-# Copyright (C) 2014-2015 all contributors <meta@public-inbox.org>
-# License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt)
+# Copyright (C) 2014-2019 all contributors <meta@public-inbox.org>
+# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
#
# Used for displaying the HTML web interface.
# See Documentation/design_www.txt for this.
package PublicInbox::View;
use strict;
use warnings;
-use URI::Escape qw/uri_escape_utf8/;
-use Date::Parse qw/str2time/;
-use Encode::MIME::Header;
-use Plack::Util;
-use PublicInbox::Hval qw/ascii_html/;
+use bytes (); # only for bytes::length
+use PublicInbox::MsgTime qw(msg_datestamp);
+use PublicInbox::Hval qw/ascii_html obfuscate_addrs/;
use PublicInbox::Linkify;
-use PublicInbox::MID qw/mid_clean id_compress mid_mime/;
+use PublicInbox::MID qw/id_compress mid_escape mids references/;
use PublicInbox::MsgIter;
use PublicInbox::Address;
use PublicInbox::WwwStream;
+use PublicInbox::Reply;
+use PublicInbox::ViewDiff qw(flush_diff);
require POSIX;
-
+use Time::Local qw(timegm);
+use PublicInbox::SearchMsg qw(subject_normalized);
+use constant COLS => 72;
use constant INDENT => ' ';
use constant TCHILD => '` ';
sub th_pfx ($) { $_[0] == 0 ? '' : TCHILD };
# public functions: (unstable)
+
sub msg_html {
- my ($ctx, $mime, $footer) = @_;
+ my ($ctx, $mime, $more, $smsg) = @_;
my $hdr = $mime->header_obj;
- my $tip = _msg_html_prepare($hdr, $ctx);
+ my $ibx = $ctx->{-inbox};
+ $ctx->{-obfs_ibx} = $ibx->{obfuscate} ? $ibx : undef;
+ my $tip = _msg_html_prepare($hdr, $ctx, $more, 0);
+ my $end = 2;
PublicInbox::WwwStream->response($ctx, 200, sub {
my ($nr, undef) = @_;
if ($nr == 1) {
- $tip . multipart_text_as_html($mime, '') .
- '</pre><hr />'
- } elsif ($nr == 2) {
+ # $more cannot be true w/o $smsg being defined:
+ my $upfx = $more ? '../'.mid_escape($smsg->mid).'/' : '';
+ $tip . multipart_text_as_html($mime, $upfx, $ctx) .
+ '</pre><hr>'
+ } elsif ($more && @$more) {
+ ++$end;
+ msg_html_more($ctx, $more, $nr);
+ } elsif ($nr == $end) {
# fake an EOF if generating the footer fails;
# we want to at least show the message if something
# here crashes:
});
}
+sub msg_page {
+ my ($ctx) = @_;
+ my $mid = $ctx->{mid};
+ my $ibx = $ctx->{-inbox};
+ my ($first, $more);
+ my $smsg;
+ if (my $over = $ibx->over) {
+ my ($id, $prev);
+ $smsg = $over->next_by_mid($mid, \$id, \$prev);
+ $first = $ibx->msg_by_smsg($smsg) if $smsg;
+ if ($first) {
+ my $next = $over->next_by_mid($mid, \$id, \$prev);
+ $more = [ $id, $prev, $next ] if $next;
+ }
+ return unless $first;
+ } else {
+ $first = $ibx->msg_by_mid($mid) or return;
+ }
+ msg_html($ctx, PublicInbox::MIME->new($first), $more, $smsg);
+}
+
+sub msg_html_more {
+ my ($ctx, $more, $nr) = @_;
+ my $str = eval {
+ my ($id, $prev, $smsg) = @$more;
+ my $mid = $ctx->{mid};
+ my $ibx = $ctx->{-inbox};
+ $smsg = $ibx->smsg_mime($smsg);
+ my $next = $ibx->over->next_by_mid($mid, \$id, \$prev);
+ @$more = $next ? ($id, $prev, $next) : ();
+ if ($smsg) {
+ my $mime = $smsg->{mime};
+ my $upfx = '../' . mid_escape($smsg->mid) . '/';
+ _msg_html_prepare($mime->header_obj, $ctx, $more, $nr) .
+ multipart_text_as_html($mime, $upfx, $ctx) .
+ '</pre><hr>'
+ } else {
+ '';
+ }
+ };
+ if ($@) {
+ warn "Error lookup up additional messages: $@\n";
+ $str = '<pre>Error looking up additional messages</pre>';
+ }
+ $str;
+}
+
# /$INBOX/$MESSAGE_ID/#R
sub msg_reply {
my ($ctx, $hdr) = @_;
my $p_url =
'https://en.wikipedia.org/wiki/Posting_style#Interleaved_style';
- my ($arg, $link) = mailto_arg_link($hdr);
+ my $info = '';
+ my $ibx = $ctx->{-inbox};
+ if (my $url = $ibx->{infourl}) {
+ $url = PublicInbox::Hval::prurl($ctx->{env}, $url);
+ $info = qq(\n List information: <a\nhref="$url">$url</a>\n);
+ }
+
+ my ($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) {
+ $link = <<EOF;
+
+* If your mail client supports setting the <b>In-Reply-To</b> header
+ via mailto: links, try the <a
+href="$link">mailto: link</a>
+EOF
+ }
+
push @$arg, '/path/to/YOUR_REPLY';
- $arg = join(" \\\n ", '', @$arg);
+ $arg = ascii_html(join(" \\\n ", '', @$arg));
<<EOF
-<pre
-id=R>You may reply publically to <a
-href=#t>this message</a> via
-plain-text email using any one of the following methods:
+<hr><pre
+id=R><b>Reply instructions:</b>
+
+You may reply publically to <a
+href=#t>this message</a> via plain-text email
+using any one of the following methods:
* Save the following mbox file, import it into your mail client,
- and reply-to-all from there: <a
+ and $reply_to_all from there: <a
href=raw>mbox</a>
Avoid top-posting and favor interleaved quoting:
<a
href="$p_url">$p_url</a>
-
-* Reply to all the recipients using the <b>--to</b>, <b>--cc</b>,
- and <b>--in-reply-to</b> switches of git-send-email(1):
+$info
+* Reply using the <b>--to</b>, <b>--cc</b>, and <b>--in-reply-to</b>
+ switches of git-send-email(1):
git send-email$arg
<a
href="$se_url">$se_url</a>
-
-* If your mail client supports setting the <b>In-Reply-To</b> header
- via mailto: links, try the <a
-href="$link">mailto: link</a></pre>
+$link</pre>
EOF
}
sub in_reply_to {
my ($hdr) = @_;
- my $irt = $hdr->header_raw('In-Reply-To');
-
- return mid_clean($irt) if (defined $irt);
+ my $refs = references($hdr);
+ $refs->[-1];
+}
- my $refs = $hdr->header_raw('References');
- if ($refs && $refs =~ /<([^>]+)>\s*\z/s) {
- return $1;
- }
- undef;
+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 ($$) {
+sub _hdr_names_html ($$) {
my ($hdr, $field) = @_;
my $val = $hdr->header($field) or return '';
ascii_html(join(', ', PublicInbox::Address::names($val)));
# this is already inside a <pre>
sub index_entry {
- my ($mime, $ctx, $more) = @_;
- my $srch = $ctx->{srch};
- my $hdr = $mime->header_obj;
- my $subj = $hdr->header('Subject');
-
- my $mid_raw = mid_clean(mid_mime($mime));
+ my ($smsg, $ctx, $more) = @_;
+ my $subj = $smsg->subject;
+ my $mid_raw = $smsg->mid;
my $id = id_compress($mid_raw, 1);
my $id_m = 'm'.$id;
- my $mid = PublicInbox::Hval->new_msgid($mid_raw);
my $root_anchor = $ctx->{root_anchor} || '';
- my $irt = in_reply_to($hdr);
+ my $irt;
+ my $obfs_ibx = $ctx->{-obfs_ibx};
+ $subj = '(no subject)' if $subj eq '';
my $rv = "<a\nhref=#e$id\nid=m$id>*</a> ";
$subj = '<b>'.ascii_html($subj).'</b>';
+ obfuscate_addrs($obfs_ibx, $subj) if $obfs_ibx;
$subj = "<u\nid=u>$subj</u>" if $root_anchor eq $id_m;
$rv .= $subj . "\n";
- $rv .= _th_index_lite($mid_raw, $irt, $id, $ctx);
+ $rv .= _th_index_lite($mid_raw, \$irt, $id, $ctx);
my @tocc;
- foreach my $f (qw(To Cc)) {
- my $dst = _hdr_names($hdr, $f);
- push @tocc, "$f: $dst" if $dst ne '';
- }
- $rv .= "From: "._hdr_names($hdr, 'From').' @ '._msg_date($hdr)." UTC";
+ my $ds = $smsg->ds; # for v1 non-Xapian/SQLite users
+ # deleting {mime} is critical to memory use,
+ # the rest of the fields saves about 400K as we iterate across 1K msgs
+ my ($mime) = delete @$smsg{qw(mime ds ts blob subject)};
+
+ my $hdr = $mime->header_obj;
+ my $from = _hdr_names_html($hdr, 'From');
+ obfuscate_addrs($obfs_ibx, $from) if $obfs_ibx;
+ $rv .= "From: $from @ ".fmt_ts($ds)." UTC";
my $upfx = $ctx->{-upfx};
- my $mhref = $upfx . $mid->as_href . '/';
+ 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 && $irt) {
+ if (!$mapping && (defined($irt) || defined($irt = in_reply_to($hdr)))) {
my $mirt = PublicInbox::Hval->new_msgid($irt);
- my $href = $upfx . $mirt->as_href . '/';
+ my $href = $upfx . $mirt->{href}. '/';
my $html = $mirt->as_html;
- $rv .= qq(In-Reply-To: <<a\nhref="$href/">$html</a>>\n)
+ $rv .= qq(In-Reply-To: <<a\nhref="$href">$html</a>>\n)
}
$rv .= "\n";
# scan through all parts, looking for displayable text
- msg_iter($mime, sub { $rv .= add_text_body($mhref, $_[0]) });
+ my $ibx = $ctx->{-inbox};
+ msg_iter($mime, sub { $rv .= add_text_body($mhref, $ctx, $_[0]) });
# add the footer
$rv .= "\n<a\nhref=#$id_m\nid=e$id>^</a> ".
"<a\nhref=\"$mhref\">permalink</a>" .
- " / <a\nhref=\"${mhref}raw\">raw</a>" .
- " / <a\nhref=\"${mhref}#R\">reply</a>";
+ " <a\nhref=\"${mhref}raw\">raw</a>" .
+ " <a\nhref=\"${mhref}#R\">reply</a>";
+
+ my $hr;
if (my $pct = $ctx->{pct}) { # used by SearchView.pm
- $rv .= " [relevance $pct->{$mid_raw}%]";
+ $rv .= "\t[relevance $pct->{$mid_raw}%]";
+ $hr = 1;
} elsif ($mapping) {
- my $threaded = 'threaded';
+ my $nested = 'nested';
my $flat = 'flat';
my $end = '';
if ($ctx->{flat}) {
+ $hr = 1;
$flat = "<b>$flat</b>";
} else {
- $threaded = "<b>$threaded</b>";
+ $nested = "<b>$nested</b>";
}
- $rv .= " / [<a\nhref=\"${mhref}T/#u\">$flat</a>";
- $rv .= "|<a\nhref=\"${mhref}t/#u\">$threaded</a>]";
- $rv .= " / <a\nhref=#r$id>$ctx->{s_nr}</a>";
+ $rv .= "\t[<a\nhref=\"${mhref}T/#u\">$flat</a>";
+ $rv .= "|<a\nhref=\"${mhref}t/#u\">$nested</a>]";
+ $rv .= " <a\nhref=#r$id>$ctx->{s_nr}</a>";
+ } else {
+ $hr = $ctx->{-hr};
}
- $rv .= $more ? "\n\n" : "\n";
+ $rv .= $more ? '</pre><hr><pre>' : '</pre>' if $hr;
+ $rv;
}
sub pad_link ($$;$) {
my $rv = '';
my $mapping = $ctx->{mapping} or return $rv;
my $pad = ' ';
- # map = [children, attr, node, idx, level]
- my $map = $mapping->{$mid_raw};
- my $nr_c = scalar @{$map->[0]};
+ my $mid_map = $mapping->{$mid_raw};
+ defined $mid_map or
+ return 'public-inbox BUG: '.ascii_html($mid_raw).' not mapped';
+ my ($attr, $node, $idx, $level) = @$mid_map;
+ my $children = $node->{children};
+ my $nr_c = scalar @$children;
my $nr_s = 0;
- my $level = $map->[4];
- my $idx = $map->[3];
- if (defined $irt) {
- my $irt_map = $mapping->{$irt};
- my $siblings = $irt_map->[0];
+ my $siblings;
+ if (my $smsg = $node->{smsg}) {
+ # delete saves about 200KB on a 1K message thread
+ if (my $refs = delete $smsg->{references}) {
+ ($$irt) = ($refs =~ m/<([^>]+)>\z/);
+ }
+ }
+ my $irt_map = $mapping->{$$irt} if defined $$irt;
+ if (defined $irt_map) {
+ $siblings = $irt_map->[1]->{children};
$nr_s = scalar(@$siblings) - 1;
- $rv .= $pad . $irt_map->[1];
+ $rv .= $pad . $irt_map->[0];
if ($idx > 0) {
my $prev = $siblings->[$idx - 1];
- my $pmid = $prev->messageid;
+ my $pmid = $prev->{id};
if ($idx > 2) {
my $s = ($idx - 1). ' preceding siblings ...';
$rv .= pad_link($pmid, $level, $s);
} elsif ($idx == 2) {
- my $ppmid = $siblings->[0]->messageid;
- $rv .= $pad . $mapping->{$ppmid}->[1];
+ my $ppmid = $siblings->[0]->{id};
+ $rv .= $pad . $mapping->{$ppmid}->[0];
}
- $rv .= $pad . $mapping->{$pmid}->[1];
+ $rv .= $pad . $mapping->{$pmid}->[0];
}
}
my $s_s = nr_to_s($nr_s, 'sibling', 'siblings');
my $s_c = nr_to_s($nr_c, 'reply', 'replies');
- my $this = $map->[1];
- $this =~ s!\n\z!</b>\n!s;
- $this =~ s!<a\nhref.*</a> !!s; # no point in duplicating subject
- $rv .= "<b>@ $this";
- my $node = $map->[2];
- if (my $child = $node->child) {
- my $cmid = $child->messageid;
- $rv .= $pad . $mapping->{$cmid}->[1];
+ $attr =~ s!\n\z!</b>\n!s;
+ $attr =~ s!<a\nhref.*</a> !!s; # no point in duplicating subject
+ $attr =~ s!<a\nhref=[^>]+>([^<]+)</a>!$1!s; # no point linking to self
+ $rv .= "<b>@ $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 ."<a\nhref=#r$id>$s_s, $s_c; $ctx->{s_nr}</a>\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 . '<pre>' . index_entry($mime, $ctx, 0) . '</pre>' . $end;
+ $beg . '<pre>' . index_entry($smsg, $ctx, 0) . '</pre>' . $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));
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 = '<hr /><pre>';
+ my $skel = '<hr><pre>';
$skel .= $nr == 1 ? 'only message in thread' : 'end of thread';
- $skel .= ", back to <a\nhref=\"../../\">index</a>";
- $skel .= "\n<a\nid=t>$nr+ messages in thread:</a> (download: ";
- $skel .= "<a\nhref=\"../t.mbox.gz\">mbox.gz</a>";
+ $skel .= ", back to <a\nhref=\"../../\">index</a>\n\n";
+ $skel .= "<b\nid=t>Thread overview:</b> ";
+ $skel .= $nr == 1 ? '(only message)' : "$nr+ messages";
+ $skel .= " (download: <a\nhref=\"../t.mbox.gz\">mbox.gz</a>";
$skel .= " / follow: <a\nhref=\"../t.atom\">Atom feed</a>)\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);
$skel .= '</pre>';
- return stream_thread($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'));
- $ctx->{-html_tip} = '<pre>'.index_entry($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);
+ $ctx->{-html_tip} = '<pre>'.index_entry($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;
$msgs = undef;
- '</pre>'.$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;
}
# 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($$quot);
# we use a <span> here to allow users to specify their own
# color for quoted text
$rv = $l->linkify_2(ascii_html($rv));
+ $$quot = undef;
$$s .= qq(<span\nclass="q">) . $rv . '</span>'
}
-sub attach_link ($$$$) {
- my ($upfx, $ct, $p, $fn) = @_;
+sub attach_link ($$$$;$) {
+ 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<a\nhref="$upfx$idx-$sfn">[-- Attachment #$idx: );
+ my $ret = qq($nl<a\nhref="$upfx$idx-$sfn">);
+ 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, "</a>\n");
+ $desc = ascii_html($desc);
+ $ret .= ($desc eq '') ? "$ts --]" : "$desc --]\n[-- $ts --]";
+ $ret .= "</a>\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 = "<pre\nid=b>"; # 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 .=
+"<pre>WARNING: multiple messages refer to this Message-ID\n</pre>";
+ }
+ $rv .= "<pre\nid=b>"; # anchor for body start
+ } else {
+ $rv .= '<pre>';
+ }
+ 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 $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));
- } 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;
- $v =~ s/(\@[^,]+,) /$1\n\t/g if ($h eq 'Cc' || $h eq 'To');
- $rv .= "$h: $v\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)) {
+ 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" if $v ne '';
+ }
+ if (defined($v = $hdr->header('Subject')) && ($v ne '')) {
+ $v = ascii_html($v);
+ obfuscate_addrs($obfs_ibx, $v) if $obfs_ibx;
+ if ($over) {
+ $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 $over;
+ $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);
- $rv .= 'Message-ID: <' . $mid->as_html . '> ';
- $rv .= "(<a\nhref=\"raw\">raw</a>)\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 .= "<a\nhref=\"../$href/\">";
+ $rv .= "<$mhtml></a> ";
+ $rv .= "(<a\nhref=\"../$href/raw\">raw</a>)\n";
+ } else {
+ $rv .= "Message-ID: <$mhtml> ";
+ $rv .= "(<a\nhref=\"raw\">raw</a>)\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(<a\nhref="${tpfx}T/#u">expand</a> ) .
- qq(/ <a\nhref="${tpfx}t.mbox.gz">mbox.gz</a> ) .
- qq(/ <a\nhref="${tpfx}t.atom">Atom feed</a>);
+ my $mid = mids($hdr)->[0];
+ my $ibx = $ctx->{-inbox};
+ my ($nr, $msgs) = $ibx->over->get_thread($mid);
+ my $expand = qq(expand[<a\nhref="${tpfx}T/#u">flat</a>) .
+ qq(|<a\nhref="${tpfx}t/#u">nested</a>] ) .
+ qq(<a\nhref="${tpfx}t.mbox.gz">mbox.gz</a> ) .
+ qq(<a\nhref="${tpfx}t.atom">Atom feed</a>);
my $parent = in_reply_to($hdr);
+ $$dst .= "\n<b>Thread overview: </b>";
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";
- $$dst .= qq! / <a\nhref="#b">[top]</a>)\n!;
+ $$dst .= "$nr+ messages / $expand";
+ $$dst .= qq! <a\nhref="#b">top</a>\n!;
- my $subj = $srch->subject_path($hdr->header('Subject'));
- $ctx->{seen} = { $subj => 1 };
+ my $subj = $hdr->header('Subject');
+ defined $subj or $subj = '';
+ $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);
+ my $refs = references($hdr);
+ my $irt = pop @$refs;
if (defined $irt) {
- my $v = PublicInbox::Hval->new_msgid($irt, 1);
+ my $v = PublicInbox::Hval->new_msgid($irt);
my $html = $v->as_html;
- my $href = $v->as_href;
+ my $href = $v->{href};
$rv .= "In-Reply-To: <";
$rv .= "<a\nhref=\"../$href/\">$html</a>>\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";
- }
+ 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 ? " <a\nhref=\"$upfx\">index</a>" : '';
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 = "<a\nhref=\"$upfx$p/\"\nrel=prev>parent</a> ";
- } 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 = "<a\nhref=\"$upfx$n/\"\nrel=next>next</a>";
}
- if ($next) {
- my $n = PublicInbox::Hval->new_msgid($next)->as_href;
- $irt .= "<a\nhref=\"$upfx$n/\"\nrel=next>next</a> ";
- } 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 = "<a\nhref=\"$upfx$prev/\"\n" .
+ 'rel=prev>prev parent</a>';
+ $parent = '';
+ } else {
+ $prev = "<a\nhref=\"$upfx$prev/\"\n" .
+ 'rel=prev>prev</a>';
+ $parent = " <a\nhref=\"$u\">parent</a>" if $u;
+ }
+ } elsif ($u) { # unlikely
+ $parent = " <a\nhref=\"$u\"\nrel=prev>parent</a>";
+ }
+ $irt = "$next $prev$parent ";
} else {
$irt = '';
}
$irt .= $idx;
}
-sub linkify_ref_nosrch {
- my $v = PublicInbox::Hval->new_msgid($_[0], 1);
+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};
"<<a\nhref=\"../$href/\">$html</a>>";
}
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: <<a\nhref="$upfx$href/">$html</a>>]};
}
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 {
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 .= "<b><a\nid=r\nhref=\"#t\">".
+ $$dst .= "<b>$d<a\nid=r\nhref=\"#t\">".
"$attr [this message]</a></b>\n";
- return;
+ return 1;
+ } else {
+ $ctx->{prev_msg} = $mid;
}
} else {
$ctx->{next_msg} ||= $mid;
# Subject is never undef, this mail was loaded from
# our Xapian which would've resulted in '' if it were
# really missing (and Filter rejects empty subjects)
- my $s = $hdr->header('Subject');
- my $h = $ctx->{srch}->subject_path($s);
- if ($ctx->{seen}->{$h}) {
- $s = undef;
+ my @subj = split(/ /, subject_normalized($smsg->subject));
+
+ # remove common suffixes from the subject if it matches the previous,
+ # so we do not show redundant text at the end.
+ my $prev_subj = $ctx->{prev_subj} || [];
+ $ctx->{prev_subj} = [ @subj ];
+ my $omit = dedupe_subject($prev_subj, \@subj, '" ');
+ my $end;
+ if (@subj) {
+ my $subj = join(' ', @subj);
+ $subj = ascii_html($subj);
+ obfuscate_addrs($obfs_ibx, $subj) if $obfs_ibx;
+ $end = "$subj</a> $omit$f\n"
} else {
- $ctx->{seen}->{$h} = 1;
- $s = PublicInbox::Hval->new($s);
- $s = $s->as_html;
+ $end = "$f</a>\n";
}
- my $m = PublicInbox::Hval->new_msgid($mid);
+ my $m;
my $id = '';
my $mapping = $ctx->{mapping};
- my $end = defined($s) ? "$s</a> $f\n" : "$f</a>\n";
if ($mapping) {
my $map = $mapping->{$mid};
$id = id_compress($mid, 1);
$m = '#m'.$id;
- $map->[1] = "$d<a\nhref=\"$m\">$end";
+ $map->[0] = "$d<a\nhref=\"$m\">$end";
$id = "\nid=r".$id;
} else {
- $m = $ctx->{-upfx}.$m->as_href.'/';
+ $m = $ctx->{-upfx}.mid_escape($mid).'/';
}
$$dst .= $d . "<a\nhref=\"$m\"$id>" . $end;
+ 1;
}
-sub skel_dump {
+sub _skel_ghost {
my ($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] = $d . qq{<<a\nhref=#r$id>$html</a>>\n};
+ $d .= qq{<<a\nhref="$href"\nid=r$id>$html</a>>\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;
- $$dst .= $ncp;
- return;
- }
- my $d = $ctx->{pct} ? ' [irrelevant] ' # search result
- : ' [not found] ';
- $d .= indent_for($level) . th_pfx($level);
- my $upfx = $ctx->{-upfx};
- my $m = PublicInbox::Hval->new_msgid($mid);
- my $href = $upfx . $m->as_href . '/';
- my $html = $m->as_html;
-
- if ($map) {
- my $id = id_compress($mid, 1);
- $map->[1] = $d . qq{<<a\nhref=#r$id>$html</a>>\n};
- $d .= qq{<<a\nhref="$href"\nid=r$id>$html</a>>\n};
- } else {
- $d .= qq{<<a\nhref="$href">$html</a>>\n};
- }
- $$dst .= $d;
+ $d .= qq{<<a\nhref="$href">$html</a>>\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 $fh->write("\n[No topics in range]</pre>") unless scalar @$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}};
- $mid = PublicInbox::Hval->new_msgid($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} = '<pre>[No topics in range]</pre>';
+ return 404;
+ }
- $subj = PublicInbox::Hval->new($subj)->as_html;
- $cur->[1] .= "<a\nhref=\"$mid/T/#u\"><b>$subj</b></a>\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 $mbox = qq(<a\nhref="$mid/t.mbox.gz">mbox.gz</a>);
- my $atom = qq(<a\nhref="$mid/t.atom">Atom</a>);
- $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 $mbox = qq(<a\nhref="$href/t.mbox.gz">mbox.gz</a>);
+ my $atom = qq(<a\nhref="$href/t.atom">Atom</a>);
+ my $s = "<a\nhref=\"$href/T/$anchor\">$top</a>\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;
+ $s .= "<a\nhref=\"$href/T/#u\">$subj</a>$omit\n";
}
+ push @out, $s;
}
- push @recent, $cur if $cur;
- @recent = map { $_->[1] } sort { $b->[0] <=> $a->[0] } @recent;
- $fh->write(join('', @recent) . '</pre>');
+ $ctx->{-html_tip} = '<pre>' . join("\n", @out) . '</pre>';
+ 200;
}
-sub emit_index_topics {
- my ($ctx) = @_;
- my ($off) = (($ctx->{cgi}->param('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 ($$) {
+ my ($ctx, $latest) = @_;
+ delete $ctx->{qp} or return;
+ my $next = $ctx->{next_page} || '';
+ my $prev = $ctx->{prev_page} || '';
+ if ($prev) {
+ $next = $next ? "$next " : ' ';
+ $prev .= qq! <a\nhref='$latest'>latest</a>!;
+ }
+ "<hr><pre>page: $next$prev</pre>";
+}
+
+sub index_nav { # callback for WwwStream
+ my (undef, $ctx) = @_;
+ pagination_footer($ctx, '.')
+}
+
+sub paginate_recent ($$) {
+ 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);
+ $ctx->{next_page} = qq!<a\nhref="?t=$s"\nrel=next>next</a>!;
}
+ if (defined($newest) && (defined($before) || defined($after))) {
+ my $s = ts2str($newest);
+ $ctx->{prev_page} = qq!<a\nhref="?t=$s-"\nrel=prev>prev</a>!;
+ }
+ $msgs;
+}
- emit_topics($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($ctx), *index_nav);
}
sub thread_adj_level {
}
sub ghost_index_entry {
- my ($ctx, $level, $mid) = @_;
+ my ($ctx, $level, $node) = @_;
my ($beg, $end) = thread_adj_level($ctx, $level);
- $beg . '<pre>'. ghost_parent($ctx->{-upfx}, $mid) . '</pre>' . $end;
+ $beg . '<pre>'. ghost_parent($ctx->{-upfx}, $node->{id})
+ . '</pre>' . $end;
}
1;