X-Git-Url: http://www.git.stargrave.org/?p=public-inbox.git;a=blobdiff_plain;f=lib%2FPublicInbox%2FView.pm;h=58851edc59f5226bc47a4c6ba3f7627a19ba3e00;hp=c151f221cc9ea1667fafd20253157c1bc9d7c7a5;hb=55b22e6414b78336633ca243c113c712af2c3bcd;hpb=f886760b1b5c79181143743814a576a8fa5cf4d9

diff --git a/lib/PublicInbox/View.pm b/lib/PublicInbox/View.pm
index c151f221..58851edc 100644
--- a/lib/PublicInbox/View.pm
+++ b/lib/PublicInbox/View.pm
@@ -15,7 +15,9 @@ use PublicInbox::Address;
 use PublicInbox::WwwStream;
 use PublicInbox::Reply;
 require POSIX;
+use Time::Local qw(timegm);
 
+use constant COLS => 72;
 use constant INDENT => '  ';
 use constant TCHILD => '` ';
 sub th_pfx ($) { $_[0] == 0 ? '' : TCHILD };
@@ -57,20 +59,16 @@ sub msg_page {
 	my ($ctx) = @_;
 	my $mid = $ctx->{mid};
 	my $ibx = $ctx->{-inbox};
-	my ($first, $more, $head, $tail, $db);
+	my ($first, $more);
 	my $smsg;
 	if (my $srch = $ibx->search) {
-		$srch->retry_reopen(sub {
-			($head, $tail, $db) = $srch->each_smsg_by_mid($mid);
-			for (; !defined($first) && $head != $tail; $head++) {
-				my @args = ($head, $db, $mid);
-				$smsg = PublicInbox::SearchMsg->get(@args);
-				$first = $ibx->msg_by_smsg($smsg);
-			}
-			if ($head != $tail) {
-				$more = [ $head, $tail, $db ];
-			}
-		});
+		my ($id, $prev);
+		$smsg = $srch->next_by_mid($mid, \$id, \$prev);
+		$first = $ibx->msg_by_smsg($smsg) if $smsg;
+		if ($first) {
+			my $next = $srch->next_by_mid($mid, \$id, \$prev);
+			$more = [ $id, $prev, $next ] if $next;
+		}
 		return unless $first;
 	} else {
 		$first = $ibx->msg_by_mid($mid) or return;
@@ -81,18 +79,11 @@ sub msg_page {
 sub msg_html_more {
 	my ($ctx, $more, $nr) = @_;
 	my $str = eval {
-		my $smsg;
-		my ($head, $tail, $db) = @$more;
+		my ($id, $prev, $smsg) = @$more;
 		my $mid = $ctx->{mid};
-		for (; !defined($smsg) && $head != $tail; $head++) {
-			my $m = PublicInbox::SearchMsg->get($head, $db, $mid);
-			$smsg = $ctx->{-inbox}->smsg_mime($m);
-		}
-		if ($head == $tail) { # done
-			@$more = ();
-		} else {
-			$more->[0] = $head;
-		}
+		$smsg = $ctx->{-inbox}->smsg_mime($smsg);
+		my $next = $ctx->{srch}->next_by_mid($mid, \$id, \$prev);
+		@$more = $next ? ($id, $prev, $next) : ();
 		if ($smsg) {
 			my $mime = $smsg->{mime};
 			my $upfx = '../' . mid_escape($smsg->mid) . '/';
@@ -174,6 +165,24 @@ sub in_reply_to {
 	$refs->[-1];
 }
 
+sub fold_addresses ($) {
+	return $_[0] if length($_[0]) <= COLS;
+	# try to fold on commas after non-word chars before $lim chars,
+	# Try to get the "," preceeded by ">" or ")", but avoid folding
+	# on the comma where somebody uses "Lastname, Firstname".
+	# We also try to keep the last and penultimate addresses in
+	# the list on the same line if possible, hence the extra \z
+	# Fall back to folding on spaces at $lim + 1 chars
+	my $lim = COLS - 8; # 8 = "\t" display width
+	my $too_long = $lim + 1;
+	$_[0] =~ s/\s*\z//s; # Email::Simple doesn't strip trailing spaces
+	$_[0] = join("\n\t",
+		($_[0] =~ /(.{0,$lim}\W(?:,|\z)|
+				.{1,$lim}(?:,|\z)|
+				.{1,$lim}|
+				.{$too_long,}?)(?:\s|\z)/xgo));
+}
+
 sub _hdr_names_html ($$) {
 	my ($hdr, $field) = @_;
 	my $val = $hdr->header($field) or return '';
@@ -208,13 +217,6 @@ sub index_entry {
 	my @tocc;
 	my $mime = $smsg->{mime};
 	my $hdr = $mime->header_obj;
-	foreach my $f (qw(To Cc)) {
-		my $dst = _hdr_names_html($hdr, $f);
-		if ($dst ne '') {
-			obfuscate_addrs($obfs_ibx, $dst) if $obfs_ibx;
-			push @tocc, "$f: $dst";
-		}
-	}
 	my $from = _hdr_names_html($hdr, 'From');
 	obfuscate_addrs($obfs_ibx, $from) if $obfs_ibx;
 	$rv .= "From: $from @ ".fmt_ts($smsg->ds)." UTC";
@@ -222,7 +224,24 @@ sub index_entry {
 	my $mhref = $upfx . mid_escape($mid_raw) . '/';
 	$rv .= qq{ (<a\nhref="$mhref">permalink</a> / };
 	$rv .= qq{<a\nhref="${mhref}raw">raw</a>)\n};
-	$rv .= '  '.join('; +', @tocc) . "\n" if @tocc;
+	my $to = fold_addresses(_hdr_names_html($hdr, 'To'));
+	my $cc = fold_addresses(_hdr_names_html($hdr, 'Cc'));
+	my ($tlen, $clen) = (length($to), length($cc));
+	my $to_cc = '';
+	if (($tlen + $clen) > COLS) {
+		$to_cc .= '  To: '.$to."\n" if $tlen;
+		$to_cc .= '  Cc: '.$cc."\n" if $clen;
+	} else {
+		if ($tlen) {
+			$to_cc .= '  To: '.$to;
+			$to_cc .= '; <b>+Cc:</b> '.$cc if $clen;
+		} else {
+			$to_cc .= '  Cc: '.$cc if $clen;
+		}
+		$to_cc .= "\n";
+	}
+	obfuscate_addrs($obfs_ibx, $to_cc) if $obfs_ibx;
+	$rv .= $to_cc;
 
 	my $mapping = $ctx->{mapping};
 	if (!$mapping && (defined($irt) || defined($irt = in_reply_to($hdr)))) {
@@ -408,9 +427,7 @@ sub thread_html {
 	my ($ctx) = @_;
 	my $mid = $ctx->{mid};
 	my $srch = $ctx->{srch};
-	my $sres = $srch->get_thread($mid);
-	my $msgs = $sres->{msgs};
-	my $nr = $sres->{total};
+	my ($nr, $msgs) = $srch->get_thread($mid);
 	return missing_thread($ctx) if $nr == 0;
 	my $skel = '<hr><pre>';
 	$skel .= $nr == 1 ? 'only message in thread' : 'end of thread';
@@ -603,29 +620,43 @@ sub _msg_html_prepare {
 		$ctx->{-upfx} = '../';
 	}
 	my @title;
-	foreach my $h (qw(From To Cc Subject Date)) {
-		my $v = $hdr->header($h);
-		defined($v) && ($v ne '') or next;
+	my $v;
+	if (defined($v = $hdr->header('From'))) {
 		$v = PublicInbox::Hval->new($v);
-
-		if ($h eq 'From') {
-			my @n = PublicInbox::Address::names($v->raw);
-			$title[1] = ascii_html(join(', ', @n));
-			obfuscate_addrs($obfs_ibx, $title[1]) if $obfs_ibx;
-		} elsif ($h eq 'Subject') {
-			$title[0] = $v->as_html;
-			if ($srch) {
-				$rv .= qq($h: <a\nhref="#r"\nid=t>);
-				$rv .= $v->as_html . "</a>\n";
-				next;
-			}
-		}
+		my @n = PublicInbox::Address::names($v->raw);
+		$title[1] = ascii_html(join(', ', @n));
 		$v = $v->as_html;
+		if ($obfs_ibx) {
+			obfuscate_addrs($obfs_ibx, $v);
+			obfuscate_addrs($obfs_ibx, $title[1]);
+		}
+		$rv .= "From: $v\n" if $v ne '';
+	}
+	foreach my $h (qw(To Cc)) {
+		defined($v = $hdr->header($h)) or next;
+		fold_addresses($v);
+		$v = ascii_html($v);
 		obfuscate_addrs($obfs_ibx, $v) if $obfs_ibx;
-		$rv .= "$h: $v\n";
-
+		$rv .= "$h: $v\n" if $v ne '';
+	}
+	if (defined($v = $hdr->header('Subject')) && ($v ne '')) {
+		$v = ascii_html($v);
+		obfuscate_addrs($obfs_ibx, $v) if $obfs_ibx;
+		if ($srch) {
+			$rv .= qq(Subject: <a\nhref="#r"\nid=t>$v</a>\n);
+		} else {
+			$rv .= "Subject: $v\n";
+		}
+		$title[0] = $v;
+	} else { # dummy anchor for thread skeleton at bottom of page
+		$rv .= qq(<a\nhref="#r"\nid=t></a>) if $srch;
+		$title[0] = '(no subject)';
+	}
+	if (defined($v = $hdr->header('Date'))) {
+		$v = ascii_html($v);
+		obfuscate_addrs($obfs_ibx, $v) if $obfs_ibx; # possible :P
+		$rv .= "Date: $v\n";
 	}
-	$title[0] ||= '(no subject)';
 	$ctx->{-title_html} = join(' - ', @title);
 	foreach (@$mids) {
 		my $mid = PublicInbox::Hval->new_msgid($_) ;
@@ -649,8 +680,7 @@ sub thread_skel {
 	my ($dst, $ctx, $hdr, $tpfx) = @_;
 	my $srch = $ctx->{srch};
 	my $mid = mids($hdr)->[0];
-	my $sres = $srch->get_thread($mid);
-	my $nr = $sres->{total};
+	my ($nr, $msgs) = $srch->get_thread($mid);
 	my $expand = qq(expand[<a\nhref="${tpfx}T/#u">flat</a>) .
 	                qq(|<a\nhref="${tpfx}t/#u">nested</a>]  ) .
 			qq(<a\nhref="${tpfx}t.mbox.gz">mbox.gz</a>  ) .
@@ -680,12 +710,11 @@ sub thread_skel {
 	$ctx->{prev_attr} = '';
 	$ctx->{prev_level} = 0;
 	$ctx->{dst} = $dst;
-	$sres = $sres->{msgs};
 
 	# reduce hash lookups in skel_dump
 	my $ibx = $ctx->{-inbox};
 	$ctx->{-obfs_ibx} = $ibx->{obfuscate} ? $ibx : undef;
-	walk_thread(thread_results($ctx, $sres), $ctx, *skel_dump);
+	walk_thread(thread_results($ctx, $msgs), $ctx, *skel_dump);
 
 	$ctx->{parent_msg} = $parent;
 }
@@ -694,7 +723,8 @@ sub _parent_headers {
 	my ($hdr, $srch) = @_;
 	my $rv = '';
 
-	my $irt = in_reply_to($hdr);
+	my $refs = references($hdr);
+	my $irt = pop @$refs;
 	if (defined $irt) {
 		my $v = PublicInbox::Hval->new_msgid($irt);
 		my $html = $v->as_html;
@@ -707,22 +737,9 @@ sub _parent_headers {
 	# we show the thread skeleton at the bottom, instead.
 	return $rv if $srch;
 
-	my $refs = $hdr->header_raw('References');
-	if ($refs) {
-		# avoid redundant URLs wasting bandwidth
-		my %seen;
-		$seen{$irt} = 1 if defined $irt;
-		my @refs;
-		my @raw_refs = ($refs =~ /<([^>]+)>/g);
-		foreach my $ref (@raw_refs) {
-			next if $seen{$ref};
-			$seen{$ref} = 1;
-			push @refs, linkify_ref_nosrch($ref);
-		}
-
-		if (@refs) {
-			$rv .= 'References: '. join("\n\t", @refs) . "\n";
-		}
+	if (@$refs) {
+		@$refs = map { linkify_ref_nosrch($_) } @$refs;
+		$rv .= 'References: '. join("\n\t", @$refs) . "\n";
 	}
 	$rv;
 }
@@ -1036,59 +1053,81 @@ sub dump_topics {
 	200;
 }
 
+sub ts2str ($) {
+	my ($ts) = @_;
+	POSIX::strftime('%Y%m%d%H%M%S', gmtime($ts));
+}
+
+sub str2ts ($) {
+	my ($yyyy, $mon, $dd, $hh, $mm, $ss) = unpack('A4A2A2A2A2A2', $_[0]);
+	timegm($ss, $mm, $hh, $dd, $mon - 1, $yyyy);
+}
+
+sub pagination_footer ($$) {
+	my ($ctx, $latest) = @_;
+	delete $ctx->{qp} or return;
+	my $next = $ctx->{next_page} || '';
+	my $prev = $ctx->{prev_page} || '';
+	if ($prev) {
+		$next = $next ? "$next " : '     ';
+		$prev .= qq! <a\nhref='$latest'>latest</a>!;
+	}
+	"<hr><pre>page: $next$prev</pre>";
+}
+
 sub index_nav { # callback for WwwStream
 	my (undef, $ctx) = @_;
-	delete $ctx->{qp} or return;
-	my ($next, $prev);
-	$next = $prev = '    ';
-	my $latest = '';
+	pagination_footer($ctx, '.')
+}
 
-	my $next_o = $ctx->{-next_o};
-	if ($next_o) {
-		$next = qq!<a\nhref="?o=$next_o"\nrel=next>next</a>!;
+sub paginate_recent ($$) {
+	my ($ctx, $lim) = @_;
+	my $t = $ctx->{qp}->{t} || '';
+	my $opts = { limit => $lim };
+	my ($after, $before);
+
+	# Xapian uses '..' but '-' is perhaps friendier to URL linkifiers
+	# if only $after exists "YYYYMMDD.." because "." could be skipped
+	# if interpreted as an end-of-sentence
+	$t =~ s/\A(\d{8,14})-// and $after = str2ts($1);
+	$t =~ /\A(\d{8,14})\z/ and $before = str2ts($1);
+
+	my $ibx = $ctx->{-inbox};
+	my $msgs = $ibx->recent($opts, $after, $before);
+	my $nr = scalar @$msgs;
+	if ($nr < $lim && defined($after)) {
+		$after = $before = undef;
+		$msgs = $ibx->recent($opts);
+		$nr = scalar @$msgs;
 	}
-	if (my $cur_o = $ctx->{-cur_o}) {
-		$latest = qq! <a\nhref=.>latest</a>!;
-
-		my $o = $cur_o - ($next_o - $cur_o);
-		if ($o > 0) {
-			$prev = qq!<a\nhref="?o=$o"\nrel=prev>prev</a>!;
-		} elsif ($o == 0) {
-			$prev = qq!<a\nhref=.\nrel=prev>prev</a>!;
+	my $more = $nr == $lim;
+	my ($newest, $oldest);
+	if ($nr) {
+		$newest = $msgs->[0]->{ts};
+		$oldest = $msgs->[-1]->{ts};
+		# if we only had $after, our SQL query in ->recent ordered
+		if ($newest < $oldest) {
+			($oldest, $newest) = ($newest, $oldest);
+			$more = 0 if defined($after) && $after < $oldest;
 		}
 	}
-	"<hr><pre>page: $next $prev$latest</pre>";
+	if (defined($oldest) && $more) {
+		my $s = ts2str($oldest);
+		$ctx->{next_page} = qq!<a\nhref="?t=$s"\nrel=next>next</a>!;
+	}
+	if (defined($newest) && (defined($before) || defined($after))) {
+		my $s = ts2str($newest);
+		$ctx->{prev_page} = qq!<a\nhref="?t=$s-"\nrel=prev>prev</a>!;
+	}
+	$msgs;
 }
 
 sub index_topics {
 	my ($ctx) = @_;
-	my ($off) = (($ctx->{qp}->{o} || '0') =~ /(\d+)/);
-	my $lim = 200;
-	my $opts = { offset => $off, limit => $lim };
-
-	$ctx->{order} = [];
-	my $srch = $ctx->{srch};
-
-	my $qs = '';
-	# this complicated bit cuts loading time by over 400ms on my system:
-	if ($off == 0) {
-		my ($min, $max) = $ctx->{-inbox}->mm->minmax;
-		my $n = $max - $lim;
-		$n = $min if $n < $min;
-		for (; $qs eq '' && $n >= $min; --$n) {
-			my $smsg = $srch->lookup_article($n) or next;
-			$qs = POSIX::strftime('d:%Y%m%d..', gmtime($smsg->ts));
-		}
-	}
-
-	my $sres = $srch->query($qs, $opts);
-	$sres = $sres->{msgs};
-	my $nr = scalar @$sres;
-	if ($nr) {
-		walk_thread(thread_results($ctx, $sres), $ctx, *acc_topic);
+	my $msgs = paginate_recent($ctx, 200); # 200 is our window
+	if (@$msgs) {
+		walk_thread(thread_results($ctx, $msgs), $ctx, *acc_topic);
 	}
-	$ctx->{-next_o} = $off + $nr;
-	$ctx->{-cur_o} = $off;
 	PublicInbox::WwwStream->response($ctx, dump_topics($ctx), *index_nav);
 }