1 # This library is free software; you can redistribute it and/or modify
2 # it under the same terms as Perl itself.
4 # This license differs from the rest of public-inbox
6 # Our own jwz-style threading class based on Mail::Thread from CPAN.
7 # Mail::Thread is unmaintained and unavailable on some distros.
8 # We also do not want pruning or subject grouping, since we want
9 # to encourage strict threading and hopefully encourage people
10 # to use proper In-Reply-To/References.
12 # This includes fixes from several open bugs for Mail::Thread
14 # Avoid circular references
15 # - https://rt.cpan.org/Public/Bug/Display.html?id=22817
17 # And avoid recursion in recurse_down:
18 # - https://rt.cpan.org/Ticket/Display.html?id=116727
19 # - http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=833479
20 package PublicInbox::SearchThread;
23 use PublicInbox::MID qw($MID_EXTRACT);
26 my ($msgs, $ordersub, $ctx) = @_;
29 # Sadly, we sort here anyways since the fill-in-the-blanks References:
30 # can be shakier if somebody used In-Reply-To with multiple, disparate
31 # messages. So, take the client Date: into account since we can't
32 # alway determine ordering when somebody uses multiple In-Reply-To.
33 # We'll trust the client Date: header here instead of the Received:
34 # time since this is for display (and not retrieval)
35 _add_message($id_table, $_) for sort { $a->{ds} <=> $b->{ds} } @$msgs;
36 my $ibx = $ctx->{-inbox};
37 my $rootset = [ grep {
38 !delete($_->{parent}) && $_->visible($ibx)
39 } values %$id_table ];
41 $rootset = $ordersub->($rootset);
42 $_->order_children($ordersub, $ctx) for @$rootset;
46 sub _get_cont_for_id ($$) {
47 my ($id_table, $mid) = @_;
48 $id_table->{$mid} ||= PublicInbox::SearchThread::Msg->new($mid);
51 sub _add_message ($$) {
52 my ($id_table, $smsg) = @_;
55 my $this = _get_cont_for_id($id_table, $smsg->{mid});
56 $this->{smsg} = $smsg;
58 # saves around 4K across 1K messages
59 # TODO: move this to a more appropriate place, breaks tests
60 # if we do it during psgi_cull
63 # B. For each element in the message's References field:
64 defined(my $refs = $smsg->{references}) or return;
66 # This loop exists to help fill in gaps left from missing
67 # messages. It is not needed in a perfect world where
68 # everything is perfectly referenced, only the last ref
71 foreach my $ref ($refs =~ m/$MID_EXTRACT/go) {
72 # Find a Container object for the given Message-ID
73 my $cont = _get_cont_for_id($id_table, $ref);
75 # Link the References field's Containers together in
76 # the order implied by the References header
78 # * If they are already linked don't change the
80 # * Do not add a link if adding that link would
83 !$cont->{parent} && # already linked
84 !$cont->has_descendent($prev) # would loop
86 $prev->add_child($cont);
91 # C. Set the parent of this message to be the last element in
93 if (defined $prev && !$this->has_descendent($prev)) { # would loop
94 $prev->add_child($this);
98 package PublicInbox::SearchThread::Msg;
106 children => {}, # becomes an array when sorted by ->order(...)
113 while (my $cont = shift @q) {
114 return $cont if $cont->{smsg};
115 push @q, values %{$cont->{children}};
121 my ($self, $child) = @_;
122 croak "Cowardly refusing to become my own parent: $self"
125 my $cid = $child->{id};
128 if (defined(my $parent = $child->{parent})) {
129 delete $parent->{children}->{$cid};
132 $self->{children}->{$cid} = $child;
133 $child->{parent} = $self;
137 my ($self, $child) = @_;
138 my %seen; # loop prevention
140 return 1 if $self == $child || $seen{$child}++;
141 $child = $child->{parent};
146 # Do not show/keep ghosts iff they have no children. Sometimes
147 # a ghost Message-ID is the result of a long header line
148 # being folded/mangled by a MUA, and not a missing message.
150 my ($self, $ibx) = @_;
151 ($self->{smsg} ||= eval { $ibx->smsg_by_mid($self->{id}) }) ||
152 (scalar values %{$self->{children}});
156 my ($cur, $ordersub, $ctx) = @_;
158 my %seen = ($cur => 1); # self-referential loop prevention
160 my $ibx = $ctx->{-inbox};
161 while (defined($cur = shift @q)) {
162 my $c = $cur->{children}; # The hashref here...
164 $c = [ grep { !$seen{$_}++ && visible($_, $ibx) } values %$c ];
165 $c = $ordersub->($c) if scalar @$c > 1;
166 $cur->{children} = $c; # ...becomes an arrayref