here to allow users to specify their own
# color for quoted text
$$quot = undef;
$$s .= qq() . $rv . ''
}
sub attach_link ($$$$;$) {
my ($ctx, $ct, $p, $fn, $err) = @_;
my ($part, $depth, @idx) = @$p;
my $nl = $idx[-1] > 1 ? "\n" : '';
my $idx = join('.', @idx);
my $size = bytes::length($part->body);
# hide attributes normally, unless we want to aid users in
# spotting MUA problems:
$ct =~ s/;.*// unless $err;
$ct = ascii_html($ct);
my $desc = $part->header('Content-Description');
$desc = $fn unless defined $desc;
$desc = '' unless defined $desc;
my $sfn;
if (defined $fn && $fn =~ /\A$PublicInbox::Hval::FN\z/o) {
$sfn = $fn;
} elsif ($ct eq 'text/plain') {
$sfn = 'a.txt';
} else {
$sfn = 'a.bin';
}
my $rv = $ctx->{obuf};
$$rv .= qq($nl{mhref}$idx-$sfn">);
if ($err) {
$$rv .= "[-- Warning: decoded text below may be mangled --]\n";
}
$$rv .= "[-- Attachment #$idx: ";
my $ts = "Type: $ct, Size: $size bytes";
$desc = ascii_html($desc);
$$rv .= ($desc eq '') ? "$ts --]" : "$desc --]\n[-- $ts --]";
$$rv .= "\n";
undef;
}
sub add_text_body { # callback for msg_iter
my ($p, $ctx) = @_;
my $upfx = $ctx->{mhref};
my $ibx = $ctx->{-inbox};
# $p - from msg_iter: [ Email::MIME, depth, @idx ]
my ($part, $depth, @idx) = @$p;
my $ct = $part->content_type || 'text/plain';
my $fn = $part->filename;
my ($s, $err) = msg_part_text($part, $ct);
return attach_link($ctx, $ct, $p, $fn) unless defined $s;
# makes no difference to browsers, and don't screw up filename
# link generation in diffs with the extra '%0D'
$s =~ s/\r\n/\n/sg;
# always support diff-highlighting, but we can't linkify hunk
# headers for solver unless some coderepo are configured:
my $diff;
if ($s =~ /^(?:diff|---|\+{3}) /ms) {
# diffstat anchors do not link across attachments or messages:
$idx[0] = $upfx . $idx[0] if $upfx ne '';
$ctx->{-apfx} = join('/', @idx);
$ctx->{-anchors} = {}; # attr => filename
$diff = 1;
delete $ctx->{-long_path};
my $spfx;
if ($ibx->{-repo_objs}) {
if (index($upfx, '//') >= 0) { # absolute URL (Atom feeds)
$spfx = $upfx;
$spfx =~ s!/([^/]*)/\z!/!;
} else {
my $n_slash = $upfx =~ tr!/!/!;
if ($n_slash == 0) {
$spfx = '../';
} elsif ($n_slash == 1) {
$spfx = '';
} else { # nslash == 2
$spfx = '../../';
}
}
}
$ctx->{-spfx} = $spfx;
};
# some editors don't put trailing newlines at the end:
$s .= "\n" unless $s =~ /\n\z/s;
# split off quoted and unquoted blocks:
my @sections = split(/((?:^>[^\n]*\n)+)/sm, $s);
$s = '';
my $rv = $ctx->{obuf};
if (defined($fn) || $depth > 0 || $err) {
# badly-encoded message with $err? tell the world about it!
attach_link($ctx, $ct, $p, $fn, $err);
$$rv .= "\n";
}
my $l = $ctx->{-linkify} //= PublicInbox::Linkify->new;
foreach my $cur (@sections) {
if ($cur =~ /\A>/) {
flush_quote($rv, $l, \$cur);
} elsif ($diff) {
flush_diff($rv, $ctx, \$cur);
} else {
# regular lines, OK
$$rv .= $l->to_html($cur);
$cur = undef;
}
}
obfuscate_addrs($ibx, $$rv) if $ibx->{obfuscate};
}
sub _msg_page_prepare_obuf {
my ($hdr, $ctx, $nr) = @_;
my $over = $ctx->{-inbox}->over;
my $obfs_ibx = $ctx->{-obfs_ibx};
my $rv = '';
my $mids = mids_for_index($hdr);
if ($nr == 0) {
if ($ctx->{more}) {
$rv .=
"WARNING: multiple messages have this Message-ID\n
";
}
$rv .= ""; # anchor for body start
} else {
$rv .= '';
}
if ($over) {
$ctx->{-upfx} = '../';
}
my @title; # (Subject[0], From[0])
for my $v ($hdr->header('From')) {
my @n = PublicInbox::Address::names($v);
$v = ascii_html($v);
$title[1] //= ascii_html(join(', ', @n));
if ($obfs_ibx) {
obfuscate_addrs($obfs_ibx, $v);
obfuscate_addrs($obfs_ibx, $title[1]);
}
$rv .= "From: $v\n" if $v ne '';
}
foreach my $h (qw(To Cc)) {
for my $v ($hdr->header($h)) {
fold_addresses($v);
$v = ascii_html($v);
obfuscate_addrs($obfs_ibx, $v) if $obfs_ibx;
$rv .= "$h: $v\n" if $v ne '';
}
}
my @subj = $hdr->header('Subject');
if (@subj) {
my $v = ascii_html(shift @subj);
obfuscate_addrs($obfs_ibx, $v) if $obfs_ibx;
$rv .= 'Subject: ';
$rv .= $over ? qq($v\n) : "$v\n";
$title[0] = $v;
for $v (@subj) { # multi-Subject message :<
$v = ascii_html($v);
obfuscate_addrs($obfs_ibx, $v) if $obfs_ibx;
$rv .= "Subject: $v\n";
}
} else { # dummy anchor for thread skeleton at bottom of page
$rv .= qq() if $over;
$title[0] = '(no subject)';
}
for my $v ($hdr->header('Date')) {
$v = ascii_html($v);
obfuscate_addrs($obfs_ibx, $v) if $obfs_ibx; # possible :P
$rv .= "Date: $v\n";
}
$ctx->{-title_html} = join(' - ', @title);
if (scalar(@$mids) == 1) { # common case
my $mhtml = ascii_html($mids->[0]);
$rv .= "Message-ID: <$mhtml> ";
$rv .= "(raw)\n";
} else {
# X-Alt-Message-ID can happen if a message is injected from
# public-inbox-nntpd because of multiple Message-ID headers.
my $lnk = PublicInbox::Linkify->new;
my $s = '';
for my $h (qw(Message-ID X-Alt-Message-ID)) {
$s .= "$h: $_\n" for ($hdr->header_raw($h));
}
$lnk->linkify_mids('..', \$s, 1);
$rv .= $s;
}
$rv .= _parent_headers($hdr, $over);
$rv .= "\n";
\$rv;
}
sub SKEL_EXPAND () {
qq(expand[flat) .
qq(|nested] ) .
qq(mbox.gz ) .
qq(Atom feed);
}
sub thread_skel ($$$) {
my ($skel, $ctx, $hdr) = @_;
my $mid = mids($hdr)->[0];
my $ibx = $ctx->{-inbox};
my ($nr, $msgs) = $ibx->over->get_thread($mid);
my $parent = in_reply_to($hdr);
$$skel .= "\nThread overview: ";
if ($nr <= 1) {
if (defined $parent) {
$$skel .= SKEL_EXPAND."\n ";
$$skel .= ghost_parent('../', $parent) . "\n";
} else {
$$skel .= '[no followups] '.SKEL_EXPAND."\n";
}
$ctx->{next_msg} = undef;
$ctx->{parent_msg} = $parent;
return;
}
$$skel .= $nr;
$$skel .= '+ messages / '.SKEL_EXPAND.qq! top\n!;
# nb: mutt only shows the first Subject in the index pane
# when multiple Subject: headers are present, so we follow suit:
my $subj = $hdr->header('Subject') // '';
$subj = '(no subject)' if $subj eq '';
$ctx->{prev_subj} = [ split(/ /, subject_normalized($subj)) ];
$ctx->{cur} = $mid;
$ctx->{prev_attr} = '';
$ctx->{prev_level} = 0;
$ctx->{skel} = $skel;
# reduce hash lookups in skel_dump
$ctx->{-obfs_ibx} = $ibx->{obfuscate} ? $ibx : undef;
walk_thread(thread_results($ctx, $msgs), $ctx, \&skel_dump);
$ctx->{parent_msg} = $parent;
}
sub _parent_headers {
my ($hdr, $over) = @_;
my $rv = '';
my @irt = $hdr->header_raw('In-Reply-To');
my $refs;
if (@irt) {
my $lnk = PublicInbox::Linkify->new;
$rv .= "In-Reply-To: $_\n" for @irt;
$lnk->linkify_mids('..', \$rv);
} else {
$refs = references($hdr);
my $irt = pop @$refs;
if (defined $irt) {
my $html = ascii_html($irt);
my $href = mid_href($irt);
$rv .= "In-Reply-To: <";
$rv .= "$html>\n";
}
}
# do not display References: if search is present,
# we show the thread skeleton at the bottom, instead.
return $rv if $over;
$refs //= references($hdr);
if (@$refs) {
@$refs = map { linkify_ref_no_over($_) } @$refs;
$rv .= 'References: '. join("\n\t", @$refs) . "\n";
}
$rv;
}
# returns a string buffer via ->getline
sub html_footer {
my ($ctx, $hdr) = @_;
my $ibx = $ctx->{-inbox};
my $upfx = '../';
my $skel = " index";
my $rv = '';
if ($ibx->over) {
$skel .= "\n";
thread_skel(\$skel, $ctx, $hdr);
my ($next, $prev);
my $parent = ' ';
$next = $prev = ' ';
if (my $n = $ctx->{next_msg}) {
$n = mid_href($n);
$next = "next";
}
my $u;
my $par = $ctx->{parent_msg};
if ($par) {
$u = mid_href($par);
$u = "$upfx$u/";
}
if (my $p = $ctx->{prev_msg}) {
$prev = mid_href($p);
if ($p && $par && $p eq $par) {
$prev = "prev parent';
$parent = '';
} else {
$prev = "prev';
$parent = " parent" if $u;
}
} elsif ($u) { # unlikely
$parent = " parent";
}
$rv .= "$next $prev$parent ";
}
$rv .= qq(reply);
$rv .= $skel;
$rv .= '
';
$rv .= msg_reply($ctx, $hdr);
}
sub linkify_ref_no_over {
my ($mid) = @_;
my $href = mid_href($mid);
my $html = ascii_html($mid);
"<$html>";
}
sub anchor_for {
my ($msgid) = @_;
'm' . id_compress($msgid, 1);
}
sub ghost_parent {
my ($upfx, $mid) = @_;
my $href = mid_href($mid);
my $html = ascii_html($mid);
qq{[parent not found: <$html>]};
}
sub indent_for {
my ($level) = @_;
$level ? INDENT x ($level - 1) : '';
}
sub find_mid_root {
my ($ctx, $level, $node, $idx) = @_;
++$ctx->{root_idx} if $level == 0;
if ($node->{id} eq $ctx->{mid}) {
$ctx->{found_mid_at} = $ctx->{root_idx};
return 0;
}
1;
}
sub strict_loose_note ($) {
my ($nr) = @_;
my $msg =
" -- strict thread matches above, loose matches on Subject: below --\n";
if ($nr > PublicInbox::Over::DEFAULT_LIMIT()) {
$msg .=
" -- use mbox.gz link to download all $nr messages --\n";
}
$msg;
}
sub thread_results {
my ($ctx, $msgs) = @_;
require PublicInbox::SearchThread;
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
# in get_thread, so we never noticed it until now. And even when
# reindexing is fixed, we'll keep this code until a SCHEMA_VERSION
# bump since reindexing is expensive and users may not do it
# loose threading could've returned too many results,
# put the root the message we care about at the top:
my $mid = $ctx->{mid};
if (defined($mid) && scalar(@$rootset) > 1) {
$ctx->{root_idx} = -1;
my $nr = scalar @$msgs;
walk_thread($rootset, $ctx, \&find_mid_root);
my $idx = $ctx->{found_mid_at};
if (defined($idx) && $idx != 0) {
my $tip = splice(@$rootset, $idx, 1);
@$rootset = reverse @$rootset;
unshift @$rootset, $tip;
$ctx->{sl_note} = strict_loose_note($nr);
}
}
$rootset
}
sub missing_thread {
my ($ctx) = @_;
require PublicInbox::ExtMsg;
PublicInbox::ExtMsg::ext_msg($ctx);
}
sub dedupe_subject {
my ($prev_subj, $subj, $val) = @_;
my $omit = ''; # '"' denotes identical text omitted
my (@prev_pop, @curr_pop);
while (@$prev_subj && @$subj && $subj->[-1] eq $prev_subj->[-1]) {
push(@prev_pop, pop(@$prev_subj));
push(@curr_pop, pop(@$subj));
$omit ||= $val;
}
pop @$subj if @$subj && $subj->[-1] =~ /^re:\s*/i;
if (scalar(@curr_pop) == 1) {
$omit = '';
push @$prev_subj, @prev_pop;
push @$subj, @curr_pop;
}
$omit;
}
sub skel_dump { # walk_thread callback
my ($ctx, $level, $node) = @_;
my $smsg = $node->{smsg} or return _skel_ghost($ctx, $level, $node);
my $skel = $ctx->{skel};
my $cur = $ctx->{cur};
my $mid = $smsg->{mid};
if ($level == 0 && $ctx->{skel_dump_roots}++) {
$$skel .= delete($ctx->{sl_note}) || '';
}
my $f = ascii_html($smsg->{from_name});
my $obfs_ibx = $ctx->{-obfs_ibx};
obfuscate_addrs($obfs_ibx, $f) if $obfs_ibx;
my $d = fmt_ts($smsg->{ds});
my $unmatched; # if lazy-loaded by SearchThread::Msg::visible()
if (exists $ctx->{searchview}) {
if (defined(my $pct = $smsg->{pct})) {
$d .= (sprintf(' % 2u', $pct) . '%');
} else {
$unmatched = 1;
$d .= ' ';
}
}
$d .= ' ' . indent_for($level) . th_pfx($level);
my $attr = $f;
$ctx->{first_level} ||= $level;
if ($attr ne $ctx->{prev_attr} || $ctx->{prev_level} > $level) {
$ctx->{prev_attr} = $attr;
}
$ctx->{prev_level} = $level;
if ($cur) {
if ($cur eq $mid) {
delete $ctx->{cur};
$$skel .= "$d".
"$attr [this message]\n";
return 1;
} else {
$ctx->{prev_msg} = $mid;
}
} else {
$ctx->{next_msg} ||= $mid;
}
# Subject is never undef, this mail was loaded from
# our Xapian which would've resulted in '' if it were
# really missing (and Filter rejects empty subjects)
my @subj = split(/ /, subject_normalized($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} || [];
$ctx->{prev_subj} = [ @subj ];
my $omit = dedupe_subject($prev_subj, \@subj, '" ');
my $end;
if (@subj) {
my $subj = join(' ', @subj);
$subj = ascii_html($subj);
obfuscate_addrs($obfs_ibx, $subj) if $obfs_ibx;
$end = "$subj $omit$f\n"
} else {
$end = "$f\n";
}
my $m;
my $id = '';
my $mapping = $unmatched ? undef : $ctx->{mapping};
if ($mapping) {
my $map = $mapping->{$mid};
$id = id_compress($mid, 1);
$m = '#m'.$id;
$map->[0] = "$d$end";
$id = "\nid=r".$id;
} else {
$m = $ctx->{-upfx}.mid_href($mid).'/';
}
$$skel .= $d . "" . $end;
1;
}
sub _skel_ghost {
my ($ctx, $level, $node) = @_;
my $mid = $node->{id};
my $d = ' [not found] ';
$d .= ' ' if exists $ctx->{searchview};
$d .= indent_for($level) . th_pfx($level);
my $upfx = $ctx->{-upfx};
my $href = $upfx . mid_href($mid) . '/';
my $html = ascii_html($mid);
my $mapping = $ctx->{mapping};
my $map = $mapping->{$mid} if $mapping;
if ($map) {
my $id = id_compress($mid, 1);
$map->[0] = $d . qq{<$html>\n};
$d .= qq{<$html>\n};
} else {
$d .= qq{<$html>\n};
}
${$ctx->{skel}} .= $d;
1;
}
sub sort_ds {
[ sort {
(eval { $a->topmost->{smsg}->{ds} } || 0) <=>
(eval { $b->topmost->{smsg}->{ds} } || 0)
} @{$_[0]} ];
}
# accumulate recent topics if search is supported
# returns 200 if done, 404 if not
sub acc_topic { # walk_thread callback
my ($ctx, $level, $node) = @_;
my $mid = $node->{id};
my $smsg = $node->{smsg} // $ctx->{-inbox}->smsg_by_mid($mid);
if ($smsg) {
my $subj = subject_normalized($smsg->{subject});
$subj = '(no subject)' if $subj eq '';
my $ds = $smsg->{ds};
if ($level == 0) { # new, top-level topic
my $topic = [ $ds, 1, { $subj => $mid }, $subj ];
$ctx->{-cur_topic} = $topic;
push @{$ctx->{order}}, $topic;
return 1;
}
# continue existing topic
my $topic = $ctx->{-cur_topic}; # should never be undef
$topic->[0] = $ds if $ds > $topic->[0];
$topic->[1]++; # bump N+ message counter
my $seen = $topic->[2];
if (scalar(@$topic) == 3) { # parent was a ghost
push @$topic, $subj;
} elsif (!defined($seen->{$subj})) {
push @$topic, $level, $subj; # @extra messages
}
$seen->{$subj} = $mid; # latest for subject
} else { # ghost message
return 1 if $level != 0; # ignore child ghosts
my $topic = $ctx->{-cur_topic} = [ -666, 0, {} ];
push @{$ctx->{order}}, $topic;
}
1;
}
sub dump_topics {
my ($ctx) = @_;
my $order = delete $ctx->{order}; # [ ds, subj1, subj2, subj3, ... ]
if (!@$order) {
$ctx->{-html_tip} = '[No topics in range]
';
return 404;
}
my @out;
my $ibx = $ctx->{-inbox};
my $obfs_ibx = $ibx->{obfuscate} ? $ibx : undef;
# sort by recency, this allows new posts to "bump" old topics...
foreach my $topic (sort { $b->[0] <=> $a->[0] } @$order) {
my ($ds, $n, $seen, $top_subj, @extra) = @$topic;
@$topic = ();
next unless defined $top_subj; # ghost topic
my $mid = delete $seen->{$top_subj};
my $href = mid_href($mid);
my $prev_subj = [ split(/ /, $top_subj) ];
$top_subj = ascii_html($top_subj);
$ds = fmt_ts($ds);
# $n isn't the total number of posts on the topic,
# just the number of posts in the current results window
my $anchor;
if ($n == 1) {
$n = '';
$anchor = '#u'; # top of only message
} else {
$n = " ($n+ messages)";
$anchor = '#t'; # thread skeleton
}
my $mbox = qq(mbox.gz);
my $atom = qq(Atom);
my $s = "$top_subj\n" .
" $ds UTC $n - $mbox / $atom\n";
for (my $i = 0; $i < scalar(@extra); $i += 2) {
my $level = $extra[$i];
my $subj = $extra[$i + 1]; # already normalized
$mid = delete $seen->{$subj};
my @subj = split(/ /, $subj);
my @next_prev = @subj; # full copy
my $omit = dedupe_subject($prev_subj, \@subj, ' "');
$prev_subj = \@next_prev;
$subj = ascii_html($subj);
obfuscate_addrs($obfs_ibx, $subj) if $obfs_ibx;
$href = mid_href($mid);
$s .= indent_for($level) . TCHILD;
$s .= qq($subj$omit\n);
}
push @out, $s;
}
$ctx->{-html_tip} = '' . join("\n", @out) . '
';
200;
}
# only for the t= query parameter passed to overview DB
sub ts2str ($) { strftime('%Y%m%d%H%M%S', gmtime($_[0])) };
sub str2ts ($) {
my ($yyyy, $mon, $dd, $hh, $mm, $ss) = unpack('A4A2A2A2A2A2', $_[0]);
timegm($ss, $mm, $hh, $dd, $mon - 1, $yyyy);
}
sub pagination_footer ($$) {
my ($ctx, $latest) = @_;
delete $ctx->{qp} or return;
my $next = $ctx->{next_page} || '';
my $prev = $ctx->{prev_page} || '';
if ($prev) {
$next = $next ? "$next " : ' ';
$prev .= qq! latest!;
}
"
page: $next$prev
";
}
sub index_nav { # callback for WwwStream
my (undef, $ctx) = @_;
pagination_footer($ctx, '.')
}
sub paginate_recent ($$) {
my ($ctx, $lim) = @_;
my $t = $ctx->{qp}->{t} || '';
my $opts = { limit => $lim };
my ($after, $before);
# Xapian uses '..' but '-' is perhaps friendier to URL linkifiers
# if only $after exists "YYYYMMDD.." because "." could be skipped
# if interpreted as an end-of-sentence
$t =~ s/\A([0-9]{8,14})-// and $after = str2ts($1);
$t =~ /\A([0-9]{8,14})\z/ and $before = str2ts($1);
my $ibx = $ctx->{-inbox};
my $msgs = $ibx->recent($opts, $after, $before);
my $nr = scalar @$msgs;
if ($nr < $lim && defined($after)) {
$after = $before = undef;
$msgs = $ibx->recent($opts);
$nr = scalar @$msgs;
}
my $more = $nr == $lim;
my ($newest, $oldest);
if ($nr) {
$newest = $msgs->[0]->{ts};
$oldest = $msgs->[-1]->{ts};
# if we only had $after, our SQL query in ->recent ordered
if ($newest < $oldest) {
($oldest, $newest) = ($newest, $oldest);
$more = 0 if defined($after) && $after < $oldest;
}
}
if (defined($oldest) && $more) {
my $s = ts2str($oldest);
$ctx->{next_page} = qq!next!;
}
if (defined($newest) && (defined($before) || defined($after))) {
my $s = ts2str($newest);
$ctx->{prev_page} = qq!prev!;
}
$msgs;
}
sub index_topics {
my ($ctx) = @_;
my $msgs = paginate_recent($ctx, 200); # 200 is our window
if (@$msgs) {
walk_thread(thread_results($ctx, $msgs), $ctx, \&acc_topic);
}
PublicInbox::WwwStream->response($ctx, dump_topics($ctx), \&index_nav);
}
sub thread_adj_level {
my ($ctx, $level) = @_;
my $max = $ctx->{cur_level};
if ($level <= 0) {
return ('', '') if $max == 0; # flat output
# reset existing lists
my $beg = $max > 1 ? ('' x ($max - 1)) : '';
$ctx->{cur_level} = 0;
("$beg", '');
} elsif ($level == $max) { # continue existing list
qw(
);
} elsif ($level < $max) {
my $beg = $max > 1 ? ('' x ($max - $level)) : '';
$ctx->{cur_level} = $level;
("$beg", '');
} else { # ($level > $max) # start a new level
$ctx->{cur_level} = $level;
my $beg = ($max ? '' : '') . '- ';
($beg, '
');
}
}
sub ghost_index_entry {
my ($ctx, $level, $node) = @_;
my ($beg, $end) = thread_adj_level($ctx, $level);
$beg . '
'. ghost_parent($ctx->{-upfx}, $node->{id})
. '
' . $end;
}
1;