X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FSearchView.pm;h=584b88ed5d4e03b06e5f64a326a83d20ca86e5e9;hb=53046dfabf5cea98662dc2e2297947be793262ce;hp=50a2c01c4749f05ab54dc73a5d05ee04747d4884;hpb=53e8cfbe4e06e5ae6ad61fb7e9bd45804c253a72;p=public-inbox.git

diff --git a/lib/PublicInbox/SearchView.pm b/lib/PublicInbox/SearchView.pm
index 50a2c01c..584b88ed 100644
--- a/lib/PublicInbox/SearchView.pm
+++ b/lib/PublicInbox/SearchView.pm
@@ -1,55 +1,81 @@
-# Copyright (C) 2015 all contributors <meta@public-inbox.org>
-# License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt)
+# Copyright (C) 2015-2019 all contributors <meta@public-inbox.org>
+# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
 #
 # Displays search results for the web interface
 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;
-require PublicInbox::Git;
-require PublicInbox::SearchThread;
-our $LIM = 50;
+use PublicInbox::SearchThread;
+our $LIM = 200;
+my %rmap_inc;
 
 sub noop {}
 
+sub mbox_results {
+	my ($ctx) = @_;
+	my $q = PublicInbox::SearchQuery->new($ctx->{qp});
+	my $x = $q->{x};
+	require PublicInbox::Mbox;
+	return PublicInbox::Mbox::mbox_all($ctx, $q->{'q'}) if $x eq 'm';
+	sres_top_html($ctx);
+}
+
 sub sres_top_html {
 	my ($ctx) = @_;
+	my $srch = $ctx->{-inbox}->search or
+		return PublicInbox::WWW::need($ctx, 'Search');
 	my $q = PublicInbox::SearchQuery->new($ctx->{qp});
-	my $code = 200;
+	my $x = $q->{x};
+	my $query = $q->{'q'};
+	my $o = $q->{o};
+	my $asc;
+	if ($o < 0) {
+		$asc = 1;
+		$o = -($o + 1); # so [-1] is the last element, like Perl lists
+	}
 
+	my $code = 200;
 	# double the limit for expanded views:
 	my $opts = {
-		limit => $LIM,
-		offset => $q->{o},
+		limit => $q->{l},
+		offset => $o,
 		mset => 1,
 		relevance => $q->{r},
+		asc => $asc,
 	};
-	my ($mset, $total);
+	my ($mset, $total, $err, $cb);
+retry:
 	eval {
-		$mset = $ctx->{srch}->query($q->{'q'}, $opts);
+		$mset = $srch->query($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, $ctx);
 		if ($x eq 't') {
 			$cb = mset_thread($ctx, $mset, $q);
 		} else {
@@ -59,18 +85,7 @@ sub sres_top_html {
 	PublicInbox::WwwStream->response($ctx, $code, $cb);
 }
 
-# allow undef for individual doc loads...
-sub load_doc_retry {
-	my ($srch, $mitem) = @_;
-
-	eval {
-		$srch->retry_reopen(sub {
-			PublicInbox::SearchMsg->load_doc($mitem->get_document)
-		});
-	}
-}
-
-# 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) = @_;
@@ -79,11 +94,13 @@ sub mset_summary {
 	my $pad = length("$total");
 	my $pfx = ' ' x $pad;
 	my $res = \($ctx->{-html_tip});
-	my $srch = $ctx->{srch};
+	my $ibx = $ctx->{-inbox};
+	my $srch = $ibx->search;
+	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;
-		my $smsg = load_doc_retry($srch, $m);
+		my $pct = get_pct($m);
+		my $smsg = PublicInbox::SearchMsg::from_mitem($m, $srch);
 		unless ($smsg) {
 			eval {
 				$m = "$m ".$m->get_docid . " expired\n";
@@ -93,29 +110,64 @@ sub mset_summary {
 		}
 		my $s = ascii_html($smsg->subject);
 		my $f = ascii_html($smsg->from_name);
-		my $ts = PublicInbox::View::fmt_ts($smsg->ts);
+		if ($obfs_ibx) {
+			obfuscate_addrs($obfs_ibx, $s);
+			obfuscate_addrs($obfs_ibx, $f);
+		}
+		my $date = PublicInbox::View::fmt_ts($smsg->ds);
 		my $mid = PublicInbox::Hval->new_msgid($smsg->mid)->{href};
+		$s = '(no subject)' if $s eq '';
 		$$res .= qq{$rank. <b><a\nhref="$mid/">}.
 			$s . "</a></b>\n";
-		$$res .= "$pfx  - by $f @ $ts UTC [$pct%]\n\n";
+		$$res .= "$pfx  - by $f @ $date UTC [$pct%]\n\n";
 	}
 	$$res .= search_nav_bot($mset, $q);
 	*noop;
 }
 
+# shorten "/full/path/to/Foo/Bar.pm" to "Foo/Bar.pm" so error
+# messages don't reveal FS layout info in case people use non-standard
+# installation paths
+sub path2inc ($) {
+	my $full = $_[0];
+	if (my $short = $rmap_inc{$full}) {
+		return $short;
+	} elsif (!scalar(keys %rmap_inc) && -e $full) {
+		%rmap_inc = map {; "$INC{$_}" => $_ } keys %INC;
+		# fall back to basename as last resort
+		$rmap_inc{$full} // (split('/', $full))[-1];
+	} else {
+		$full;
+	}
+}
+
 sub err_txt {
 	my ($ctx, $err) = @_;
 	my $u = $ctx->{-inbox}->base_url($ctx->{env}) . '_/text/help/';
 	$err =~ s/^\s*Exception:\s*//; # bad word to show users :P
+	$err =~ s!(\S+)!path2inc($1)!sge;
 	$err = ascii_html($err);
 	"\nBad query: <b>$err</b>\n" .
 		qq{See <a\nhref="$u">$u</a> for help on using search};
 }
 
 sub search_nav_top {
-	my ($mset, $q) = @_;
+	my ($mset, $q, $ctx) = @_;
+	my $m = $q->qs_html(x => 'm', r => undef);
+	my $rv = qq{<form\naction="?$m"\nmethod="post"><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>};
@@ -129,95 +181,137 @@ 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>]};
+	$rv .= qq{\n\t\t\t\t\t\tdownload: };
+	$rv .= qq{<input\ntype=submit\nvalue="mbox.gz"/></pre></form><pre>};
 }
 
 sub search_nav_bot {
 	my ($mset, $q) = @_;
 	my $total = $mset->get_matches_estimated;
-	my $nr = scalar $mset->items;
+	my $l = $q->{l};
+	my $rv = '</pre><hr><pre id=t>';
 	my $o = $q->{o};
-	my $end = $o + $nr;
-	my $beg = $o + 1;
-	my $rv = "</pre><hr><pre>Results $beg-$end of $total";
-	my $n = $o + $LIM;
-
-	if ($n < $total) {
-		my $qs = $q->qs_html(o => $n);
-		$rv .= qq{, <a\nhref="?$qs"\nrel=next>next</a>}
+	my $off = $o < 0 ? -($o + 1) : $o;
+	my $end = $off + $mset->size;
+	my $beg = $off + 1;
+
+	if ($beg <= $end) {
+		$rv .= "Results $beg-$end of $total";
+		$rv .= ' (estimated)' if $end != $total;
+	} else {
+		$rv .= "No more results, only $total";
 	}
-	if ($o > 0) {
-		$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>};
+	my ($next, $join, $prev);
+
+	if ($o >= 0) { # sort descending
+		my $n = $o + $l;
+		if ($n < $total) {
+			$next = $q->qs_html(o => $n, l => $l);
+		}
+		if ($o > 0) {
+			$join = $n < $total ? '/' : '       ';
+			my $p = $o - $l;
+			$prev = $q->qs_html(o => ($p > 0 ? $p : 0));
+		}
+	} else { # o < 0, sort ascending
+		my $n = $o - $l;
+
+		if (-$n < $total) {
+			$next = $q->qs_html(o => $n, l => $l);
+		}
+		if ($o < -1) {
+			$join = -$n < $total ? '/' : '       ';
+			my $p = $o + $l;
+			$prev = $q->qs_html(o => ($p < 0 ? $p : 0));
+		}
 	}
-	$rv .= '</pre>';
+
+	$rv .= qq{  <a\nhref="?$next"\nrel=next>next</a>} if $next;
+	$rv .= $join if $join;
+	$rv .= qq{<a\nhref="?$prev"\nrel=prev>prev</a>} if $prev;
+
+	my $rev = $q->qs_html(o => $o < 0 ? 0 : -1);
+	$rv .= qq{ | <a\nhref="?$rev">reverse results</a></pre>};
+}
+
+sub sort_relevance {
+	[ sort {
+		(eval { $b->topmost->{smsg}->{pct} } // 0) <=>
+		(eval { $a->topmost->{smsg}->{pct} } // 0)
+	} @{$_[0]} ]
+}
+
+sub get_pct ($) {
+	# Capped at "99%" since "100%" takes an extra column in the
+	# thread skeleton view.  <xapian/mset.h> says the value isn't
+	# very meaningful, anyways.
+	my $n = $_[0]->get_percent;
+	$n > 99 ? 99 : $n;
+}
+
+sub load_msgs {
+	my ($mset) = @_;
+	[ map {
+		my $mi = $_;
+		my $smsg = PublicInbox::SearchMsg::from_mitem($mi);
+		$smsg->{pct} = get_pct($mi);
+		$smsg;
+	} ($mset->items) ]
 }
 
 sub mset_thread {
 	my ($ctx, $mset, $q) = @_;
-	my %pct;
-	my $msgs = $ctx->{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 $th = PublicInbox::SearchThread->new($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 $msgs = $ctx->{-inbox}->search->retry_reopen(\&load_msgs, $mset);
+	my $r = $q->{r};
+	my $rootset = PublicInbox::SearchThread::thread($msgs,
+		$r ? \&sort_relevance : \&PublicInbox::View::sort_ds,
+		$ctx);
 	my $skel = search_nav_bot($mset, $q). "<pre>";
-	my $inbox = $ctx->{-inbox};
 	$ctx->{-upfx} = '';
 	$ctx->{anchor_idx} = 1;
 	$ctx->{cur_level} = 0;
-	$ctx->{dst} = \$skel;
+	$ctx->{skel} = \$skel;
 	$ctx->{mapping} = {};
-	$ctx->{pct} = \%pct;
+	$ctx->{searchview} = 1;
 	$ctx->{prev_attr} = '';
 	$ctx->{prev_level} = 0;
-	$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);
 
-	my $mime;
-	sub {
-		return unless $msgs;
-		while ($mime = shift @$msgs) {
-			$mime = $inbox->msg_by_smsg($mime) and last;
-		}
-		if ($mime) {
-			$mime = Email::MIME->new($mime);
-			return PublicInbox::View::index_entry($mime, $ctx,
-				scalar @$msgs);
-		}
-		$msgs = undef;
-		$skel .= "\n</pre>";
-	};
+	@$msgs = reverse @$msgs if $r;
+	$ctx->{msgs} = $msgs;
+	\&mset_thread_i;
+}
+
+# callback for PublicInbox::WwwStream::getline
+sub mset_thread_i {
+	my ($nr, $ctx) = @_;
+	my $msgs = $ctx->{msgs} or return;
+	while (my $smsg = pop @$msgs) {
+		$ctx->{-inbox}->smsg_mime($smsg) or next;
+		return PublicInbox::View::index_entry($smsg, $ctx,
+							scalar @$msgs);
+	}
+	my ($skel) = delete @$ctx{qw(skel msgs)};
+	$$skel .= "\n</pre>";
 }
 
 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";
@@ -232,34 +326,40 @@ sub ctx_prepare {
 
 sub adump {
 	my ($cb, $mset, $q, $ctx) = @_;
-	my $ibx = $ctx->{-inbox};
-	my @items = $mset->items;
-	$ctx->{search_query} = $q;
-	my $srch = $ctx->{srch};
-	PublicInbox::WwwAtomStream->response($ctx, 200, sub {
-		while (my $x = shift @items) {
-			$x = load_doc_retry($srch, $x);
-			$x = $ibx->msg_by_smsg($x) and
-					return Email::MIME->new($x);
-		}
-		return undef;
-	});
+	$ctx->{items} = [ $mset->items ];
+	$ctx->{search_query} = $q; # used by WwwAtomStream::atom_header
+	$ctx->{srch} = $ctx->{-inbox}->search;
+	PublicInbox::WwwAtomStream->response($ctx, 200, \&adump_i);
+}
+
+# callback for PublicInbox::WwwAtomStream::getline
+sub adump_i {
+	my ($ctx) = @_;
+	while (my $mi = shift @{$ctx->{items}}) {
+		my $smsg = eval {
+			PublicInbox::SearchMsg::from_mitem($mi, $ctx->{srch});
+		} or next;
+		$ctx->{-inbox}->smsg_mime($smsg) and return $smsg;
+	}
 }
 
 package PublicInbox::SearchQuery;
 use strict;
 use warnings;
-use PublicInbox::Hval;
-use PublicInbox::MID qw(mid_escape);
+use URI::Escape qw(uri_escape);
+use PublicInbox::MID qw(MID_ESC);
 
 sub new {
 	my ($class, $qp) = @_;
 
 	my $r = $qp->{r};
+	my ($l) = (($qp->{l} || '') =~ /([0-9]+)/);
+	$l = $LIM if !$l || $l > $LIM;
 	bless {
 		q => $qp->{'q'},
 		x => $qp->{x} || '',
-		o => (($qp->{o} || '0') =~ /(\d+)/),
+		o => (($qp->{o} || '0') =~ /(-?[0-9]+)/),
+		l => $l,
 		r => (defined $r && $r ne '0'),
 	}, $class;
 }
@@ -275,18 +375,21 @@ 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";
 
 	if (my $o = $self->{o}) { # ignore o == 0
 		$qs .= "&amp;o=$o";
 	}
+	if (my $l = $self->{l}) {
+		$qs .= "&amp;l=$l" unless $l == $LIM;
+	}
 	if (my $r = $self->{r}) {
 		$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;
 }