require POSIX;
use Time::Local qw(timegm);
+use constant COLS => 72;
use constant INDENT => ' ';
use constant TCHILD => '` ';
sub th_pfx ($) { $_[0] == 0 ? '' : TCHILD };
my ($arg, $link, $reply_to_all) =
PublicInbox::Reply::mailto_arg_link($ibx, $hdr);
+ if (ref($arg) eq 'SCALAR') {
+ return '<pre id=R>'.ascii_html($$arg).'</pre>';
+ }
# mailto: link only works if address obfuscation is disabled
if ($link) {
$refs->[-1];
}
+sub fold_addresses ($) {
+ return $_[0] if length($_[0]) <= COLS;
+ # try to fold on commas after non-word chars before $lim chars,
+ # Try to get the "," preceeded by ">" or ")", but avoid folding
+ # on the comma where somebody uses "Lastname, Firstname".
+ # We also try to keep the last and penultimate addresses in
+ # the list on the same line if possible, hence the extra \z
+ # Fall back to folding on spaces at $lim + 1 chars
+ my $lim = COLS - 8; # 8 = "\t" display width
+ my $too_long = $lim + 1;
+ $_[0] =~ s/\s*\z//s; # Email::Simple doesn't strip trailing spaces
+ $_[0] = join("\n\t",
+ ($_[0] =~ /(.{0,$lim}\W(?:,|\z)|
+ .{1,$lim}(?:,|\z)|
+ .{1,$lim}|
+ .{$too_long,}?)(?:\s|\z)/xgo));
+}
+
sub _hdr_names_html ($$) {
my ($hdr, $field) = @_;
my $val = $hdr->header($field) or return '';
my @tocc;
my $mime = $smsg->{mime};
my $hdr = $mime->header_obj;
- foreach my $f (qw(To Cc)) {
- my $dst = _hdr_names_html($hdr, $f);
- if ($dst ne '') {
- obfuscate_addrs($obfs_ibx, $dst) if $obfs_ibx;
- push @tocc, "$f: $dst";
- }
- }
my $from = _hdr_names_html($hdr, 'From');
obfuscate_addrs($obfs_ibx, $from) if $obfs_ibx;
$rv .= "From: $from @ ".fmt_ts($smsg->ds)." UTC";
my $mhref = $upfx . mid_escape($mid_raw) . '/';
$rv .= qq{ (<a\nhref="$mhref">permalink</a> / };
$rv .= qq{<a\nhref="${mhref}raw">raw</a>)\n};
- $rv .= ' '.join('; +', @tocc) . "\n" if @tocc;
+ my $to = fold_addresses(_hdr_names_html($hdr, 'To'));
+ my $cc = fold_addresses(_hdr_names_html($hdr, 'Cc'));
+ my ($tlen, $clen) = (length($to), length($cc));
+ my $to_cc = '';
+ if (($tlen + $clen) > COLS) {
+ $to_cc .= ' To: '.$to."\n" if $tlen;
+ $to_cc .= ' Cc: '.$cc."\n" if $clen;
+ } else {
+ if ($tlen) {
+ $to_cc .= ' To: '.$to;
+ $to_cc .= '; <b>+Cc:</b> '.$cc if $clen;
+ } else {
+ $to_cc .= ' Cc: '.$cc if $clen;
+ }
+ $to_cc .= "\n";
+ }
+ obfuscate_addrs($obfs_ibx, $to_cc) if $obfs_ibx;
+ $rv .= $to_cc;
my $mapping = $ctx->{mapping};
if (!$mapping && (defined($irt) || defined($irt = in_reply_to($hdr)))) {
while (@q) {
my ($level, $node, $i) = splice(@q, 0, 3);
defined $node or next;
- $cb->($ctx, $level, $node, $i);
+ $cb->($ctx, $level, $node, $i) or return;
++$level;
$i = 0;
unshift @q, map { ($level, $_, $i++) } @{$node->{children}};
my ($part, $depth) = @$p; # attachment @idx is unused
my $ct = $part->content_type || 'text/plain';
my $fn = $part->filename;
+ my ($s, $err) = msg_part_text($part, $ct);
- if ($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!
- my $err = $@;
- if ($err) {
- if ($ct =~ m!\btext/plain\b!i) {
- # Try to assume UTF-8 because Alpine seems to
- # do wacky things and set charset=X-UNKNOWN
- $part->charset_set('UTF-8');
- $s = eval { $part->body_str };
-
- # If forcing charset=UTF-8 failed,
- # attach_link will warn further down...
- $s = $part->body if $@;
- } else {
- return attach_link($upfx, $ct, $p, $fn);
- }
- }
+ return attach_link($upfx, $ct, $p, $fn) unless defined $s;
my @lines = split(/^/m, $s);
$s = '';
if (defined($fn) || $depth > 0 || $err) {
+ # badly-encoded message with $err? tell the world about it!
$s .= attach_link($upfx, $ct, $p, $fn, $err);
$s .= "\n";
}
$ctx->{-upfx} = '../';
}
my @title;
- foreach my $h (qw(From To Cc Subject Date)) {
- my $v = $hdr->header($h);
- defined($v) && ($v ne '') or next;
+ my $v;
+ if (defined($v = $hdr->header('From'))) {
$v = PublicInbox::Hval->new($v);
-
- if ($h eq 'From') {
- my @n = PublicInbox::Address::names($v->raw);
- $title[1] = ascii_html(join(', ', @n));
- obfuscate_addrs($obfs_ibx, $title[1]) if $obfs_ibx;
- } elsif ($h eq 'Subject') {
- $title[0] = $v->as_html;
- if ($srch) {
- $rv .= qq($h: <a\nhref="#r"\nid=t>);
- $rv .= $v->as_html . "</a>\n";
- next;
- }
- }
+ my @n = PublicInbox::Address::names($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 $h (qw(To Cc)) {
+ defined($v = $hdr->header($h)) or next;
+ fold_addresses($v);
+ $v = ascii_html($v);
obfuscate_addrs($obfs_ibx, $v) if $obfs_ibx;
- $rv .= "$h: $v\n";
-
+ $rv .= "$h: $v\n" if $v ne '';
+ }
+ if (defined($v = $hdr->header('Subject')) && ($v ne '')) {
+ $v = ascii_html($v);
+ obfuscate_addrs($obfs_ibx, $v) if $obfs_ibx;
+ if ($srch) {
+ $rv .= qq(Subject: <a\nhref="#r"\nid=t>$v</a>\n);
+ } else {
+ $rv .= "Subject: $v\n";
+ }
+ $title[0] = $v;
+ } else { # dummy anchor for thread skeleton at bottom of page
+ $rv .= qq(<a\nhref="#r"\nid=t></a>) if $srch;
+ $title[0] = '(no subject)';
+ }
+ if (defined($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)';
$ctx->{-title_html} = join(' - ', @title);
foreach (@$mids) {
my $mid = PublicInbox::Hval->new_msgid($_) ;
my ($hdr, $srch) = @_;
my $rv = '';
- my $irt = in_reply_to($hdr);
+ my $refs = references($hdr);
+ my $irt = pop @$refs;
if (defined $irt) {
my $v = PublicInbox::Hval->new_msgid($irt);
my $html = $v->as_html;
# 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);
- }
-
- if (@refs) {
- $rv .= 'References: '. join("\n\t", @refs) . "\n";
- }
+ if (@$refs) {
+ @$refs = map { linkify_ref_nosrch($_) } @$refs;
+ $rv .= 'References: '. join("\n\t", @$refs) . "\n";
}
$rv;
}
$level ? INDENT x ($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, $msgs) = @_;
require PublicInbox::SearchThread;
- PublicInbox::SearchThread::thread($msgs, *sort_ds, $ctx->{-inbox});
+ 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 {
my $cur = $ctx->{cur};
my $mid = $smsg->{mid};
+ if ($level == 0 && $ctx->{skel_dump_roots}++) {
+ $$dst .= delete $ctx->{sl_note} || '';
+ }
+
my $f = ascii_html($smsg->from_name);
my $obfs_ibx = $ctx->{-obfs_ibx};
obfuscate_addrs($obfs_ibx, $f) if $obfs_ibx;
delete $ctx->{cur};
$$dst .= "<b>$d<a\nid=r\nhref=\"#t\">".
"$attr [this message]</a></b>\n";
- return;
+ return 1;
} else {
$ctx->{prev_msg} = $mid;
}
$m = $ctx->{-upfx}.mid_escape($mid).'/';
}
$$dst .= $d . "<a\nhref=\"$m\"$id>" . $end;
+ 1;
}
sub _skel_ghost {
}
my $dst = $ctx->{dst};
$$dst .= $d;
+ 1;
}
sub sort_ds {
$topic = [ $ds, 1, { $subj => $mid }, $subj ];
$ctx->{-cur_topic} = $topic;
push @{$ctx->{order}}, $topic;
- return;
+ return 1;
}
$topic = $ctx->{-cur_topic}; # should never be undef
}
$seen->{$subj} = $mid; # latest for subject
} else { # ghost message
- return if $level != 0; # ignore child ghosts
+ return 1 if $level != 0; # ignore child ghosts
$topic = [ -666, 0, {} ];
$ctx->{-cur_topic} = $topic;
push @{$ctx->{order}}, $topic;
}
+ 1;
}
sub dump_topics {