+#
+# 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;
+use PublicInbox::Hval qw/ascii_html obfuscate_addrs/;
use PublicInbox::View;
-use POSIX qw/strftime/;
-our $LIM = 25;
+use PublicInbox::WwwAtomStream;
+use PublicInbox::MID qw(MID_ESC);
+use PublicInbox::MIME;
+require PublicInbox::Git;
+require PublicInbox::SearchThread;
+our $LIM = 200;
-sub sres_top_html {
- my ($ctx, $q) = @_;
- my $cgi = $ctx->{cgi};
- # $q ||= $cgi->param('q');
- my $o = int($cgi->param('o') || 0);
- my $r = $cgi->param('r');
- $r = (defined $r && $r ne '0');
- my $opts = { limit => $LIM, offset => $o, mset => 1, relevance => $r };
- my $mset = $ctx->{srch}->query($q, $opts);
- my $total = $mset->get_matches_estimated;
- my $query = PublicInbox::Hval->new_oneline($q);
- my $qh = $query->as_html;
- my $res = "$qh - search results" .
- qq{} .
- PublicInbox::View::PRE_WRAP;
+sub sres_top_html {
+ my ($ctx) = @_;
+ my $q = PublicInbox::SearchQuery->new($ctx->{qp});
+ my $x = $q->{x};
+ my $query = $q->{'q'};
- my $foot = $ctx->{footer} || '';
- $foot = qq{Back to index.};
- if ($total == 0) {
- $res .= "\n\n[No results found]
$foot";
+ my $code = 200;
+ # double the limit for expanded views:
+ my $opts = {
+ limit => $LIM,
+ offset => $q->{o},
+ mset => 1,
+ relevance => $q->{r},
+ };
+ my ($mset, $total, $err, $cb);
+retry:
+ eval {
+ $mset = $ctx->{srch}->query($query, $opts);
+ $total = $mset->get_matches_estimated;
+ };
+ $err = $@;
+ ctx_prepare($q, $ctx);
+ if ($err) {
+ $code = 400;
+ $ctx->{-html_tip} = ''.err_txt($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]
";
+ $cb = *noop;
} else {
- $q = $query->as_href;
- $q =~ s/%20/+/g; # improve URL readability
- my $qp = "?q=$q";
- $qp .= "&o=$o" if $o;
-
- $res .= "Search results ordered by [";
- if ($r) {
- $res .= qq{date|relevance};
+ return adump($_[0], $mset, $q, $ctx) if $x eq 'A';
+
+ $ctx->{-html_tip} = search_nav_top($mset, $q, $ctx);
+ if ($x eq 't') {
+ $cb = mset_thread($ctx, $mset, $q);
} else {
- $qp .= '&r';
- $res .= qq{date|relevance};
- }
- $res .= "]\n\n";
-
- dump_mset(\$res, $mset);
- my $nr = scalar $mset->items;
- my $end = $o + $nr;
- my $beg = $o + 1;
- $res .= "
";
- $res .= "Results $beg-$end of $total";
-
- my $n = $o + $LIM;
- if ($n < $total) {
- $qp = "q=$q&o=$n";
- $qp .= "&r" if $r;
- $res .= qq{, next}
+ $cb = mset_summary($ctx, $mset, $q);
}
- if ($o > 0) {
- $res .= $n < $total ? '/' : ', ';
- my $p = $o - $LIM;
- $qp = "q=$q";
- $qp .= "&o=$p" if $p > 0;
- $qp .= "&r" if $r;
- $res .= qq{prev};
- }
- $res .= "\n\n" . $foot;
}
+ PublicInbox::WwwStream->response($ctx, $code, $cb);
+}
- $res .= "
";
- [200, ['Content-Type'=>'text/html; charset=UTF-8'], [$res]];
+# 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)
+ });
+ }
}
-sub dump_mset {
- my ($res, $mset) = @_;
+# display non-nested search results similar to what users expect from
+# regular WWW search engines:
+sub mset_summary {
+ my ($ctx, $mset, $q) = @_;
my $total = $mset->get_matches_estimated;
my $pad = length("$total");
my $pfx = ' ' x $pad;
+ my $res = \($ctx->{-html_tip});
+ my $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 $s = PublicInbox::Hval->new_oneline($smsg->subject);
- my $f = $smsg->from_name;
- $f = PublicInbox::Hval->new_oneline($f)->as_html;
- my $d = strftime('%Y-%m-%d %H:%M', gmtime($smsg->ts));
- my $mid = PublicInbox::Hval->new_msgid($smsg->mid)->as_href;
- $$res .= qq{$rank. }.
- $s->as_html . "\n";
- $$res .= "$pfx - by $f @ $d UTC [$pct%]\n\n";
+ 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 $date = PublicInbox::View::fmt_ts($smsg->ds);
+ my $mid = PublicInbox::Hval->new_msgid($smsg->mid)->{href};
+ $$res .= qq{$rank. }.
+ $s . "\n";
+ $$res .= "$pfx - by $f @ $date UTC [$pct%]\n\n";
+ }
+ $$res .= search_nav_bot($mset, $q);
+ *noop;
+}
+
+sub err_txt {
+ my ($ctx, $err) = @_;
+ my $u = $ctx->{-inbox}->base_url($ctx->{env}) . '_/text/help/';
+ $err =~ s/^\s*Exception:\s*//; # bad word to show users :P
+ $err = ascii_html($err);
+ "\nBad query: $err\n" .
+ qq{See $u for help on using search};
+}
+
+sub search_nav_top {
+ my ($mset, $q, $ctx) = @_;
+ my $m = $q->qs_html(x => 'm', r => undef);
+ my $rv = qq{
};
+}
+
+sub search_nav_bot {
+ my ($mset, $q) = @_;
+ my $total = $mset->get_matches_estimated;
+ my $o = $q->{o};
+ my $end = $o + $mset->size;
+ my $beg = $o + 1;
+ 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 + $LIM;
+
+ if ($n < $total) {
+ my $qs = $q->qs_html(o => $n);
+ $rv .= qq{ next}
+ }
+ if ($o > 0) {
+ $rv .= $n < $total ? '/' : ' ';
+ my $p = $o - $LIM;
+ 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 $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,
+ $r ? sort_relevance(\%pct) : *PublicInbox::View::sort_ds,
+ $ctx);
+ my $skel = search_nav_bot($mset, $q). "";
+ my $inbox = $ctx->{-inbox};
+ $ctx->{-upfx} = '';
+ $ctx->{anchor_idx} = 1;
+ $ctx->{cur_level} = 0;
+ $ctx->{dst} = \$skel;
+ $ctx->{mapping} = {};
+ $ctx->{pct} = \%pct;
+ $ctx->{prev_attr} = '';
+ $ctx->{prev_level} = 0;
+ $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;
+ sub {
+ return unless $msgs;
+ my $smsg;
+ while (my $m = pop @$msgs) {
+ $smsg = $inbox->smsg_mime($m) and last;
+ }
+ if ($smsg) {
+ return PublicInbox::View::index_entry($smsg, $ctx,
+ scalar @$msgs);
+ }
+ $msgs = undef;
+ $skel .= "\n
";
+ };
+}
+
+sub ctx_prepare {
+ my ($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{} if $q->{r};
+ if (my $x = $q->{x}) {
+ $x = ascii_html($x);
+ $extra .= qq{};
+ }
+ $ctx->{-extra_form_html} = $extra;
+}
+
+sub adump {
+ my ($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 = $ibx->smsg_mime($x) and return $x;
+ }
+ return undef;
+ });
+}
+
+package PublicInbox::SearchQuery;
+use strict;
+use warnings;
+use URI::Escape qw(uri_escape);
+use PublicInbox::Hval;
+use PublicInbox::MID qw(MID_ESC);
+
+sub new {
+ my ($class, $qp) = @_;
+
+ my $r = $qp->{r};
+ bless {
+ q => $qp->{'q'},
+ x => $qp->{x} || '',
+ o => (($qp->{o} || '0') =~ /(\d+)/),
+ r => (defined $r && $r ne '0'),
+ }, $class;
+}
+
+sub qs_html {
+ my ($self, %over) = @_;
+
+ if (keys %over) {
+ my $tmp = bless { %$self }, ref($self);
+ foreach my $k (keys %over) {
+ $tmp->{$k} = $over{$k};
+ }
+ $self = $tmp;
+ }
+
+ 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 $r = $self->{r}) {
+ $qs .= "&r";
+ }
+ if (my $x = $self->{x}) {
+ $qs .= "&x=$x" if ($x eq 't' || $x eq 'A' || $x eq 'm');
}
+ $qs;
}
1;