepare_obuf($mime->header_obj, $ctx, $nr);
+	multipart_text_as_html($mime, $ctx);
 	${delete $ctx->{obuf}} .= '</pre><hr>';
 }
 
@@ -225,7 +210,7 @@ sub index_entry {
 	obfuscate_addrs($obfs_ibx, $from) if $obfs_ibx;
 	$rv .= "From: $from @ ".fmt_ts($ds)." UTC";
 	my $upfx = $ctx->{-upfx};
-	my $mhref = $upfx . mid_escape($mid_raw) . '/';
+	my $mhref = $upfx . mid_href($mid_raw) . '/';
 	$rv .= qq{ (<a\nhref="$mhref">permalink</a> / };
 	$rv .= qq{<a\nhref="${mhref}raw">raw</a>)\n};
 	my $to = fold_addresses(_hdr_names_html($hdr, 'To'));
@@ -249,9 +234,8 @@ sub index_entry {
 
 	my $mapping = $ctx->{mapping};
 	if (!$mapping && (defined($irt) || defined($irt = in_reply_to($hdr)))) {
-		my $mirt = PublicInbox::Hval->new_msgid($irt);
-		my $href = $upfx . $mirt->{href}. '/';
-		my $html = $mirt->as_html;
+		my $href = $upfx . mid_href($irt) . '/';
+		my $html = ascii_html($irt);
 		$rv .= qq(In-Reply-To: &lt;<a\nhref="$href">$html</a>&gt;\n)
 	}
 	$rv .= "\n";
@@ -316,7 +300,7 @@ sub _th_index_lite {
 	if (my $smsg = $node->{smsg}) {
 		# delete saves about 200KB on a 1K message thread
 		if (my $refs = delete $smsg->{references}) {
-			($$irt) = ($refs =~ m/<([^>]+)>\z/);
+			($$irt) = ($refs =~ m/$MID_EXTRACT\z/o);
 		}
 	}
 	my $irt_map = $mapping->{$$irt} if defined $$irt;
@@ -428,7 +412,7 @@ sub stream_thread ($$) {
 	return missing_thread($ctx) unless $smsg;
 
 	$ctx->{-obfs_ibx} = $ibx->{obfuscate} ? $ibx : undef;
-	$ctx->{-title_html} = ascii_html($smsg->subject);
+	$ctx->{-title_html} = ascii_html($smsg->{subject});
 	$ctx->{-html_tip} = thread_index_entry($ctx, $level, $smsg);
 	$ctx->{-queue} = \@q;
 	PublicInbox::WwwStream->response($ctx, 200, \&stream_thread_i);
@@ -474,7 +458,7 @@ sub thread_html {
 		$smsg = $ibx->smsg_mime($m) and last;
 	}
 	return missing_thread($ctx) unless $smsg;
-	$ctx->{-title_html} = ascii_html($smsg->subject);
+	$ctx->{-title_html} = ascii_html($smsg->{subject});
 	$ctx->{-html_tip} = '<pre>'.index_entry($smsg, $ctx, scalar @$msgs);
 	$ctx->{msgs} = $msgs;
 	PublicInbox::WwwStream->response($ctx, 200, \&thread_html_i);
@@ -492,25 +476,10 @@ sub thread_html_i { # PublicInbox::WwwStream::getline callback
 }
 
 sub multipart_text_as_html {
-	my (undef, $mhref, $ctx) = @_; # $mime = $_[0]
-	$ctx->{mhref} = $mhref;
+	# ($mime, $ctx) = @_; # msg_iter will do "$_[0] = undef"
 
 	# scan through all parts, looking for displayable text
-	msg_iter($_[0], \&add_text_body, $ctx, 1);
-}
-
-sub flush_quote {
-	my ($s, $l, $quot) = @_;
-
-	# show everything in the full version with anchor from
-	# short version (see above)
-	my $rv = $l->linkify_1($$quot);
-
-	# we use a <span> here to allow users to specify their own
-	# color for quoted text
-	$rv = $l->linkify_2(ascii_html($rv));
-	$$quot = undef;
-	$$s .= qq(<span\nclass="q">) . $rv . '</span>'
+	msg_iter($_[0], \&add_text_body, $_[1], 1);
 }
 
 sub attach_link ($$$$;$) {
@@ -524,9 +493,6 @@ sub attach_link ($$$$;$) {
 	# spotting MUA problems:
 	$ct =~ s/;.*// unless $err;
 	$ct = ascii_html($ct);
-	my $desc = $part->header('Content-Description');
-	$desc = $fn unless defined $desc;
-	$desc = '' unless defined $desc;
 	my $sfn;
 	if (defined $fn && $fn =~ /\A$PublicInbox::Hval::FN\z/o) {
 		$sfn = $fn;
@@ -538,10 +504,13 @@ sub attach_link ($$$$;$) {
 	my $rv = $ctx->{obuf};
 	$$rv .= qq($nl<a\nhref="$ctx->{mhref}$idx-$sfn">);
 	if ($err) {
-		$$rv .= "[-- Warning: decoded text below may be mangled --]\n";
+		$$rv .= <<EOF;
+[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
+EOF
 	}
 	$$rv .= "[-- Attachment #$idx: ";
 	my $ts = "Type: $ct, Size: $size bytes";
+	my $desc = $part->header('Content-Description') // $fn // '';
 	$desc = ascii_html($desc);
 	$$rv .= ($desc eq '') ? "$ts --]" : "$desc --]\n[-- $ts --]";
 	$$rv .= "</a>\n";
@@ -563,6 +532,9 @@ sub add_text_body { # callback for msg_iter
 	# link generation in diffs with the extra '%0D'
 	$s =~ s/\r\n/\n/sg;
 
+	# will be escaped to `&#8226;' in HTML
+	obfuscate_addrs($ibx, $s, "\x{2022}") if $ibx->{obfuscate};
+
 	# always support diff-highlighting, but we can't linkify hunk
 	# headers for solver unless some coderepo are configured:
 	my $diff;
@@ -571,7 +543,7 @@ sub add_text_body { # callback for msg_iter
 		$idx[0] = $upfx . $idx[0] if $upfx ne '';
 		$ctx->{-apfx} = join('/', @idx);
 		$ctx->{-anchors} = {}; # attr => filename
-		$ctx->{-diff} = $diff = [];
+		$diff = 1;
 		delete $ctx->{-long_path};
 		my $spfx;
 		if ($ibx->{-repo_objs}) {
@@ -596,34 +568,33 @@ sub add_text_body { # callback for msg_iter
 	$s .= "\n" unless $s =~ /\n\z/s;
 
 	# split off quoted and unquoted blocks:
-	my @sections = split(/((?:^>[^\n]*\n)+)/sm, $s);
-	$s = '';
+	my @sections = PublicInbox::MsgIter::split_quotes($s);
+	undef $s; # free memory
 	my $rv = $ctx->{obuf};
 	if (defined($fn) || $depth > 0 || $err) {
 		# badly-encoded message with $err? tell the world about it!
 		attach_link($ctx, $ct, $p, $fn, $err);
 		$$rv .= "\n";
 	}
-	my $l = PublicInbox::Linkify->new;
+	my $l = $ctx->{-linkify} //= PublicInbox::Linkify->new;
 	foreach my $cur (@sections) {
 		if ($cur =~ /\A>/) {
-			flush_quote($rv, $l, \$cur);
+			# we use a <span> here to allow users to specify
+			# their own color for quoted text
+			$$rv .= qq(<span\nclass="q">);
+			$$rv .= $l->to_html($cur);
+			$$rv .= '</span>';
 		} elsif ($diff) {
-			@$diff = split(/^/m, $cur);
-			$cur = undef;
-			flush_diff($rv, $ctx, $l);
+			flush_diff($ctx, \$cur);
 		} else {
 			# regular lines, OK
-			$l->linkify_1($cur);
-			$$rv .= $l->linkify_2(ascii_html($cur));
-			$cur = undef;
+			$$rv .= $l->to_html($cur);
 		}
+		undef $cur; # free memory
 	}
-
-	obfuscate_addrs($ibx, $$rv) if $ibx->{obfuscate};
 }
 
-sub _msg_html_prepare_obuf {
+sub _msg_page_prepare_obuf {
 	my ($hdr, $ctx, $nr) = @_;
 	my $over = $ctx->{-inbox}->over;
 	my $obfs_ibx = $ctx->{-obfs_ibx};
@@ -643,10 +614,9 @@ sub _msg_html_prepare_obuf {
 	}
 	my @title; # (Subject[0], From[0])
 	for my $v ($hdr->header('From')) {
-		$v = PublicInbox::Hval->new($v);
-		my @n = PublicInbox::Address::names($v->raw);
+		my @n = PublicInbox::Address::names($v);
+		$v = ascii_html($v);
 		$title[1] //= ascii_html(join(', ', @n));
-		$v = $v->as_html;
 		if ($obfs_ibx) {
 			obfuscate_addrs($obfs_ibx, $v);
 			obfuscate_addrs($obfs_ibx, $title[1]);
@@ -663,16 +633,15 @@ sub _msg_html_prepare_obuf {
 	}
 	my @subj = $hdr->header('Subject');
 	if (@subj) {
-		for my $v (@subj) {
+		my $v = ascii_html(shift @subj);
+		obfuscate_addrs($obfs_ibx, $v) if $obfs_ibx;
+		$rv .= 'Subject: ';
+		$rv .= $over ? qq(<a\nhref="#r"\nid=t>$v</a>\n) : "$v\n";
+		$title[0] = $v;
+		for $v (@subj) { # multi-Subject message :<
 			$v = ascii_html($v);
 			obfuscate_addrs($obfs_ibx, $v) if $obfs_ibx;
-			$rv .= 'Subject: ';
-			if ($over) {
-				$rv .= qq(<a\nhref="#r"\nid=t>$v</a>\n);
-			} else {
-				$rv .= "$v\n";
-			}
-			$title[0] //= $v;
+			$rv .= "Subject: $v\n";
 		}
 	} else { # dummy anchor for thread skeleton at bottom of page
 		$rv .= qq(<a\nhref="#r"\nid=t></a>) if $over;
@@ -685,8 +654,7 @@ sub _msg_html_prepare_obuf {
 	}
 	$ctx->{-title_html} = join(' - ', @title);
 	if (scalar(@$mids) == 1) { # common case
-		my $mid = PublicInbox::Hval->new_msgid($mids->[0]);
-		my $mhtml = $mid->as_html;
+		my $mhtml = ascii_html($mids->[0]);
 		$rv .= "Message-ID: &lt;$mhtml&gt; ";
 		$rv .= "(<a\nhref=\"raw\">raw</a>)\n";
 	} else {
@@ -702,7 +670,7 @@ sub _msg_html_prepare_obuf {
 	}
 	$rv .= _parent_headers($hdr, $over);
 	$rv .= "\n";
-	$ctx->{obuf} = \$rv;
+	\$rv;
 }
 
 sub SKEL_EXPAND () {
@@ -764,9 +732,8 @@ sub _parent_headers {
 		$refs = references($hdr);
 		my $irt = pop @$refs;
 		if (defined $irt) {
-			my $v = PublicInbox::Hval->new_msgid($irt);
-			my $html = $v->as_html;
-			my $href = $v->{href};
+			my $html = ascii_html($irt);
+			my $href = mid_href($irt);
 			$rv .= "In-Reply-To: &lt;";
 			$rv .= "<a\nhref=\"../$href/\">$html</a>&gt;\n";
 		}
@@ -786,9 +753,8 @@ sub _parent_headers {
 
 # returns a string buffer via ->getline
 sub html_footer {
-	my ($ctx) = @_;
+	my ($ctx, $hdr) = @_;
 	my $ibx = $ctx->{-inbox};
-	my $hdr = delete $ctx->{hdr};
 	my $upfx = '../';
 	my $skel = " <a\nhref=\"$upfx\">index</a>";
 	my $rv = '<pre>';
@@ -800,17 +766,17 @@ sub html_footer {
 		$next = $prev = '    ';
 
 		if (my $n = $ctx->{next_msg}) {
-			$n = PublicInbox::Hval->new_msgid($n)->{href};
+			$n = mid_href($n);
 			$next = "<a\nhref=\"$upfx$n/\"\nrel=next>next</a>";
 		}
 		my $u;
 		my $par = $ctx->{parent_msg};
 		if ($par) {
-			$u = PublicInbox::Hval->new_msgid($par)->{href};
+			$u = mid_href($par);
 			$u = "$upfx$u/";
 		}
 		if (my $p = $ctx->{prev_msg}) {
-			$prev = PublicInbox::Hval->new_msgid($p)->{href};
+			$prev = mid_href($p);
 			if ($p && $par && $p eq $par) {
 				$prev = "<a\nhref=\"$upfx$prev/\"\n" .
 					'rel=prev>prev parent</a>';
@@ -832,9 +798,9 @@ sub html_footer {
 }
 
 sub linkify_ref_no_over {
-	my $v = PublicInbox::Hval->new_msgid($_[0]);
-	my $html = $v->as_html;
-	my $href = $v->{href};
+	my ($mid) = @_;
+	my $href = mid_href($mid);
+	my $html = ascii_html($mid);
 	"&lt;<a\nhref=\"../$href/\">$html</a>&gt;";
 }
 
@@ -846,9 +812,8 @@ sub anchor_for {
 sub ghost_parent {
 	my ($upfx, $mid) = @_;
 
-	$mid = PublicInbox::Hval->new_msgid($mid);
-	my $href = $mid->{href};
-	my $html = $mid->as_html;
+	my $href = mid_href($mid);
+	my $html = ascii_html($mid);
 	qq{[parent not found: &lt;<a\nhref="$upfx$href/">$html</a>&gt;]};
 }
 
@@ -917,20 +882,20 @@ sub missing_thread {
 sub dedupe_subject {
 	my ($prev_subj, $subj, $val) = @_;
 
-	my $omit = ''; # '"' denotes identical text omitted
+	my $omit; # '"' denotes identical text omitted
 	my (@prev_pop, @curr_pop);
 	while (@$prev_subj && @$subj && $subj->[-1] eq $prev_subj->[-1]) {
 		push(@prev_pop, pop(@$prev_subj));
 		push(@curr_pop, pop(@$subj));
-		$omit ||= $val;
+		$omit //= $val;
 	}
 	pop @$subj if @$subj && $subj->[-1] =~ /^re:\s*/i;
 	if (scalar(@curr_pop) == 1) {
-		$omit = '';
+		$omit = undef;
 		push @$prev_subj, @prev_pop;
 		push @$subj, @curr_pop;
 	}
-	$omit;
+	$omit // '';
 }
 
 sub skel_dump { # walk_thread callback
@@ -945,7 +910,7 @@ sub skel_dump { # walk_thread callback
 		$$skel .= delete($ctx->{sl_note}) || '';
 	}
 
-	my $f = ascii_html($smsg->from_name);
+	my $f = ascii_html($smsg->{from_name});
 	my $obfs_ibx = $ctx->{-obfs_ibx};
 	obfuscate_addrs($obfs_ibx, $f) if $obfs_ibx;
 
@@ -984,7 +949,7 @@ sub skel_dump { # walk_thread callback
 	# 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 @subj = split(/ /, subject_normalized($smsg->subject));
+	my @subj = split(/ /, subject_normalized($smsg->{subject}));
 	# remove common suffixes from the subject if it matches the previous,
 	# so we do not show redundant text at the end.
 	my $prev_subj = $ctx->{prev_subj} || [];
@@ -1009,7 +974,7 @@ sub skel_dump { # walk_thread callback
 		$map->[0] = "$d<a\nhref=\"$m\">$end";
 		$id = "\nid=r".$id;
 	} else {
-		$m = $ctx->{-upfx}.mid_escape($mid).'/';
+		$m = $ctx->{-upfx}.mid_href($mid).'/';
 	}
 	$$skel .=  $d . "<a\nhref=\"$m\"$id>" . $end;
 	1;
@@ -1023,9 +988,8 @@ sub _skel_ghost {
 	$d .= '    '  if exists $ctx->{searchview};
 	$d .= indent_for($level) . th_pfx($level);
 	my $upfx = $ctx->{-upfx};
-	my $m = PublicInbox::Hval->new_msgid($mid);
-	my $href = $upfx . $m->{href} . '/';
-	my $html = $m->as_html;
+	my $href = $upfx . mid_href($mid) . '/';
+	my $html = ascii_html($mid);
 
 	my $mapping = $ctx->{mapping};
 	my $map = $mapping->{$mid} if $mapping;
@@ -1042,8 +1006,8 @@ sub _skel_ghost {
 
 sub sort_ds {
 	[ sort {
-		(eval { $a->topmost->{smsg}->ds } || 0) <=>
-		(eval { $b->topmost->{smsg}->ds } || 0)
+		(eval { $a->topmost->{smsg}->{ds} } || 0) <=>
+		(eval { $b->topmost->{smsg}->{ds} } || 0)
 	} @{$_[0]} ];
 }
 
@@ -1052,35 +1016,32 @@ sub sort_ds {
 sub acc_topic { # walk_thread callback
 	my ($ctx, $level, $node) = @_;
 	my $mid = $node->{id};
-	my $x = $node->{smsg} || $ctx->{-inbox}->smsg_by_mid($mid);
-	my ($subj, $ds);
-	my $topic;
-	if ($x) {
-		$subj = $x->subject;
-		$subj = subject_normalized($subj);
+	my $smsg = $node->{smsg} // $ctx->{-inbox}->smsg_by_mid($mid);
+	if ($smsg) {
+		my $subj = subject_normalized($smsg->{subject});
 		$subj = '(no subject)' if $subj eq '';
-		$ds = $x->ds;
-		if ($level == 0) {
-			$topic = [ $ds, 1, { $subj => $mid }, $subj ];
+		my $ds = $smsg->{ds};
+		if ($level == 0) { # new, top-level topic
+			my $topic = [ $ds, 1, { $subj => $mid }, $subj ];
 			$ctx->{-cur_topic} = $topic;
 			push @{$ctx->{order}}, $topic;
 			return 1;
 		}
 
-		$topic = $ctx->{-cur_topic}; # should never be undef
+		# continue existing topic
+		my $topic = $ctx->{-cur_topic}; # should never be undef
 		$topic->[0] = $ds if $ds > $topic->[0];
-		$topic->[1]++;
+		$topic->[1]++; # bump N+ message counter
 		my $seen = $topic->[2];
 		if (scalar(@$topic) == 3) { # parent was a ghost
 			push @$topic, $subj;
-		} elsif (!$seen->{$subj}) {
-			push @$topic, $level, $subj;
+		} elsif (!defined($seen->{$subj})) {
+			push @$topic, $level, $subj; # @extra messages
 		}
 		$seen->{$subj} = $mid; # latest for subject
 	} else { # ghost message
 		return 1 if $level != 0; # ignore child ghosts
-		$topic = [ -666, 0, {} ];
-		$ctx->{-cur_topic} = $topic;
+		my $topic = $ctx->{-cur_topic} = [ -666, 0, {} ];
 		push @{$ctx->{order}}, $topic;
 	}
 	1;
@@ -1089,7 +1050,7 @@ sub acc_topic { # walk_thread callback
 sub dump_topics {
 	my ($ctx) = @_;
 	my $order = delete $ctx->{order}; # [ ds, subj1, subj2, subj3, ... ]
-	if (!@$order) {
+	unless ($order) {
 		$ctx->{-html_tip} = '<pre>[No topics in range]</pre>';
 		return 404;
 	}
@@ -1100,13 +1061,13 @@ sub dump_topics {
 
 	# sort by recency, this allows new posts to "bump" old topics...
 	foreach my $topic (sort { $b->[0] <=> $a->[0] } @$order) {
-		my ($ds, $n, $seen, $top, @ex) = @$topic;
+		my ($ds, $n, $seen, $top_subj, @extra) = @$topic;
 		@$topic = ();
-		next unless defined $top;  # ghost topic
-		my $mid = delete $seen->{$top};
-		my $href = mid_escape($mid);
-		my $prev_subj = [ split(/ /, $top) ];
-		$top = PublicInbox::Hval->new($top)->as_html;
+		next unless defined $top_subj;  # ghost topic
+		my $mid = delete $seen->{$top_subj};
+		my $href = mid_href($mid);
+		my $prev_subj = [ split(/ /, $top_subj) ];
+		$top_subj = ascii_html($top_subj);
 		$ds = fmt_ts($ds);
 
 		# $n isn't the total number of posts on the topic,
@@ -1122,21 +1083,21 @@ sub dump_topics {
 
 		my $mbox = qq(<a\nhref="$href/t.mbox.gz">mbox.gz</a>);
 		my $atom = qq(<a\nhref="$href/t.atom">Atom</a>);
-		my $s = "<a\nhref=\"$href/T/$anchor\">$top</a>\n" .
+		my $s = "<a\nhref=\"$href/T/$anchor\">$top_subj</a>\n" .
 			" $ds UTC $n - $mbox / $atom\n";
-		for (my $i = 0; $i < scalar(@ex); $i += 2) {
-			my $level = $ex[$i];
-			my $subj = $ex[$i + 1];
+		for (my $i = 0; $i < scalar(@extra); $i += 2) {
+			my $level = $extra[$i];
+			my $subj = $extra[$i + 1]; # already normalized
 			$mid = delete $seen->{$subj};
-			my @subj = split(/ /, subject_normalized($subj));
+			my @subj = split(/ /, $subj);
 			my @next_prev = @subj; # full copy
 			my $omit = dedupe_subject($prev_subj, \@subj, ' &#34;');
 			$prev_subj = \@next_prev;
-			$subj = ascii_html(join(' ', @subj));
+			$subj = ascii_html($subj);
 			obfuscate_addrs($obfs_ibx, $subj) if $obfs_ibx;
-			$href = mid_escape($mid);
+			$href = mid_href($mid);
 			$s .= indent_for($level) . TCHILD;
-			$s .= "<a\nhref=\"$href/T/#u\">$subj</a>$omit\n";
+			$s .= qq(<a\nhref="$href/T/#u">$subj</a>$omit\n);
 		}
 		push @out, $s;
 	}