[PATCH 0/2] improve "&x=t" search results

[PATCH 0/2] improve "&x=t" search results

[Eric Wong]

We weren't using SearchThread::thread correctly in SearchView,
so make accept SearchThread::thread the WWW $ctx arg, instead
(future changes along those lines on the horizon).

We now show the percentage relevance in the thread skeleton.

This is not bringing us closer to "mairix -t", yet.  But it
could be expanded into something in-between "mairix -t" behavior
and regular "mairix" behavior.

Eric Wong (2):
  searchthread: fix usage of user-supplied parameter
  view: show percentage in search results thread skeleton

 lib/PublicInbox/SearchThread.pm |  8 +++++---
 lib/PublicInbox/SearchView.pm   |  2 +-
 lib/PublicInbox/View.pm         | 23 ++++++++++++++++-------
 t/psgi_v2.t                     |  2 +-
 4 files changed, 23 insertions(+), 12 deletions(-)

[PATCH 1/2] searchthread: fix usage of user-supplied parameter

[Eric Wong]

Instead of only passing an Inbox object, we'll pass the $ctx
reference as PublicInbox::SearchView::mset_thread did.

So although mset_thread was wrong, we now make it's usage
of SearchThread::thread correct and update other callers to
favor the new style of passing the entire $ctx (with ->{-inbox})
instead of just the Inbox object.

This makes the thread skeleton at the bottom of the search
page to show subjects of messages, but unfortunately links to
non-existent #anchors.  The next commit will fix that.

While we're at it, favor "\&foo" over "*foo" since the former
makes the code reference (aka "function pointer) obvious so it
won't be confused for other things named "foo" in that
scope (e.g. $foo/@foo/%foo).
---
 lib/PublicInbox/SearchThread.pm | 8 +++++---
 lib/PublicInbox/SearchView.pm   | 2 +-
 lib/PublicInbox/View.pm         | 3 +--
 3 files changed, 7 insertions(+), 6 deletions(-)

diff --git a/lib/PublicInbox/SearchThread.pm b/lib/PublicInbox/SearchThread.pm
index 931bd579..ab2f1a84 100644
--- a/lib/PublicInbox/SearchThread.pm
+++ b/lib/PublicInbox/SearchThread.pm
@@ -22,7 +22,7 @@ use strict;
 use warnings;
 
 sub thread {
-	my ($msgs, $ordersub, $ibx) = @_;
+	my ($msgs, $ordersub, $ctx) = @_;
 	my $id_table = {};
 
 	# Sadly, we sort here anyways since the fill-in-the-blanks References:
@@ -32,12 +32,13 @@ sub thread {
 	# We'll trust the client Date: header here instead of the Received:
 	# time since this is for display (and not retrieval)
 	_add_message($id_table, $_) for sort { $a->{ds} <=> $b->{ds} } @$msgs;
+	my $ibx = $ctx->{-inbox};
 	my $rootset = [ grep {
 			!delete($_->{parent}) && $_->visible($ibx)
 		} values %$id_table ];
 	$id_table = undef;
 	$rootset = $ordersub->($rootset);
-	$_->order_children($ordersub, $ibx) for @$rootset;
+	$_->order_children($ordersub, $ctx) for @$rootset;
 	$rootset;
 }
 
@@ -151,10 +152,11 @@ sub visible ($$) {
 }
 
 sub order_children {
-	my ($cur, $ordersub, $ibx) = @_;
+	my ($cur, $ordersub, $ctx) = @_;
 
 	my %seen = ($cur => 1); # self-referential loop prevention
 	my @q = ($cur);
+	my $ibx = $ctx->{-inbox};
 	while (defined($cur = shift @q)) {
 		my $c = $cur->{children}; # The hashref here...
 
diff --git a/lib/PublicInbox/SearchView.pm b/lib/PublicInbox/SearchView.pm
index 78f2bd8b..566808e1 100644
--- a/lib/PublicInbox/SearchView.pm
+++ b/lib/PublicInbox/SearchView.pm
@@ -276,7 +276,7 @@ sub mset_thread {
 	} ($mset->items) ]});
 	my $r = $q->{r};
 	my $rootset = PublicInbox::SearchThread::thread($msgs,
-		$r ? sort_relevance(\%pct) : *PublicInbox::View::sort_ds,
+		$r ? sort_relevance(\%pct) : \&PublicInbox::View::sort_ds,
 		$ctx);
 	my $skel = search_nav_bot($mset, $q). "<pre>";
 	$ctx->{-upfx} = '';
diff --git a/lib/PublicInbox/View.pm b/lib/PublicInbox/View.pm
index 39b04174..33f71990 100644
--- a/lib/PublicInbox/View.pm
+++ b/lib/PublicInbox/View.pm
@@ -881,8 +881,7 @@ sub strict_loose_note ($) {
 sub thread_results {
 	my ($ctx, $msgs) = @_;
 	require PublicInbox::SearchThread;
-	my $ibx = $ctx->{-inbox};
-	my $rootset = PublicInbox::SearchThread::thread($msgs, *sort_ds, $ibx);
+	my $rootset = PublicInbox::SearchThread::thread($msgs, \&sort_ds, $ctx);
 
 	# FIXME: `tid' is broken on --reindex, so that needs to be fixed
 	# and preserved in the future.  This bug is hidden by `sid' matches

[PATCH 2/2] view: show percentage in search results thread skeleton

[Eric Wong]

The displays the Xapian ->get_percent value in the skeleton to
improve scanning of relevancy; irrelevant results do not display
that.

This fixes broken #anchor links introduced in the previous
commit, irrelevant messages now link to the /$INBOX/$MESSAGE_ID page.
---
 lib/PublicInbox/View.pm | 20 +++++++++++++++-----
 t/psgi_v2.t             |  2 +-
 2 files changed, 16 insertions(+), 6 deletions(-)

diff --git a/lib/PublicInbox/View.pm b/lib/PublicInbox/View.pm
index 33f71990..5924a2a7 100644
--- a/lib/PublicInbox/View.pm
+++ b/lib/PublicInbox/View.pm
@@ -950,7 +950,18 @@ sub skel_dump {
 	my $obfs_ibx = $ctx->{-obfs_ibx};
 	obfuscate_addrs($obfs_ibx, $f) if $obfs_ibx;
 
-	my $d = fmt_ts($smsg->{ds}) . ' ' . indent_for($level) . th_pfx($level);
+	my $d = fmt_ts($smsg->{ds});
+	my $unmatched; # if lazy-loaded by SearchThread::Msg::visible()
+	if (my $pct = $ctx->{pct}) {
+		$pct = $pct->{$smsg->{mid}};
+		if (defined $pct) {
+			$d .= (sprintf(' % 3u', $pct) . '%');
+		} else {
+			$unmatched = 1;
+			$d .= '     ';
+		}
+	}
+	$d .= ' ' . indent_for($level) . th_pfx($level);
 	my $attr = $f;
 	$ctx->{first_level} ||= $level;
 
@@ -976,7 +987,6 @@ sub skel_dump {
 	# our Xapian which would've resulted in '' if it were
 	# really missing (and Filter rejects empty subjects)
 	my @subj = split(/ /, subject_normalized($smsg->subject));
-
 	# remove common suffixes from the subject if it matches the previous,
 	# so we do not show redundant text at the end.
 	my $prev_subj = $ctx->{prev_subj} || [];
@@ -993,7 +1003,7 @@ sub skel_dump {
 	}
 	my $m;
 	my $id = '';
-	my $mapping = $ctx->{mapping};
+	my $mapping = $unmatched ? undef : $ctx->{mapping};
 	if ($mapping) {
 		my $map = $mapping->{$mid};
 		$id = id_compress($mid, 1);
@@ -1011,8 +1021,8 @@ sub _skel_ghost {
 	my ($ctx, $level, $node) = @_;
 
 	my $mid = $node->{id};
-	my $d = $ctx->{pct} ? '    [irrelevant] ' # search result
-			    : '     [not found] ';
+	my $d = '     [not found] ';
+	$d .= '     '  if exists $ctx->{pct};
 	$d .= indent_for($level) . th_pfx($level);
 	my $upfx = $ctx->{-upfx};
 	my $m = PublicInbox::Hval->new_msgid($mid);
diff --git a/t/psgi_v2.t b/t/psgi_v2.t
index 8e81e89b..10176516 100644
--- a/t/psgi_v2.t
+++ b/t/psgi_v2.t
@@ -155,7 +155,7 @@ test_psgi(sub { $www->call(@_) }, sub {
 	is($res->code, 200, 'success with threaded search');
 	my $raw = $res->content;
 	ok($raw =~ s/\A.*>Results 1-3 of 3\b//s, 'got all results');
-	my @over = ($raw =~ m/\d{4}-\d+-\d+\s+\d+:\d+ (.+)$/gm);
+	my @over = ($raw =~ m/\d{4}-\d+-\d+\s+\d+:\d+ +(?:\d+\% )?(.+)$/gm);
 	is_deeply(\@over, [ '<a', '` <a', '` <a' ], 'threaded messages show up');
 
 	local $SIG{__WARN__} = 'DEFAULT';