@ $attr";
+ if ($nr_c) {
+ my $cmid = $children->[0]->{id};
+ $rv .= $pad . $mapping->{$cmid}->[0];
if ($nr_c > 2) {
my $s = ($nr_c - 1). ' more replies';
$rv .= pad_link($cmid, $level + 1, $s);
- } elsif (my $cn = $child->next) {
- $rv .= $pad . $mapping->{$cn->messageid}->[1];
+ } elsif (my $cn = $children->[1]) {
+ $rv .= $pad . $mapping->{$cn->{id}}->[0];
}
}
- if (my $next = $node->next) {
- my $nmid = $next->messageid;
- $rv .= $pad . $mapping->{$nmid}->[1];
+
+ my $next = $siblings->[$idx+1] if $siblings && $idx >= 0;
+ if ($next) {
+ my $nmid = $next->{id};
+ $rv .= $pad . $mapping->{$nmid}->[0];
my $nnext = $nr_s - $idx;
if ($nnext > 2) {
my $s = ($nnext - 1).' subsequent siblings';
$rv .= pad_link($nmid, $level, $s);
- } elsif (my $nn = $next->next) {
- $rv .= $pad . $mapping->{$nn->messageid}->[1];
+ } elsif (my $nn = $siblings->[$idx + 2]) {
+ $rv .= $pad . $mapping->{$nn->{id}}->[0];
}
}
$rv .= $pad ."$s_s, $s_c; $ctx->{s_nr}\n";
}
sub walk_thread {
- my ($th, $ctx, $cb) = @_;
- my @q = map { (0, $_) } $th->rootset;
+ my ($rootset, $ctx, $cb) = @_;
+ my @q = map { (0, $_, -1) } @$rootset;
while (@q) {
- my $level = shift @q;
- my $node = shift @q or next;
- $cb->($ctx, $level, $node);
- unshift @q, $level+1, $node->child, $level, $node->next;
+ my ($level, $node, $i) = splice(@q, 0, 3);
+ defined $node or next;
+ $cb->($ctx, $level, $node, $i) or return;
+ ++$level;
+ $i = 0;
+ unshift @q, map { ($level, $_, $i++) } @{$node->{children}};
}
}
sub pre_thread {
- my ($ctx, $level, $node) = @_;
- my $mapping = $ctx->{mapping};
- my $idx = -1;
- if (my $parent = $node->parent) {
- my $m = $mapping->{$parent->messageid}->[0];
- $idx = scalar @$m;
- push @$m, $node;
- }
- $mapping->{$node->messageid} = [ [], '', $node, $idx, $level ];
+ my ($ctx, $level, $node, $idx) = @_;
+ $ctx->{mapping}->{$node->{id}} = [ '', $node, $idx, $level ];
skel_dump($ctx, $level, $node);
}
sub thread_index_entry {
- my ($ctx, $level, $mime) = @_;
+ my ($ctx, $level, $smsg) = @_;
my ($beg, $end) = thread_adj_level($ctx, $level);
- $beg . '' . index_entry($mime, $ctx, 0) . '
' . $end;
+ $beg . '' . index_entry($smsg, $ctx, 0) . '
' . $end;
}
sub stream_thread ($$) {
- my ($th, $ctx) = @_;
- my $inbox = $ctx->{-inbox};
- my $mime;
- my @q = map { (0, $_) } $th->rootset;
+ my ($rootset, $ctx) = @_;
+ my $ibx = $ctx->{-inbox};
+ my @q = map { (0, $_) } @$rootset;
my $level;
+ my $smsg;
while (@q) {
$level = shift @q;
my $node = shift @q or next;
- unshift @q, $level+1, $node->child, $level, $node->next;
- $mime = $inbox->msg_by_mid($node->messageid) and last;
+ my $cl = $level + 1;
+ unshift @q, map { ($cl, $_) } @{$node->{children}};
+ $smsg = $ibx->smsg_mime($node->{smsg}) and last;
}
- return missing_thread($ctx) unless $mime;
+ return missing_thread($ctx) unless $smsg;
- $mime = Email::MIME->new($mime);
- $ctx->{-title_html} = ascii_html($mime->header('Subject'));
- $ctx->{-html_tip} = thread_index_entry($ctx, $level, $mime);
+ $ctx->{-obfs_ibx} = $ibx->{obfuscate} ? $ibx : undef;
+ $ctx->{-title_html} = ascii_html($smsg->subject);
+ $ctx->{-html_tip} = thread_index_entry($ctx, $level, $smsg);
+ $smsg = undef;
PublicInbox::WwwStream->response($ctx, 200, sub {
return unless $ctx;
while (@q) {
$level = shift @q;
my $node = shift @q or next;
- unshift @q, $level+1, $node->child, $level, $node->next;
- my $mid = $node->messageid;
- if ($mime = $inbox->msg_by_mid($mid)) {
- $mime = Email::MIME->new($mime);
- return thread_index_entry($ctx, $level, $mime);
+ my $cl = $level + 1;
+ unshift @q, map { ($cl, $_) } @{$node->{children}};
+ if ($smsg = $ibx->smsg_mime($node->{smsg})) {
+ return thread_index_entry($ctx, $level, $smsg);
} else {
- return ghost_index_entry($ctx, $level, $mid);
+ return ghost_index_entry($ctx, $level, $node);
}
}
my $ret = join('', thread_adj_level($ctx, 0));
@@ -316,66 +439,63 @@ sub stream_thread ($$) {
sub thread_html {
my ($ctx) = @_;
my $mid = $ctx->{mid};
- my $sres = $ctx->{srch}->get_thread($mid, { asc => 1 });
- my $msgs = load_results($sres);
- my $nr = $sres->{total};
+ my $ibx = $ctx->{-inbox};
+ my ($nr, $msgs) = $ibx->over->get_thread($mid);
return missing_thread($ctx) if $nr == 0;
my $skel = '
';
$skel .= $nr == 1 ? 'only message in thread' : 'end of thread';
- $skel .= ", back to index";
- $skel .= "\n$nr+ messages in thread: (download: ";
- $skel .= "mbox.gz";
+ $skel .= ", back to index\n\n";
+ $skel .= "Thread overview: ";
+ $skel .= $nr == 1 ? '(only message)' : "$nr+ messages";
+ $skel .= " (download: mbox.gz";
$skel .= " / follow: Atom feed)\n";
+ $skel .= "-- links below jump to the message on this page --\n";
$ctx->{-upfx} = '../../';
$ctx->{cur_level} = 0;
$ctx->{dst} = \$skel;
$ctx->{prev_attr} = '';
$ctx->{prev_level} = 0;
$ctx->{root_anchor} = anchor_for($mid);
- $ctx->{seen} = {};
$ctx->{mapping} = {};
- $ctx->{s_nr} = "$nr+ messages in thread";
+ $ctx->{s_nr} = ($nr > 1 ? "$nr+ messages" : 'only message')
+ .' in thread';
+
+ my $rootset = thread_results($ctx, $msgs);
+
+ # reduce hash lookups in pre_thread->skel_dump
+ $ctx->{-obfs_ibx} = $ibx->{obfuscate} ? $ibx : undef;
+ walk_thread($rootset, $ctx, *pre_thread);
- my $th = thread_results($msgs);
- walk_thread($th, $ctx, *pre_thread);
$skel .= '
';
- return stream_thread($th, $ctx) unless $ctx->{flat};
-
- # flat display: lazy load the full message from mini_mime:
- my $inbox = $ctx->{-inbox};
- my $mime;
- while ($mime = shift @$msgs) {
- $mime = $inbox->msg_by_mid(mid_clean(mid_mime($mime))) and last;
- }
- return missing_thread($ctx) unless $mime;
- $mime = Email::MIME->new($mime);
- $ctx->{-title_html} = ascii_html($mime->header('Subject'));
- $ctx->{-html_tip} = ''.index_entry($mime, $ctx, scalar @$msgs);
- $mime = undef;
+ return stream_thread($rootset, $ctx) unless $ctx->{flat};
+
+ # flat display: lazy load the full message from smsg
+ my $smsg;
+ while (my $m = shift @$msgs) {
+ $smsg = $ibx->smsg_mime($m) and last;
+ }
+ return missing_thread($ctx) unless $smsg;
+ $ctx->{-title_html} = ascii_html($smsg->subject);
+ $ctx->{-html_tip} = ''.index_entry($smsg, $ctx, scalar @$msgs);
+ $smsg = undef;
PublicInbox::WwwStream->response($ctx, 200, sub {
return unless $msgs;
- while ($mime = shift @$msgs) {
- $mid = mid_clean(mid_mime($mime));
- $mime = $inbox->msg_by_mid($mid) and last;
- }
- if ($mime) {
- $mime = Email::MIME->new($mime);
- return index_entry($mime, $ctx, scalar @$msgs);
+ $smsg = undef;
+ while (my $m = shift @$msgs) {
+ $smsg = $ibx->smsg_mime($m) and last;
}
+ return index_entry($smsg, $ctx, scalar @$msgs) if $smsg;
$msgs = undef;
- '
'.$skel;
+ $skel;
});
}
sub multipart_text_as_html {
- my ($mime, $upfx) = @_;
+ my ($mime, $upfx, $ctx) = @_;
my $rv = "";
# scan through all parts, looking for displayable text
- msg_iter($mime, sub {
- my ($p) = @_;
- $rv .= add_text_body($upfx, $p);
- });
+ msg_iter($mime, sub { $rv .= add_text_body($upfx, $ctx, $_[0]) });
$rv;
}
@@ -384,273 +504,323 @@ sub flush_quote {
# show everything in the full version with anchor from
# short version (see above)
- my $rv = $l->linkify_1(join('', @$quot));
- @$quot = ();
+ my $rv = $l->linkify_1($$quot);
# we use a here to allow users to specify their own
# color for quoted text
$rv = $l->linkify_2(ascii_html($rv));
+ $$quot = undef;
$$s .= qq() . $rv . ''
}
-sub attach_link ($$$$) {
- my ($upfx, $ct, $p, $fn) = @_;
+sub attach_link ($$$$;$) {
+ my ($upfx, $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);
- $ct ||= 'text/plain';
- $ct =~ s/;.*//; # no attributes
+
+ # 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[[:alnum:]][\w\.-]+[[:alnum:]]\z/) {
+ 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 @ret = qq($nl[-- Attachment #$idx: );
+ my $ret = qq($nl);
+ if ($err) {
+ $ret .=
+"[-- Warning: decoded text below may be mangled --]\n";
+ }
+ $ret .= "[-- Attachment #$idx: ";
my $ts = "Type: $ct, Size: $size bytes";
- push(@ret, ($desc eq '') ? "$ts --]" : "$desc --]\n[-- $ts --]");
- join('', @ret, "\n");
+ $desc = ascii_html($desc);
+ $ret .= ($desc eq '') ? "$ts --]" : "$desc --]\n[-- $ts --]";
+ $ret .= "\n";
}
sub add_text_body {
- my ($upfx, $p) = @_; # from msg_iter: [ Email::MIME, depth, @idx ]
+ my ($upfx, $ctx, $p) = @_;
+ my $ibx = $ctx->{-inbox};
+ my $obfs_ibx = $ibx->{obfuscate} ? $ibx : undef;
+ # $p - from msg_iter: [ Email::MIME, depth, @idx ]
my ($part, $depth, @idx) = @$p;
- my $ct = $part->content_type;
+ my $ct = $part->content_type || 'text/plain';
my $fn = $part->filename;
+ my ($s, $err) = msg_part_text($part, $ct);
+
+ return attach_link($upfx, $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
+ $ctx->{-diff} = $diff = [];
+ 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;
+ };
- if (defined $ct && $ct =~ m!\btext/x?html\b!i) {
- return attach_link($upfx, $ct, $p, $fn);
- }
-
- my $s = eval { $part->body_str };
-
- # badly-encoded message? tell the world about it!
- return attach_link($upfx, $ct, $p, $fn) if $@;
+ # some editors don't put trailing newlines at the end:
+ $s .= "\n" unless $s =~ /\n\z/s;
- my @lines = split(/^/m, $s);
+ # split off quoted and unquoted blocks:
+ my @sections = split(/((?:^>[^\n]*\n)+)/sm, $s);
$s = '';
- if (defined($fn) || $depth > 0) {
- $s .= attach_link($upfx, $ct, $p, $fn);
+ if (defined($fn) || $depth > 0 || $err) {
+ # badly-encoded message with $err? tell the world about it!
+ $s .= attach_link($upfx, $ct, $p, $fn, $err);
$s .= "\n";
}
- my @quot;
my $l = PublicInbox::Linkify->new;
- while (defined(my $cur = shift @lines)) {
- if ($cur !~ /^>/) {
- # show the previously buffered quote inline
- flush_quote(\$s, $l, \@quot) if @quot;
-
- # regular line, OK
- $cur = $l->linkify_1($cur);
- $cur = ascii_html($cur);
- $s .= $l->linkify_2($cur);
+ foreach my $cur (@sections) {
+ if ($cur =~ /\A>/) {
+ flush_quote(\$s, $l, \$cur);
+ } elsif ($diff) {
+ @$diff = split(/^/m, $cur);
+ $cur = undef;
+ flush_diff(\$s, $ctx, $l);
} else {
- push @quot, $cur;
+ # regular lines, OK
+ $l->linkify_1($cur);
+ $s .= $l->linkify_2(ascii_html($cur));
+ $cur = undef;
}
}
- my $end = "\n";
- if (@quot) {
- $end = '';
- flush_quote(\$s, $l, \@quot);
- }
- $s =~ s/[ \t]+$//sgm; # kill per-line trailing whitespace
- $s =~ s/\A\n+//s; # kill leading blank lines
- $s =~ s/\s+\z//s; # kill all trailing spaces
- $s .= $end;
+ obfuscate_addrs($obfs_ibx, $s) if $obfs_ibx;
+ $s;
}
sub _msg_html_prepare {
- my ($hdr, $ctx) = @_;
- my $srch = $ctx->{srch} if $ctx;
+ my ($hdr, $ctx, $more, $nr) = @_;
my $atom = '';
- my $rv = ""; # anchor for body start
-
- if ($srch) {
+ my $over = $ctx->{-inbox}->over;
+ my $obfs_ibx = $ctx->{-obfs_ibx};
+ my $rv = '';
+ my $mids = mids($hdr);
+ my $multiple = scalar(@$mids) > 1; # zero, one, infinity
+ if ($nr == 0) {
+ if ($more) {
+ $rv .=
+"WARNING: multiple messages refer to this Message-ID\n
";
+ }
+ $rv .= ""; # anchor for body start
+ } else {
+ $rv .= '';
+ }
+ if ($over) {
$ctx->{-upfx} = '../';
}
- my @title;
- my $mid = $hdr->header_raw('Message-ID');
- $mid = PublicInbox::Hval->new_msgid($mid);
- foreach my $h (qw(From To Cc Subject Date)) {
- my $v = $hdr->header($h);
- defined($v) && ($v ne '') or next;
+ my @title; # (Subject[0], From[0])
+ for my $v ($hdr->header('From')) {
$v = PublicInbox::Hval->new($v);
-
- if ($h eq 'From') {
- my @n = PublicInbox::Address::names($v->raw);
- $title[1] = ascii_html(join(', ', @n));
- } elsif ($h eq 'Subject') {
- $title[0] = $v->as_html;
- if ($srch) {
- $rv .= qq($h: );
- $rv .= $v->as_html . "\n";
- next;
+ my @n = PublicInbox::Address::names($v->raw);
+ $title[1] //= ascii_html(join(', ', @n));
+ $v = $v->as_html;
+ 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) {
+ for my $v (@subj) {
+ $v = ascii_html($v);
+ obfuscate_addrs($obfs_ibx, $v) if $obfs_ibx;
+ $rv .= 'Subject: ';
+ if ($over) {
+ $rv .= qq($v\n);
+ } else {
+ $rv .= "$v\n";
}
+ $title[0] //= $v;
}
- $v = $v->as_html;
- $v =~ s/(\@[^,]+,) /$1\n\t/g if ($h eq 'Cc' || $h eq 'To');
- $rv .= "$h: $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";
}
- $title[0] ||= '(no subject)';
$ctx->{-title_html} = join(' - ', @title);
- $rv .= 'Message-ID: <' . $mid->as_html . '> ';
- $rv .= "(raw)\n";
- $rv .= _parent_headers($hdr, $srch);
+ foreach (@$mids) {
+ my $mid = PublicInbox::Hval->new_msgid($_) ;
+ my $mhtml = $mid->as_html;
+ if ($multiple) {
+ my $href = $mid->{href};
+ $rv .= "Message-ID: ";
+ $rv .= "";
+ $rv .= "<$mhtml> ";
+ $rv .= "(raw)\n";
+ } else {
+ $rv .= "Message-ID: <$mhtml> ";
+ $rv .= "(raw)\n";
+ }
+ }
+ $rv .= _parent_headers($hdr, $over);
$rv .= "\n";
}
sub thread_skel {
my ($dst, $ctx, $hdr, $tpfx) = @_;
- my $srch = $ctx->{srch};
- my $mid = mid_clean($hdr->header_raw('Message-ID'));
- my $sres = $srch->get_thread($mid);
- my $nr = $sres->{total};
- my $expand = qq(expand ) .
- qq(/ mbox.gz ) .
- qq(/ Atom feed);
+ my $mid = mids($hdr)->[0];
+ my $ibx = $ctx->{-inbox};
+ my ($nr, $msgs) = $ibx->over->get_thread($mid);
+ my $expand = qq(expand[flat) .
+ qq(|nested] ) .
+ qq(mbox.gz ) .
+ qq(Atom feed);
my $parent = in_reply_to($hdr);
+ $$dst .= "\nThread overview: ";
if ($nr <= 1) {
if (defined $parent) {
- $$dst .= "($expand)\n ";
+ $$dst .= "$expand\n ";
$$dst .= ghost_parent("$tpfx../", $parent) . "\n";
} else {
- $$dst .= "[no followups, yet] ($expand)\n";
+ $$dst .= "[no followups] $expand\n";
}
$ctx->{next_msg} = undef;
$ctx->{parent_msg} = $parent;
return;
}
- $$dst .= "$nr+ messages in thread ($expand";
- $$dst .= qq! / [top])\n!;
+ $$dst .= "$nr+ messages / $expand";
+ $$dst .= qq! top\n!;
- my $subj = $srch->subject_path($hdr->header('Subject'));
- $ctx->{seen} = { $subj => 1 };
+ # 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->{dst} = $dst;
- walk_thread(thread_results(load_results($sres)), $ctx, *skel_dump);
+
+ # 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, $srch) = @_;
+ my ($hdr, $over) = @_;
my $rv = '';
-
- my $irt = in_reply_to($hdr);
- if (defined $irt) {
- my $v = PublicInbox::Hval->new_msgid($irt);
- my $html = $v->as_html;
- my $href = $v->as_href;
- $rv .= "In-Reply-To: <";
- $rv .= "$html>\n";
+ 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 $v = PublicInbox::Hval->new_msgid($irt);
+ my $html = $v->as_html;
+ my $href = $v->{href};
+ $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 $srch;
-
- my $refs = $hdr->header_raw('References');
- if ($refs) {
- # avoid redundant URLs wasting bandwidth
- my %seen;
- $seen{$irt} = 1 if defined $irt;
- my @refs;
- my @raw_refs = ($refs =~ /<([^>]+)>/g);
- foreach my $ref (@raw_refs) {
- next if $seen{$ref};
- $seen{$ref} = 1;
- push @refs, linkify_ref_nosrch($ref);
- }
+ return $rv if $over;
- if (@refs) {
- $rv .= 'References: '. join("\n\t", @refs) . "\n";
- }
+ $refs //= references($hdr);
+ if (@$refs) {
+ @$refs = map { linkify_ref_no_over($_) } @$refs;
+ $rv .= 'References: '. join("\n\t", @$refs) . "\n";
}
$rv;
}
-sub squote_maybe ($) {
- my ($val) = @_;
- if ($val =~ m{([^\w@\./,\%\+\-])}) {
- $val =~ s/(['!])/'\\$1'/g; # '!' for csh
- return "'$val'";
- }
- $val;
-}
-
-sub mailto_arg_link {
- my ($hdr) = @_;
- my %cc; # everyone else
- my $to; # this is the From address
-
- foreach my $h (qw(From To Cc)) {
- my $v = $hdr->header($h);
- defined($v) && ($v ne '') or next;
- my @addrs = PublicInbox::Address::emails($v);
- foreach my $address (@addrs) {
- my $dst = lc($address);
- $cc{$dst} ||= $address;
- $to ||= $dst;
- }
- }
- my @arg;
-
- my $subj = $hdr->header('Subject') || '';
- $subj = "Re: $subj" unless $subj =~ /\bRe:/i;
- my $mid = $hdr->header_raw('Message-ID');
- push @arg, '--in-reply-to='.ascii_html(squote_maybe(mid_clean($mid)));
- my $irt = uri_escape_utf8($mid);
- delete $cc{$to};
- push @arg, '--to=' . ascii_html($to);
- $to = uri_escape_utf8($to);
- $subj = uri_escape_utf8($subj);
- my $cc = join(',', sort values %cc);
- push @arg, '--cc=' . ascii_html($cc);
- $cc = uri_escape_utf8($cc);
- my $href = "mailto:$to?In-Reply-To=$irt&Cc=${cc}&Subject=$subj";
- $href =~ s/%20/+/g;
-
- (\@arg, ascii_html($href));
-}
-
sub html_footer {
my ($hdr, $standalone, $ctx, $rhref) = @_;
- my $srch = $ctx->{srch} if $ctx;
+ my $ibx = $ctx->{-inbox} if $ctx;
my $upfx = '../';
my $tpfx = '';
my $idx = $standalone ? " index" : '';
my $irt = '';
- if ($idx && $srch) {
+ if ($idx && $ibx->over) {
$idx .= "\n";
thread_skel(\$idx, $ctx, $hdr, $tpfx);
- my $p = $ctx->{parent_msg};
- my $next = $ctx->{next_msg};
- if ($p) {
- $p = PublicInbox::Hval->new_msgid($p);
- $p = $p->as_href;
- $irt = "parent ";
- } else {
- $irt = ' ' x length('parent ');
+ my ($next, $prev);
+ my $parent = ' ';
+ $next = $prev = ' ';
+
+ if (my $n = $ctx->{next_msg}) {
+ $n = PublicInbox::Hval->new_msgid($n)->{href};
+ $next = "next";
}
- if ($next) {
- my $n = PublicInbox::Hval->new_msgid($next)->as_href;
- $irt .= "next ";
- } else {
- $irt .= ' ' x length('next ');
+ my $u;
+ my $par = $ctx->{parent_msg};
+ if ($par) {
+ $u = PublicInbox::Hval->new_msgid($par)->{href};
+ $u = "$upfx$u/";
}
+ if (my $p = $ctx->{prev_msg}) {
+ $prev = PublicInbox::Hval->new_msgid($p)->{href};
+ if ($p && $par && $p eq $par) {
+ $prev = "prev parent';
+ $parent = '';
+ } else {
+ $prev = "prev';
+ $parent = " parent" if $u;
+ }
+ } elsif ($u) { # unlikely
+ $parent = " parent";
+ }
+ $irt = "$next $prev$parent ";
} else {
$irt = '';
}
@@ -659,10 +829,10 @@ sub html_footer {
$irt .= $idx;
}
-sub linkify_ref_nosrch {
+sub linkify_ref_no_over {
my $v = PublicInbox::Hval->new_msgid($_[0]);
my $html = $v->as_html;
- my $href = $v->as_href;
+ my $href = $v->{href};
"<$html>";
}
@@ -673,39 +843,68 @@ sub anchor_for {
sub ghost_parent {
my ($upfx, $mid) = @_;
- # 'subject dummy' is used internally by Mail::Thread
- return '[no common parent]' if ($mid eq 'subject dummy');
$mid = PublicInbox::Hval->new_msgid($mid);
- my $href = $mid->as_href;
+ my $href = $mid->{href};
my $html = $mid->as_html;
qq{[parent not found: <$html>]};
}
sub indent_for {
my ($level) = @_;
- INDENT x ($level - 1);
+ $level ? INDENT x ($level - 1) : '';
}
-sub load_results {
- my ($sres) = @_;
-
- [ map { $_->mini_mime } @{delete $sres->{msgs}} ];
+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 msg_timestamp {
- my ($hdr) = @_;
- my $ts = eval { str2time($hdr->header('Date')) };
- defined($ts) ? $ts : 0;
+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 ($msgs) = @_;
- require PublicInbox::Thread;
- my $th = PublicInbox::Thread->new(@$msgs);
- $th->thread;
- $th->order(*sort_ts);
- $th
+ my ($ctx, $msgs) = @_;
+ require PublicInbox::SearchThread;
+ my $ibx = $ctx->{-inbox};
+ my $rootset = PublicInbox::SearchThread::thread($msgs, *sort_ds, $ibx);
+
+ # 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 {
@@ -714,39 +913,60 @@ sub missing_thread {
PublicInbox::ExtMsg::ext_msg($ctx);
}
-sub _msg_date {
- my ($hdr) = @_;
- my $ts = $hdr->header('X-PI-TS') || msg_timestamp($hdr);
- fmt_ts($ts);
-}
-
sub fmt_ts { POSIX::strftime('%Y-%m-%d %k:%M', gmtime($_[0])) }
-sub _skel_header {
- my ($ctx, $hdr, $level) = @_;
+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 {
+ my ($ctx, $level, $node) = @_;
+ my $smsg = $node->{smsg} or return _skel_ghost($ctx, $level, $node);
my $dst = $ctx->{dst};
my $cur = $ctx->{cur};
- my $mid = mid_clean($hdr->header_raw('Message-ID'));
- my $f = ascii_html($hdr->header('X-PI-From'));
- my $d = _msg_date($hdr) . ' ' . indent_for($level) . th_pfx($level);
+ my $mid = $smsg->{mid};
+
+ if ($level == 0 && $ctx->{skel_dump_roots}++) {
+ $$dst .= 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}) . ' ' . 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;
- } else {
- $attr = '';
}
$ctx->{prev_level} = $level;
if ($cur) {
if ($cur eq $mid) {
delete $ctx->{cur};
- $$dst .= $d;
- $$dst .= "".
+ $$dst .= "$d".
"$attr [this message]\n";
- return;
+ return 1;
+ } else {
+ $ctx->{prev_msg} = $mid;
}
} else {
$ctx->{next_msg} ||= $mid;
@@ -755,175 +975,244 @@ sub _skel_header {
# Subject is never undef, this mail was loaded from
# our Xapian which would've resulted in '' if it were
# really missing (and Filter rejects empty subjects)
- my $s = $hdr->header('Subject');
- my $h = $ctx->{srch}->subject_path($s);
- if ($ctx->{seen}->{$h}) {
- $s = undef;
+ my @subj = split(/ /, subject_normalized($smsg->subject));
+
+ # remove common suffixes from the subject if it matches the previous,
+ # so we do not show redundant text at the end.
+ my $prev_subj = $ctx->{prev_subj} || [];
+ $ctx->{prev_subj} = [ @subj ];
+ my $omit = dedupe_subject($prev_subj, \@subj, '" ');
+ my $end;
+ if (@subj) {
+ my $subj = join(' ', @subj);
+ $subj = ascii_html($subj);
+ obfuscate_addrs($obfs_ibx, $subj) if $obfs_ibx;
+ $end = "$subj $omit$f\n"
} else {
- $ctx->{seen}->{$h} = 1;
- $s = PublicInbox::Hval->new($s);
- $s = $s->as_html;
+ $end = "$f\n";
}
- my $m = PublicInbox::Hval->new_msgid($mid);
+ my $m;
my $id = '';
my $mapping = $ctx->{mapping};
- my $end = defined($s) ? "$s $f\n" : "$f\n";
if ($mapping) {
my $map = $mapping->{$mid};
$id = id_compress($mid, 1);
$m = '#m'.$id;
- $map->[1] = "$d$end";
+ $map->[0] = "$d$end";
$id = "\nid=r".$id;
} else {
- $m = $ctx->{-upfx}.$m->as_href.'/';
+ $m = $ctx->{-upfx}.mid_escape($mid).'/';
}
$$dst .= $d . "" . $end;
+ 1;
}
-sub skel_dump {
+sub _skel_ghost {
my ($ctx, $level, $node) = @_;
- if (my $mime = $node->message) {
- _skel_header($ctx, $mime->header_obj, $level);
+
+ my $mid = $node->{id};
+ my $d = $ctx->{pct} ? ' [irrelevant] ' # search result
+ : ' [not found] ';
+ $d .= indent_for($level) . th_pfx($level);
+ my $upfx = $ctx->{-upfx};
+ my $m = PublicInbox::Hval->new_msgid($mid);
+ my $href = $upfx . $m->{href} . '/';
+ my $html = $m->as_html;
+
+ 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 {
- my $mid = $node->messageid;
- my $dst = $ctx->{dst};
- my $mapping = $ctx->{mapping};
- my $map = $mapping->{$mid} if $mapping;
- if ($mid eq 'subject dummy') {
- my $ncp = "\t[no common parent]\n";
- $map->[1] = $ncp if $map;
- $$dst .= $ncp;
- return;
- }
- my $d = $ctx->{pct} ? ' [irrelevant] ' # search result
- : ' [not found] ';
- $d .= indent_for($level) . th_pfx($level);
- my $upfx = $ctx->{-upfx};
- my $m = PublicInbox::Hval->new_msgid($mid);
- my $href = $upfx . $m->as_href . '/';
- my $html = $m->as_html;
-
- if ($map) {
- my $id = id_compress($mid, 1);
- $map->[1] = $d . qq{<$html>\n};
- $d .= qq{<$html>\n};
- } else {
- $d .= qq{<$html>\n};
- }
- $$dst .= $d;
+ $d .= qq{<$html>\n};
}
+ my $dst = $ctx->{dst};
+ $$dst .= $d;
+ 1;
}
-sub sort_ts {
- sort {
- (eval { $a->topmost->message->header('X-PI-TS') } || 0) <=>
- (eval { $b->topmost->message->header('X-PI-TS') } || 0)
- } @_;
-}
-
-sub _tryload_ghost ($$) {
- my ($srch, $mid) = @_;
- my $smsg = $srch->lookup_mail($mid) or return;
- $smsg->mini_mime;
+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 1 if done, undef if not
-sub add_topic {
+# returns 200 if done, 404 if not
+sub acc_topic {
my ($ctx, $level, $node) = @_;
- my $srch = $ctx->{srch};
- my $mid = $node->messageid;
- my $x = $node->message || _tryload_ghost($srch, $mid);
- my ($subj, $ts);
+ my $mid = $node->{id};
+ my $x = $node->{smsg} || $ctx->{-inbox}->smsg_by_mid($mid);
+ my ($subj, $ds);
+ my $topic;
if ($x) {
- $x = $x->header_obj;
- $subj = $x->header('Subject');
- $subj = $srch->subject_normalized($subj);
- $ts = $x->header('X-PI-TS');
- } else { # ghost message, do not bump level
- $ts = -666;
- $subj = "<$mid>";
- }
- if (++$ctx->{subjs}->{$subj} == 1) {
- push @{$ctx->{order}}, [ $level, $subj ];
- }
- my $exist = $ctx->{latest}->{$subj};
- if (!$exist || $exist->[1] < $ts) {
- $ctx->{latest}->{$subj} = [ $mid, $ts ];
+ $subj = $x->subject;
+ $subj = subject_normalized($subj);
+ $subj = '(no subject)' if $subj eq '';
+ $ds = $x->ds;
+ if ($level == 0) {
+ $topic = [ $ds, 1, { $subj => $mid }, $subj ];
+ $ctx->{-cur_topic} = $topic;
+ push @{$ctx->{order}}, $topic;
+ return 1;
+ }
+
+ $topic = $ctx->{-cur_topic}; # should never be undef
+ $topic->[0] = $ds if $ds > $topic->[0];
+ $topic->[1]++;
+ my $seen = $topic->[2];
+ if (scalar(@$topic) == 3) { # parent was a ghost
+ push @$topic, $subj;
+ } elsif (!$seen->{$subj}) {
+ push @$topic, $level, $subj;
+ }
+ $seen->{$subj} = $mid; # latest for subject
+ } else { # ghost message
+ return 1 if $level != 0; # ignore child ghosts
+ $topic = [ -666, 0, {} ];
+ $ctx->{-cur_topic} = $topic;
+ push @{$ctx->{order}}, $topic;
}
+ 1;
}
-sub emit_topics {
+sub dump_topics {
my ($ctx) = @_;
- my $order = $ctx->{order};
- my $subjs = $ctx->{subjs};
- my $latest = $ctx->{latest};
- my $fh = $ctx->{fh};
- return $fh->write("\n[No topics in range]
") unless scalar @$order;
- my $pfx;
- my $prev = 0;
- my $prev_attr = '';
- my $cur;
- my @recent;
- while (defined(my $info = shift @$order)) {
- my ($level, $subj) = @$info;
- my $n = delete $subjs->{$subj};
- my ($mid, $ts) = @{delete $latest->{$subj}};
- my $href = PublicInbox::Hval->new_msgid($mid)->as_href;
- $pfx = indent_for($level);
- my $nl = $level == $prev ? "\n" : '';
- if ($nl && $cur) {
- push @recent, $cur;
- $cur = undef;
- }
- $cur ||= [ $ts, '' ];
- $cur->[0] = $ts if $ts > $cur->[0];
- $cur->[1] .= $nl . $pfx . th_pfx($level);
- if ($ts == -666) { # ghost
- $cur->[1] .= ghost_parent('', $mid) . "\n";
- next; # child will have mbox / atom link
- }
+ my $order = delete $ctx->{order}; # [ ds, subj1, subj2, subj3, ... ]
+ if (!@$order) {
+ $ctx->{-html_tip} = '[No topics in range]
';
+ return 404;
+ }
- $subj = PublicInbox::Hval->new($subj)->as_html;
- $cur->[1] .= "$subj\n";
- $ts = fmt_ts($ts);
- my $attr = " $ts UTC";
+ 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, $top, @ex) = @$topic;
+ @$topic = ();
+ next unless defined $top; # ghost topic
+ my $mid = delete $seen->{$top};
+ my $href = mid_escape($mid);
+ my $prev_subj = [ split(/ /, $top) ];
+ $top = PublicInbox::Hval->new($top)->as_html;
+ $ds = fmt_ts($ds);
# $n isn't the total number of posts on the topic,
# just the number of posts in the current results window
- $n = $n == 1 ? '' : " ($n+ messages)";
-
- if ($level == 0 || $attr ne $prev_attr) {
- my $mbox = qq(mbox.gz);
- my $atom = qq(Atom);
- $pfx .= INDENT if $level > 0;
- $cur->[1] .= $pfx . $attr . $n . " - $mbox / $atom\n";
- $prev_attr = $attr;
+ 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\n" .
+ " $ds UTC $n - $mbox / $atom\n";
+ for (my $i = 0; $i < scalar(@ex); $i += 2) {
+ my $level = $ex[$i];
+ my $subj = $ex[$i + 1];
+ $mid = delete $seen->{$subj};
+ my @subj = split(/ /, subject_normalized($subj));
+ my @next_prev = @subj; # full copy
+ my $omit = dedupe_subject($prev_subj, \@subj, ' "');
+ $prev_subj = \@next_prev;
+ $subj = ascii_html(join(' ', @subj));
+ obfuscate_addrs($obfs_ibx, $subj) if $obfs_ibx;
+ $href = mid_escape($mid);
+ $s .= indent_for($level) . TCHILD;
+ $s .= "$subj$omit\n";
}
+ push @out, $s;
}
- push @recent, $cur if $cur;
- @recent = map { $_->[1] } sort { $b->[0] <=> $a->[0] } @recent;
- $fh->write(join('', @recent) . '
');
+ $ctx->{-html_tip} = '' . join("\n", @out) . '
';
+ 200;
}
-sub emit_index_topics {
- my ($ctx) = @_;
- my ($off) = (($ctx->{qp}->{o} || '0') =~ /(\d+)/);
- $ctx->{order} = [];
- $ctx->{subjs} = {};
- $ctx->{latest} = {};
- my $max = 25;
- my %opts = ( offset => $off, limit => $max * 4 );
- while (scalar @{$ctx->{order}} < $max) {
- my $sres = $ctx->{srch}->query('', \%opts);
- my $nr = scalar @{$sres->{msgs}} or last;
- $sres = load_results($sres);
- walk_thread(thread_results($sres), $ctx, *add_topic);
- $opts{offset} += $nr;
+sub ts2str ($) {
+ my ($ts) = @_;
+ POSIX::strftime('%Y%m%d%H%M%S', gmtime($ts));
+}
+
+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!;
+ }
+ $msgs;
+}
- emit_topics($ctx);
- $opts{offset};
+sub index_topics {
+ my ($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($ctx), *index_nav);
}
sub thread_adj_level {
@@ -951,9 +1240,10 @@ sub thread_adj_level {
}
sub ghost_index_entry {
- my ($ctx, $level, $mid) = @_;
+ my ($ctx, $level, $node) = @_;
my ($beg, $end) = thread_adj_level($ctx, $level);
- $beg . ''. ghost_parent($ctx->{-upfx}, $mid) . '
' . $end;
+ $beg . ''. ghost_parent($ctx->{-upfx}, $node->{id})
+ . '
' . $end;
}
1;