X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FSearchView.pm;h=892e8fdae14038098bc9cd9ae21f5dcd97b2a1fc;hb=68f9da93b11412122fff6e6154a42ecabcdbada0;hp=fbef4116e9623f2838f17cf0c4fc31c9ff31acaf;hpb=5c5dfb7d5558ff248477fa57aef08e2a8bbd9800;p=public-inbox.git

diff --git a/lib/PublicInbox/SearchView.pm b/lib/PublicInbox/SearchView.pm
index fbef4116..892e8fda 100644
--- a/lib/PublicInbox/SearchView.pm
+++ b/lib/PublicInbox/SearchView.pm
@@ -1,98 +1,187 @@
-# Copyright (C) 2015 all contributors <meta@public-inbox.org>
-# License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt)
+# Copyright (C) 2015-2020 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 PublicInbox::SearchMsg;
-use PublicInbox::Hval qw/ascii_html/;
+use v5.10.1;
+use List::Util qw(min max);
+use URI::Escape qw(uri_unescape);
+use PublicInbox::Smsg;
+use PublicInbox::Hval qw(ascii_html obfuscate_addrs mid_href fmt_ts);
 use PublicInbox::View;
-use PublicInbox::MID qw(mid2path mid_mime);
-use Email::MIME;
-require PublicInbox::Git;
-require PublicInbox::Thread;
-our $LIM = 50;
+use PublicInbox::WwwAtomStream;
+use PublicInbox::WwwStream qw(html_oneshot);
+use PublicInbox::SearchThread;
+use PublicInbox::SearchQuery;
+use PublicInbox::Search qw(mdocid);
+my %rmap_inc;
+
+sub mbox_results {
+	my ($ctx) = @_;
+	my $q = PublicInbox::SearchQuery->new($ctx->{qp});
+	if ($ctx->{env}->{'psgi.input'}->read(my $buf, 3)) {
+		$q->{t} = 1 if $buf =~ /\Ax=[^0]/;
+	}
+	require PublicInbox::Mbox;
+	$q->{x} eq 'm' ? PublicInbox::Mbox::mbox_all($ctx, $q) :
+			sres_top_html($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},
+		thread => $q->{t},
+		asc => $asc,
 	};
-	my ($mset, $total);
-
+	my ($mset, $total, $err, $html);
+retry:
 	eval {
-		$mset = $ctx->{srch}->query($q->{q}, $opts);
+		$mset = $srch->query($query, $opts);
 		$total = $mset->get_matches_estimated;
 	};
-	my $err = $@;
-	my $res = html_start($q, $ctx) . '<pre>';
+	$err = $@;
+	ctx_prepare($q, $ctx);
 	if ($err) {
 		$code = 400;
-		$res .= err_txt($ctx, $err) . "</pre><hr /><pre>" . foot($ctx);
+		$html = '<pre>'.err_txt($ctx, $err).'</pre><hr>';
 	} elsif ($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;
-		$res .= "\n\n[No results found]</pre><hr /><pre>".foot($ctx);
+		$html = "<pre>\n[No results found]</pre><hr>";
 	} else {
-		my $x = $q->{x};
-		return sub { adump($_[0], $mset, $q, $ctx) } if ($x eq 'A');
+		return adump($_[0], $mset, $q, $ctx) if $x eq 'A';
 
-		$res .= search_nav_top($mset, $q) . "\n\n";
-		if ($x eq 't') {
-			return sub { tdump($_[0], $res, $mset, $q, $ctx) };
-		}
-		dump_mset(\$res, $mset);
-		$res .= '</pre>' . search_nav_bot($mset, $q) .
-			"\n\n" . foot($ctx);
+		$ctx->{-html_tip} = search_nav_top($mset, $q, $ctx);
+		return mset_thread($ctx, $mset, $q) if $x eq 't';
+		mset_summary($ctx, $mset, $q); # appends to {-html_tip}
+		$html = '';
 	}
-
-	$res .= "</pre></body></html>";
-	[$code, ['Content-Type'=>'text/html; charset=UTF-8'], [$res]];
+	html_oneshot($ctx, $code);
 }
 
-# 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 dump_mset {
-	my ($res, $mset) = @_;
+sub mset_summary {
+	my ($ctx, $mset, $q) = @_;
 
 	my $total = $mset->get_matches_estimated;
 	my $pad = length("$total");
 	my $pfx = ' ' x $pad;
+	my $res = \($ctx->{-html_tip});
+	my $ibx = $ctx->{-inbox};
+	my $obfs_ibx = $ibx->{obfuscate} ? $ibx : undef;
+	my @nums = @{$ibx->search->mset_to_artnums($mset)};
+	my %num2msg = map { $_->{num} => $_ } @{$ibx->over->get_all(@nums)};
+	my ($min, $max);
+
 	foreach my $m ($mset->items) {
 		my $rank = sprintf("%${pad}d", $m->get_rank + 1);
-		my $pct = $m->get_percent;
-		my $smsg = PublicInbox::SearchMsg->load_doc($m->get_document);
-		my $s = ascii_html($smsg->subject);
-		my $f = ascii_html($smsg->from_name);
-		my $ts = PublicInbox::View::fmt_ts($smsg->ts);
-		my $mid = PublicInbox::Hval->new_msgid($smsg->mid)->as_href;
+		my $pct = get_pct($m);
+		my $num = shift @nums;
+		my $smsg = delete($num2msg{$num}) or do {
+			eval {
+				$m = "$m $num expired\n";
+				$ctx->{env}->{'psgi.errors'}->print($m);
+			};
+			next;
+		};
+		$ctx->{-t_max} //= $smsg->{ts};
+
+		# only when sorting by relevance, ->items is always
+		# ordered descending:
+		$max //= $pct;
+		$min = $pct;
+
+		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 $date = fmt_ts($smsg->{ds});
+		my $mid = mid_href($smsg->{mid});
+		$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";
+		$$res .= "$pfx  - by $f @ $date UTC [$pct%]\n\n";
+	}
+	if ($q->{r}) { # for descriptions in search_nav_bot
+		$q->{-min_pct} = $min;
+		$q->{-max_pct} = $max;
+	}
+	$$res .= search_nav_bot($mset, $q);
+	undef;
+}
+
+# shorten "/full/path/to/Foo/Bar.pm" to "Foo/Bar.pm" so error
+# messages don't reveal FS layout info in case people use non-standard
+# installation paths
+sub path2inc ($) {
+	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 = '//xapian.org/docs/queryparser.html';
-	$u = PublicInbox::Hval::prurl($ctx->{cgi}->{env}, $u);
+	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);
-	"\n\nBad query: <b>$err</b>\n" .
-		qq{See <a\nhref="$u">$u</a> for Xapian query syntax};
+	"\nBad query: <b>$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, t => 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 = "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>};
@@ -106,180 +195,178 @@ 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>]};
+	if ($ctx->{-inbox}->search->has_threadid) {
+		$rv .= qq{\n\t\t\tdownload mbox.gz: } .
+			# we set name=z w/o using it since it seems required for
+			# lynx (but works fine for w3m).
+			qq{<input\ntype=submit\nname=z\n} .
+				q{value="results only"/>} .
+			qq{|<input\ntype=submit\nname=x\n} .
+				q{value="full threads"/>};
+	} else { # BOFH needs to --reindex
+		$rv .= qq{\n\t\t\t\t\t\tdownload: } .
+			qq{<input\ntype=submit\nname=z\nvalue="mbox.gz"/>}
+	}
+	$rv .= qq{</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 = "<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>}
-	}
-	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>};
-	}
-	$rv;
-}
+	my $off = $o < 0 ? -($o + 1) : $o;
+	my $end = $off + $mset->size;
+	my $beg = $off + 1;
 
-sub tdump {
-	my ($cb, $res, $mset, $q, $ctx) = @_;
-	my $fh = $cb->([200, ['Content-Type'=>'text/html; charset=UTF-8']]);
-	$fh->write($res .= '</pre>');
-	my %pct;
-	my @m = map {
-		my $i = $_;
-		my $m = PublicInbox::SearchMsg->load_doc($i->get_document);
-		$pct{$m->mid} = $i->get_percent;
-		$m = $m->mini_mime;
-		$m;
-	} ($mset->items);
-
-	my $th = PublicInbox::Thread->new(@m);
-	$th->thread;
-	if ($q->{r}) { # order by relevance
-		$th->order(sub {
-			sort { (eval { $pct{$b->topmost->messageid} } || 0)
-					<=>
-				(eval { $pct{$a->topmost->messageid} } || 0)
-			} @_;
-		});
-	} else { # order by time (default for threaded view)
-		$th->order(*PublicInbox::View::sort_ts);
+	if ($beg <= $end) {
+		my $approx = $end == $total ? '' : '~';
+		$rv .= "Results $beg-$end of $approx$total";
+	} else {
+		$rv .= "No more results, only $total";
 	}
-	my $skel = '';
-	my $state = {
-		-inbox => $ctx->{-inbox},
-		anchor_idx => 1,
-		ctx => $ctx,
-		cur_level => 0,
-		dst => \$skel,
-		fh => $fh,
-		mapping => {},
-		pct => \%pct,
-		prev_attr => '',
-		prev_level => 0,
-		seen => {},
-		srch => $ctx->{srch},
-		upfx => './',
-	};
-	$ctx->{searchview} = 1;
-	PublicInbox::View::walk_thread($th, $state,
-		*PublicInbox::View::pre_thread);
+	my ($next, $join, $prev, $nd, $pd);
 
-	PublicInbox::View::thread_entry($state, $_, 0) for @m;
+	if ($o >= 0) { # sort descending
+		my $n = $o + $l;
+		if ($n < $total) {
+			$next = $q->qs_html(o => $n, l => $l);
+			$nd = $q->{r} ? "[&lt;= $q->{-min_pct}%]" : '(older)';
+		}
+		if ($o > 0) {
+			$join = $n < $total ? ' | ' : "\t";
+			my $p = $o - $l;
+			$prev = $q->qs_html(o => ($p > 0 ? $p : 0));
+			$pd = $q->{r} ? "[&gt;= $q->{-max_pct}%]" : '(newer)';
+		}
+	} else { # o < 0, sort ascending
+		my $n = $o - $l;
 
-	$fh->write(search_nav_bot($mset, $q). "\n\n" . $skel . "\n" .
-			foot($ctx). '</pre></body></html>');
+		if (-$n < $total) {
+			$next = $q->qs_html(o => $n, l => $l);
+			$nd = $q->{r} ? "[&lt;= $q->{-min_pct}%]" : '(newer)';
+		}
+		if ($o < -1) {
+			$join = -$n < $total ? ' | ' : "\t";
+			my $p = $o + $l;
+			$prev = $q->qs_html(o => ($p < 0 ? $p : 0));
+			$pd = $q->{r} ? "[&gt;= $q->{-max_pct}%]" : '(older)';
+		}
+	}
 
-	$fh->close;
-}
+	$rv .= qq{  <a\nhref="?$next"\nrel=next>next $nd</a>} if $next;
+	$rv .= $join if $join;
+	$rv .= qq{<a\nhref="?$prev"\nrel=prev>prev $pd</a>} if $prev;
 
-sub foot {
-	my ($ctx) = @_;
-	my $foot = $ctx->{footer} || '';
-	qq{Back to <a\nhref=".">index</a>.\n$foot};
+	my $rev = $q->qs_html(o => $o < 0 ? 0 : -1);
+	$rv .= qq{ | <a\nhref="?$rev">reverse</a></pre>};
 }
 
-sub html_start {
-	my ($q, $ctx) = @_;
-	my $qh = ascii_html($q->{'q'});
-	my $A = $q->qs_html(x => 'A', r => undef);
-	my $res = '<html><head>' . PublicInbox::Hval::STYLE .
-		"<title>$qh - search results</title>" .
-		qq{<link\nrel=alternate\ntitle="Atom feed"\n} .
-		qq!href="?$A"\ntype="application/atom+xml"/></head>! .
-		qq{<body><form\naction="">} .
-		qq{<input\nname=q\nvalue="$qh"\ntype=text />};
-
-	$res .= qq{<input\ntype=hidden\nname=r />} if $q->{r};
-	if (my $x = $q->{x}) {
-		$x = ascii_html($x);
-		$res .= qq{<input\ntype=hidden\nname=x\nvalue="$x" />};
-	}
+sub sort_relevance {
+	[ sort {
+		(eval { $b->topmost->{pct} } // 0) <=>
+		(eval { $a->topmost->{pct} } // 0)
+	} @{$_[0]} ]
+}
 
-	$res .= qq{<input\ntype=submit\nvalue=search /></form>};
+sub get_pct ($) {
+	# 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 adump {
-	my ($cb, $mset, $q, $ctx) = @_;
-	my $fh = $cb->([ 200, ['Content-Type' => 'application/atom+xml']]);
+sub mset_thread {
+	my ($ctx, $mset, $q) = @_;
 	my $ibx = $ctx->{-inbox};
-	my $feed_opts = PublicInbox::Feed::get_feedopts($ctx);
-	my $x = ascii_html($q->{'q'});
-	$x = qq{$x - search results};
-	$feed_opts->{atomurl} = $feed_opts->{url} . '?'. $q->qs_html;
-	$feed_opts->{url} .= '?'. $q->qs_html(x => undef);
-	$x = PublicInbox::Feed::atom_header($feed_opts, $x);
-	$fh->write($x. PublicInbox::Feed::feed_updated());
-
-	for ($mset->items) {
-		$x = PublicInbox::SearchMsg->load_doc($_->get_document)->mid;
-		$x = mid2path($x);
-		my $s = PublicInbox::Feed::feed_entry($feed_opts, $x, $ibx);
-		$fh->write($s) if defined $s;
+	my $nshard = $ibx->search->{nshard} // 1;
+	my %pct = map { mdocid($nshard, $_) => get_pct($_) } $mset->items;
+	my $msgs = $ibx->over->get_all(keys %pct);
+	$_->{pct} = $pct{$_->{num}} for @$msgs;
+	my $r = $q->{r};
+	if ($r) { # for descriptions in search_nav_bot
+		my @pct = values %pct;
+		$q->{-min_pct} = min(@pct);
+		$q->{-max_pct} = max(@pct);
 	}
-	PublicInbox::Feed::end_feed($fh);
+	my $rootset = PublicInbox::SearchThread::thread($msgs,
+		$r ? \&sort_relevance : \&PublicInbox::View::sort_ds,
+		$ctx);
+	my $skel = search_nav_bot($mset, $q). "<pre>";
+	$ctx->{-upfx} = '';
+	$ctx->{anchor_idx} = 1;
+	$ctx->{cur_level} = 0;
+	$ctx->{skel} = \$skel;
+	$ctx->{mapping} = {};
+	$ctx->{searchview} = 1;
+	$ctx->{prev_attr} = '';
+	$ctx->{prev_level} = 0;
+	$ctx->{s_nr} = scalar(@$msgs).'+ results';
+
+	# reduce hash lookups in skel_dump
+	$ctx->{-obfs_ibx} = $ibx->{obfuscate} ? $ibx : undef;
+	PublicInbox::View::walk_thread($rootset, $ctx,
+		\&PublicInbox::View::pre_thread);
+
+	# link $INBOX_DIR/description text to "recent" view around
+	# the newest message in this result set:
+	$ctx->{-t_max} = max(map { delete $_->{ts} } @$msgs);
+
+	@$msgs = reverse @$msgs if $r;
+	$ctx->{msgs} = $msgs;
+	PublicInbox::WwwStream::aresponse($ctx, 200, \&mset_thread_i);
 }
 
-package PublicInbox::SearchQuery;
-use strict;
-use warnings;
-use PublicInbox::Hval;
-
-sub new {
-	my ($class, $qp) = @_;
-
-	my $r = $qp->{r};
-	bless {
-		q => $qp->{'q'},
-		x => $qp->{x} || '',
-		o => (($qp->{o} || '0') =~ /(\d+)/),
-		r => (defined $r && $r ne '0'),
-	}, $class;
+# callback for PublicInbox::WwwStream::getline
+sub mset_thread_i {
+	my ($ctx, $eml) = @_;
+	$ctx->zmore($ctx->html_top) if exists $ctx->{-html_tip};
+	$eml and return PublicInbox::View::eml_entry($ctx, $eml);
+	my $smsg = shift @{$ctx->{msgs}} or
+		$ctx->zmore(${delete($ctx->{skel})});
+	$smsg;
 }
 
-sub qs_html {
-	my ($self, %over) = @_;
-
-	if (keys %over) {
-		my $tmp = bless { %$self }, ref($self);
-		foreach my $k (keys %over) {
-			$tmp->{$k} = $over{$k};
-		}
-		$self = $tmp;
+sub ctx_prepare {
+	my ($q, $ctx) = @_;
+	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";
+	my $extra = '';
+	$extra .= qq{<input\ntype=hidden\nname=r />} if $q->{r};
+	if (my $x = $q->{x}) {
+		$x = ascii_html($x);
+		$extra .= qq{<input\ntype=hidden\nname=x\nvalue="$x" />};
 	}
+	$ctx->{-extra_form_html} = $extra;
+}
 
-	my $q = PublicInbox::Hval->new($self->{'q'})->as_href;
-	$q =~ s/%20/+/g; # improve URL readability
-	my $qs = "q=$q";
+sub adump {
+	my ($cb, $mset, $q, $ctx) = @_;
+	$ctx->{ids} = $ctx->{-inbox}->search->mset_to_artnums($mset);
+	$ctx->{search_query} = $q; # used by WwwAtomStream::atom_header
+	PublicInbox::WwwAtomStream->response($ctx, 200, \&adump_i);
+}
 
-	if (my $o = $self->{o}) { # ignore o == 0
-		$qs .= "&amp;o=$o";
-	}
-	if (my $r = $self->{r}) {
-		$qs .= "&amp;r";
-	}
-	if (my $x = $self->{x}) {
-		$qs .= "&amp;x=$x" if ($x eq 't' || $x eq 'A');
+# callback for PublicInbox::WwwAtomStream::getline
+sub adump_i {
+	my ($ctx) = @_;
+	while (my $num = shift @{$ctx->{ids}}) {
+		my $smsg = eval { $ctx->{-inbox}->over->get_art($num) } or next;
+		return $smsg;
 	}
-	$qs;
 }
 
 1;