X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FSearchView.pm;h=8e0c3cfc60d733cbdd5847f5a014b9018c5a9069;hb=cefae696bcce7ac04b762613b8936e2e446d6203;hp=b867d002da71f79d718971e1d03ef4acd5463177;hpb=b24abb7762a600a3cba45ccd9bcb5ec8d97dd37d;p=public-inbox.git

diff --git a/lib/PublicInbox/SearchView.pm b/lib/PublicInbox/SearchView.pm
index b867d002..8e0c3cfc 100644
--- a/lib/PublicInbox/SearchView.pm
+++ b/lib/PublicInbox/SearchView.pm
@@ -5,11 +5,12 @@
 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 PublicInbox::MID qw(mid2path mid_mime mid_clean mid_escape MID_ESC);
 use PublicInbox::MIME;
 require PublicInbox::Git;
 require PublicInbox::SearchThread;
@@ -20,8 +21,11 @@ sub noop {}
 sub sres_top_html {
 	my ($ctx) = @_;
 	my $q = PublicInbox::SearchQuery->new($ctx->{qp});
-	my $code = 200;
+	my $x = $q->{x};
+	my $query = $q->{'q'};
+	return PublicInbox::Mbox::mbox_all($ctx, $query) if $x eq 'm';
 
+	my $code = 200;
 	# double the limit for expanded views:
 	my $opts = {
 		limit => $LIM,
@@ -29,27 +33,34 @@ 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);
+		$mset = $ctx->{srch}->query($query, $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;
 	} else {
-		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 {
@@ -70,7 +81,7 @@ sub load_doc_retry {
 	}
 }
 
-# display non-threaded search results similar to what users expect from
+# display non-nested search results similar to what users expect from
 # regular WWW search engines:
 sub mset_summary {
 	my ($ctx, $mset, $q) = @_;
@@ -80,6 +91,8 @@ 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 +106,10 @@ 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 +130,22 @@ sub err_txt {
 }
 
 sub search_nav_top {
-	my ($mset, $q) = @_;
+	my ($mset, $q, $ctx) = @_;
+
+	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";
+	}
 
-	my $rv = "<pre>Search results ordered by [";
+	$rv .= 'Search results ordered by [';
 	if ($q->{r}) {
 		my $d = $q->qs_html(r => 0);
 		$rv .= qq{<a\nhref="?$d">date</a>|<b>relevance</b>};
@@ -129,13 +159,17 @@ sub search_nav_top {
 	my $x = $q->{x};
 	if ($x eq '') {
 		my $t = $q->qs_html(x => 't');
-		$rv .= qq{<b>summary</b>|<a\nhref="?$t">threaded</a>}
+		$rv .= qq{<b>summary</b>|<a\nhref="?$t">nested</a>}
 	} elsif ($q->{x} eq 't') {
 		my $s = $q->qs_html(x => '');
-		$rv .= qq{<a\nhref="?$s">summary</a>|<b>threaded</b>};
+		$rv .= qq{<a\nhref="?$s">summary</a>|<b>nested</b>};
 	}
 	my $A = $q->qs_html(x => 'A', r => undef);
 	$rv .= qq{|<a\nhref="?$A">Atom feed</a>]};
+	my $m = $q->qs_html(x => 'm', r => undef);
+	warn "m: $m\n";
+	$rv .= qq{\n\t\t\t\t\t\tdownload: };
+	$rv .= qq{<a\nhref="?$m"\nrel="nofollow">mbox.gz</a>};
 }
 
 sub search_nav_bot {
@@ -179,15 +213,17 @@ sub sort_relevance {
 sub mset_thread {
 	my ($ctx, $mset, $q) = @_;
 	my %pct;
-	my $msgs = $ctx->{srch}->retry_reopen(sub { [ map {
+	my $srch = $ctx->{srch};
+	my $msgs = $srch->retry_reopen(sub { [ map {
 		my $i = $_;
 		my $smsg = PublicInbox::SearchMsg->load_doc($i->get_document);
 		$pct{$smsg->mid} = $i->get_percent;
 		$smsg;
 	} ($mset->items) ]});
-
+	my $r = $q->{r};
 	my $rootset = PublicInbox::SearchThread::thread($msgs,
-		$q->{r} ? sort_relevance(\%pct) : *PublicInbox::View::sort_ts);
+		$r ? sort_relevance(\%pct) : *PublicInbox::View::sort_ts,
+		$srch);
 	my $skel = search_nav_bot($mset, $q). "<pre>";
 	my $inbox = $ctx->{-inbox};
 	$ctx->{-upfx} = '';
@@ -201,13 +237,16 @@ sub mset_thread {
 	$ctx->{seen} = {};
 	$ctx->{s_nr} = scalar(@$msgs).'+ results';
 
+	# 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) {
@@ -222,7 +261,9 @@ 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";
@@ -254,8 +295,9 @@ 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) = @_;
@@ -280,7 +322,7 @@ 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";
 
@@ -291,7 +333,7 @@ sub qs_html {
 		$qs .= "&amp;r";
 	}
 	if (my $x = $self->{x}) {
-		$qs .= "&amp;x=$x" if ($x eq 't' || $x eq 'A');
+		$qs .= "&amp;x=$x" if ($x eq 't' || $x eq 'A' || $x eq 'm');
 	}
 	$qs;
 }