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;
25 my ($msgs, $ordersub, $ctx) = @_;
28 # Sadly, we sort here anyways since the fill-in-the-blanks References:
29 # can be shakier if somebody used In-Reply-To with multiple, disparate
30 # messages. So, take the client Date: into account since we can't
31 # alway determine ordering when somebody uses multiple In-Reply-To.
32 # We'll trust the client Date: header here instead of the Received:
33 # time since this is for display (and not retrieval)
34 _add_message($id_table, $_) for sort { $a->{ds} <=> $b->{ds} } @$msgs;
35 my $ibx = $ctx->{-inbox};
36 my $rootset = [ grep {
37 !delete($_->{parent}) && $_->visible($ibx)
38 } values %$id_table ];
40 $rootset = $ordersub->($rootset);
41 $_->order_children($ordersub, $ctx) for @$rootset;
45 sub _get_cont_for_id ($$) {
46 my ($id_table, $mid) = @_;
47 $id_table->{$mid} ||= PublicInbox::SearchThread::Msg->new($mid);
50 sub _add_message ($$) {
51 my ($id_table, $smsg) = @_;
54 my $this = _get_cont_for_id($id_table, $smsg->{mid});
55 $this->{smsg} = $smsg;
57 # saves around 4K across 1K messages
58 # TODO: move this to a more appropriate place, breaks tests
59 # if we do it during psgi_cull
62 # B. For each element in the message's References field:
63 defined(my $refs = $smsg->{references}) or return;
65 # This loop exists to help fill in gaps left from missing
66 # messages. It is not needed in a perfect world where
67 # everything is perfectly referenced, only the last ref
70 foreach my $ref ($refs =~ m/<([^>]+)>/g) {
71 # Find a Container object for the given Message-ID
72 my $cont = _get_cont_for_id($id_table, $ref);
74 # Link the References field's Containers together in
75 # the order implied by the References header
77 # * If they are already linked don't change the
79 # * Do not add a link if adding that link would
82 !$cont->{parent} && # already linked
83 !$cont->has_descendent($prev) # would loop
85 $prev->add_child($cont);
90 # C. Set the parent of this message to be the last element in
92 if (defined $prev && !$this->has_descendent($prev)) { # would loop
93 $prev->add_child($this);
97 package PublicInbox::SearchThread::Msg;
105 children => {}, # becomes an array when sorted by ->order(...)
112 while (my $cont = shift @q) {
113 return $cont if $cont->{smsg};
114 push @q, values %{$cont->{children}};
120 my ($self, $child) = @_;
121 croak "Cowardly refusing to become my own parent: $self"
124 my $cid = $child->{id};
127 if (defined(my $parent = $child->{parent})) {
128 delete $parent->{children}->{$cid};
131 $self->{children}->{$cid} = $child;
132 $child->{parent} = $self;
136 my ($self, $child) = @_;
137 my %seen; # loop prevention
139 return 1 if $self == $child || $seen{$child}++;
140 $child = $child->{parent};
145 # Do not show/keep ghosts iff they have no children. Sometimes
146 # a ghost Message-ID is the result of a long header line
147 # being folded/mangled by a MUA, and not a missing message.
149 my ($self, $ibx) = @_;
150 ($self->{smsg} ||= eval { $ibx->smsg_by_mid($self->{id}) }) ||
151 (scalar values %{$self->{children}});
155 my ($cur, $ordersub, $ctx) = @_;
157 my %seen = ($cur => 1); # self-referential loop prevention
159 my $ibx = $ctx->{-inbox};
160 while (defined($cur = shift @q)) {
161 my $c = $cur->{children}; # The hashref here...
163 $c = [ grep { !$seen{$_}++ && visible($_, $ibx) } values %$c ];
164 $c = $ordersub->($c) if scalar @$c > 1;
165 $cur->{children} = $c; # ...becomes an arrayref