X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FSearchView.pm;h=a5974034d20a3335ddc517e50d7b9c3e181b4b98;hb=1a8eb0a43c68e64b77fdd3c4ebd977a1ab88e71b;hp=50a2c01c4749f05ab54dc73a5d05ee04747d4884;hpb=53e8cfbe4e06e5ae6ad61fb7e9bd45804c253a72;p=public-inbox.git

diff --git a/lib/PublicInbox/SearchView.pm b/lib/PublicInbox/SearchView.pm
index 50a2c01c..a5974034 100644
--- a/lib/PublicInbox/SearchView.pm
+++ b/lib/PublicInbox/SearchView.pm
@@ -5,15 +5,16 @@
 package PublicInbox::SearchView;
 use strict;
 use warnings;
+use URI::Escape qw(uri_unescape uri_escape);
 use PublicInbox::SearchMsg;
-use PublicInbox::Hval qw/ascii_html/;
+use PublicInbox::Hval qw/ascii_html obfuscate_addrs/;
 use PublicInbox::View;
 use PublicInbox::WwwAtomStream;
-use PublicInbox::MID qw(mid2path mid_mime mid_clean mid_escape);
-use Email::MIME;
+use PublicInbox::MID qw(mid2path mid_mime mid_clean mid_escape MID_ESC);
+use PublicInbox::MIME;
 require PublicInbox::Git;
 require PublicInbox::SearchThread;
-our $LIM = 50;
+our $LIM = 200;
 
 sub noop {}
 
@@ -29,19 +30,27 @@ sub sres_top_html {
 		mset => 1,
 		relevance => $q->{r},
 	};
-	my ($mset, $total);
+	my ($mset, $total, $err, $cb);
+retry:
 	eval {
 		$mset = $ctx->{srch}->query($q->{'q'}, $opts);
 		$total = $mset->get_matches_estimated;
 	};
-	my $err = $@;
+	$err = $@;
 	ctx_prepare($q, $ctx);
-	my $cb;
 	if ($err) {
 		$code = 400;
 		$ctx->{-html_tip} = '<pre>'.err_txt($ctx, $err).'</pre><hr>';
 		$cb = *noop;
 	} elsif ($total == 0) {
+		if (defined($ctx->{-uxs_retried})) {
+			# undo retry damage:
+			$q->{'q'} = $ctx->{-uxs_retried};
+		} elsif (index($q->{'q'}, '%') >= 0) {
+			$ctx->{-uxs_retried} = $q->{'q'};
+			$q->{'q'} = uri_unescape($q->{'q'});
+			goto retry;
+		}
 		$code = 404;
 		$ctx->{-html_tip} = "<pre>\n[No results found]</pre><hr>";
 		$cb = *noop;
@@ -49,7 +58,7 @@ sub sres_top_html {
 		my $x = $q->{x};
 		return adump($_[0], $mset, $q, $ctx) if $x eq 'A';
 
-		$ctx->{-html_tip} = search_nav_top($mset, $q) . "\n\n";
+		$ctx->{-html_tip} = search_nav_top($mset, $q, $ctx) . "\n\n";
 		if ($x eq 't') {
 			$cb = mset_thread($ctx, $mset, $q);
 		} else {
@@ -80,6 +89,8 @@ sub mset_summary {
 	my $pfx = ' ' x $pad;
 	my $res = \($ctx->{-html_tip});
 	my $srch = $ctx->{srch};
+	my $ibx = $ctx->{-inbox};
+	my $obfs_ibx = $ibx->{obfuscate} ? $ibx : undef;
 	foreach my $m ($mset->items) {
 		my $rank = sprintf("%${pad}d", $m->get_rank + 1);
 		my $pct = $m->get_percent;
@@ -93,6 +104,10 @@ sub mset_summary {
 		}
 		my $s = ascii_html($smsg->subject);
 		my $f = ascii_html($smsg->from_name);
+		if ($obfs_ibx) {
+			obfuscate_addrs($obfs_ibx, $s);
+			obfuscate_addrs($obfs_ibx, $f);
+		}
 		my $ts = PublicInbox::View::fmt_ts($smsg->ts);
 		my $mid = PublicInbox::Hval->new_msgid($smsg->mid)->{href};
 		$$res .= qq{$rank. <b><a\nhref="$mid/">}.
@@ -113,9 +128,22 @@ sub err_txt {
 }
 
 sub search_nav_top {
-	my ($mset, $q) = @_;
+	my ($mset, $q, $ctx) = @_;
 
-	my $rv = "<pre>Search results ordered by [";
+	my $rv = '<pre>';
+	my $initial_q = $ctx->{-uxs_retried};
+	if (defined $initial_q) {
+		my $rewritten = $q->{'q'};
+		utf8::decode($initial_q);
+		utf8::decode($rewritten);
+		$initial_q = ascii_html($initial_q);
+		$rewritten = ascii_html($rewritten);
+		$rv .= " Warning: Initial query:\n <b>$initial_q</b>\n";
+		$rv .= " returned no results, used:\n";
+		$rv .= " <b>$rewritten</b>\n instead\n\n";
+	}
+
+	$rv .= 'Search results ordered by [';
 	if ($q->{r}) {
 		my $d = $q->qs_html(r => 0);
 		$rv .= qq{<a\nhref="?$d">date</a>|<b>relevance</b>};
@@ -145,15 +173,21 @@ sub search_nav_bot {
 	my $o = $q->{o};
 	my $end = $o + $nr;
 	my $beg = $o + 1;
-	my $rv = "</pre><hr><pre>Results $beg-$end of $total";
+	my $rv = '</pre><hr><pre id=t>';
+	if ($beg <= $end) {
+		$rv .= "Results $beg-$end of $total";
+		$rv .= ' (estimated)' if $end != $total;
+	} else {
+		$rv .= "No more results, only $total";
+	}
 	my $n = $o + $LIM;
 
 	if ($n < $total) {
 		my $qs = $q->qs_html(o => $n);
-		$rv .= qq{, <a\nhref="?$qs"\nrel=next>next</a>}
+		$rv .= qq{  <a\nhref="?$qs"\nrel=next>next</a>}
 	}
 	if ($o > 0) {
-		$rv .= $n < $total ? '/' : ',      ';
+		$rv .= $n < $total ? '/' : '       ';
 		my $p = $o - $LIM;
 		my $qs = $q->qs_html(o => ($p > 0 ? $p : 0));
 		$rv .= qq{<a\nhref="?$qs"\nrel=prev>prev</a>};
@@ -161,6 +195,15 @@ sub search_nav_bot {
 	$rv .= '</pre>';
 }
 
+sub sort_relevance {
+	my ($pct) = @_;
+	sub {
+		[ sort { (eval { $pct->{$b->topmost->{id}} } || 0)
+				<=>
+			(eval { $pct->{$a->topmost->{id}} } || 0)
+	} @{$_[0]} ] };
+}
+
 sub mset_thread {
 	my ($ctx, $mset, $q) = @_;
 	my %pct;
@@ -170,19 +213,9 @@ sub mset_thread {
 		$pct{$smsg->mid} = $i->get_percent;
 		$smsg;
 	} ($mset->items) ]});
-
-	my $th = PublicInbox::SearchThread->new($msgs);
-	$th->thread;
-	if ($q->{r}) { # order by relevance
-		$th->order(sub {
-			[ sort { (eval { $pct{$b->topmost->{id}} } || 0)
-					<=>
-				(eval { $pct{$a->topmost->{id}} } || 0)
-			} @{$_[0]} ];
-		});
-	} else { # order by time (default for threaded view)
-		$th->order(*PublicInbox::View::sort_ts);
-	}
+	my $r = $q->{r};
+	my $rootset = PublicInbox::SearchThread::thread($msgs,
+		$r ? sort_relevance(\%pct) : *PublicInbox::View::sort_ts);
 	my $skel = search_nav_bot($mset, $q). "<pre>";
 	my $inbox = $ctx->{-inbox};
 	$ctx->{-upfx} = '';
@@ -196,17 +229,20 @@ sub mset_thread {
 	$ctx->{seen} = {};
 	$ctx->{s_nr} = scalar(@$msgs).'+ results';
 
-	PublicInbox::View::walk_thread($th, $ctx,
+	# reduce hash lookups in skel_dump
+	$ctx->{-obfuscate} = $ctx->{-inbox}->{obfuscate};
+	PublicInbox::View::walk_thread($rootset, $ctx,
 		*PublicInbox::View::pre_thread);
 
+	@$msgs = reverse @$msgs if $r;
 	my $mime;
 	sub {
 		return unless $msgs;
-		while ($mime = shift @$msgs) {
+		while ($mime = pop @$msgs) {
 			$mime = $inbox->msg_by_smsg($mime) and last;
 		}
 		if ($mime) {
-			$mime = Email::MIME->new($mime);
+			$mime = PublicInbox::MIME->new($mime);
 			return PublicInbox::View::index_entry($mime, $ctx,
 				scalar @$msgs);
 		}
@@ -217,7 +253,9 @@ sub mset_thread {
 
 sub ctx_prepare {
 	my ($q, $ctx) = @_;
-	my $qh = ascii_html($q->{'q'});
+	my $qh = $q->{'q'};
+	utf8::decode($qh);
+	$qh = ascii_html($qh);
 	$ctx->{-q_value_html} = $qh;
 	$ctx->{-atom} = '?'.$q->qs_html(x => 'A', r => undef);
 	$ctx->{-title_html} = "$qh - search results";
@@ -240,7 +278,7 @@ sub adump {
 		while (my $x = shift @items) {
 			$x = load_doc_retry($srch, $x);
 			$x = $ibx->msg_by_smsg($x) and
-					return Email::MIME->new($x);
+					return PublicInbox::MIME->new($x);
 		}
 		return undef;
 	});
@@ -249,8 +287,9 @@ sub adump {
 package PublicInbox::SearchQuery;
 use strict;
 use warnings;
+use URI::Escape qw(uri_escape);
 use PublicInbox::Hval;
-use PublicInbox::MID qw(mid_escape);
+use PublicInbox::MID qw(MID_ESC);
 
 sub new {
 	my ($class, $qp) = @_;
@@ -275,7 +314,7 @@ sub qs_html {
 		$self = $tmp;
 	}
 
-	my $q = mid_escape($self->{'q'});
+	my $q = uri_escape($self->{'q'}, MID_ESC);
 	$q =~ s/%20/+/g; # improve URL readability
 	my $qs = "q=$q";