#
# 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::MID qw(mid2path mid_mime mid_clean);
-use Email::MIME;
+use PublicInbox::WwwAtomStream;
+use PublicInbox::MID qw(mid2path mid_mime mid_clean mid_escape MID_ESC);
+use PublicInbox::MIME;
require PublicInbox::Git;
-require PublicInbox::Thread;
-our $LIM = 50;
+require PublicInbox::SearchThread;
+our $LIM = 200;
sub noop {}
+sub mbox_results {
+ my ($ctx) = @_;
+ my $q = PublicInbox::SearchQuery->new($ctx->{qp});
+ my $x = $q->{x};
+ return PublicInbox::Mbox::mbox_all($ctx, $q->{'q'}) if $x eq 'm';
+ sres_top_html($ctx);
+}
+
sub sres_top_html {
my ($ctx) = @_;
my $q = PublicInbox::SearchQuery->new($ctx->{qp});
- my $code = 200;
+ my $x = $q->{x};
+ my $query = $q->{'q'};
+ my $code = 200;
# double the limit for expanded views:
my $opts = {
- limit => $LIM,
+ limit => $q->{l},
offset => $q->{o},
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;
- $ctx->{-html_tip} = ''.err_txt($ctx, $err).'
';
+ $ctx->{-html_tip} = ''.err_txt($ctx, $err).'
';
$cb = *noop;
} elsif ($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} = "\n[No results found]
";
+ $ctx->{-html_tip} = "\n[No results found]
";
$cb = *noop;
} 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';
- $ctx->{-html_tip} = search_nav_top($mset, $q) . "\n\n";
+ $ctx->{-html_tip} = search_nav_top($mset, $q, $ctx);
if ($x eq 't') {
$cb = mset_thread($ctx, $mset, $q);
} else {
@@ -58,7 +77,18 @@ sub sres_top_html {
PublicInbox::WwwStream->response($ctx, $code, $cb);
}
-# display non-threaded search results similar to what users expect from
+# allow undef for individual doc loads...
+sub load_doc_retry {
+ my ($srch, $mitem) = @_;
+
+ eval {
+ $srch->retry_reopen(sub {
+ PublicInbox::SearchMsg->load_doc($mitem->get_document)
+ });
+ }
+}
+
+# display non-nested search results similar to what users expect from
# regular WWW search engines:
sub mset_summary {
my ($ctx, $mset, $q) = @_;
@@ -67,14 +97,28 @@ 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 $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 $smsg = PublicInbox::SearchMsg->load_doc($m->get_document);
+ my $smsg = load_doc_retry($srch, $m);
+ unless ($smsg) {
+ eval {
+ $m = "$m ".$m->get_docid . " expired\n";
+ $ctx->{env}->{'psgi.errors'}->print($m);
+ };
+ next;
+ }
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)->as_href;
+ my $mid = PublicInbox::Hval->new_msgid($smsg->mid)->{href};
$$res .= qq{$rank. }.
$s . "\n";
$$res .= "$pfx - by $f @ $ts UTC [$pct%]\n\n";
@@ -85,18 +129,30 @@ sub mset_summary {
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
$err = ascii_html($err);
"\nBad query: $err\n" .
- qq{See $u for Xapian query syntax};
+ qq{See $u for help on using search};
}
sub search_nav_top {
- my ($mset, $q) = @_;
+ my ($mset, $q, $ctx) = @_;
+ my $m = $q->qs_html(x => 'm', r => undef);
+ my $rv = qq{};
}
sub search_nav_bot {
@@ -124,47 +182,54 @@ sub search_nav_bot {
my $total = $mset->get_matches_estimated;
my $nr = scalar $mset->items;
my $o = $q->{o};
+ my $l = $q->{l};
my $end = $o + $nr;
my $beg = $o + 1;
- my $rv = "
Results $beg-$end of $total";
- my $n = $o + $LIM;
+ my $rv = '
';
+ if ($beg <= $end) {
+ $rv .= "Results $beg-$end of $total";
+ $rv .= ' (estimated)' if $end != $total;
+ } else {
+ $rv .= "No more results, only $total";
+ }
+ my $n = $o + $l;
if ($n < $total) {
- my $qs = $q->qs_html(o => $n);
- $rv .= qq{, next}
+ my $qs = $q->qs_html(o => $n, l => $l);
+ $rv .= qq{ next}
}
if ($o > 0) {
- $rv .= $n < $total ? '/' : ', ';
- my $p = $o - $LIM;
+ $rv .= $n < $total ? '/' : ' ';
+ my $p = $o - $l;
my $qs = $q->qs_html(o => ($p > 0 ? $p : 0));
$rv .= qq{prev};
}
$rv .= '
';
}
+sub sort_relevance {
+ my ($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;
- my @m = map {
+ my $srch = $ctx->{srch};
+ my $msgs = $srch->retry_reopen(sub { [ 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);
- }
+ 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,
+ $r ? sort_relevance(\%pct) : *PublicInbox::View::sort_ts,
+ $srch);
my $skel = search_nav_bot($mset, $q). "";
my $inbox = $ctx->{-inbox};
$ctx->{-upfx} = '';
@@ -176,21 +241,22 @@ sub mset_thread {
$ctx->{prev_attr} = '';
$ctx->{prev_level} = 0;
$ctx->{seen} = {};
- $ctx->{s_nr} = scalar(@m).'+ results';
+ $ctx->{s_nr} = scalar(@$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 $msgs = \@m;
+ @$msgs = reverse @$msgs if $r;
my $mime;
sub {
return unless $msgs;
- while ($mime = shift @$msgs) {
- my $mid = mid_clean(mid_mime($mime));
- $mime = $inbox->msg_by_mid($mid) and last;
+ 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);
}
@@ -201,7 +267,9 @@ 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";
@@ -216,38 +284,40 @@ sub ctx_prepare {
sub adump {
my ($cb, $mset, $q, $ctx) = @_;
- my $fh = $cb->([ 200, ['Content-Type' => 'application/atom+xml']]);
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 $s;
- }
- PublicInbox::Feed::end_feed($fh);
+ 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 PublicInbox::MIME->new($x);
+ }
+ return undef;
+ });
}
package PublicInbox::SearchQuery;
use strict;
use warnings;
+use URI::Escape qw(uri_escape);
use PublicInbox::Hval;
+use PublicInbox::MID qw(MID_ESC);
sub new {
my ($class, $qp) = @_;
my $r = $qp->{r};
+ my $l = $qp->{l} || '200';
+ if (! ($l =~ /(\d+)/ && $l <= $LIM)) {
+ $l = $LIM;
+ }
bless {
q => $qp->{'q'},
x => $qp->{x} || '',
o => (($qp->{o} || '0') =~ /(\d+)/),
+ l => $l,
r => (defined $r && $r ne '0'),
}, $class;
}
@@ -263,18 +333,21 @@ sub qs_html {
$self = $tmp;
}
- my $q = PublicInbox::Hval->new($self->{'q'})->as_href;
+ 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
$qs .= "&o=$o";
}
+ if (my $l = $self->{l}) {
+ $qs .= "&l=$l";
+ }
if (my $r = $self->{r}) {
$qs .= "&r";
}
if (my $x = $self->{x}) {
- $qs .= "&x=$x" if ($x eq 't' || $x eq 'A');
+ $qs .= "&x=$x" if ($x eq 't' || $x eq 'A' || $x eq 'm');
}
$qs;
}