this message 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: mbox
Avoid top-posting and favor interleaved quoting:
$p_url
$info
-* Reply to all the recipients using the --to, --cc,
- and --in-reply-to switches of git-send-email(1):
+* Reply using the --to, --cc, and --in-reply-to
+ switches of git-send-email(1):
git send-email$arg
$se_url
-
-* If your mail client supports setting the In-Reply-To header
- via mailto: links, try the mailto: link
+$link
EOF
}
sub in_reply_to {
my ($hdr) = @_;
- my $irt = $hdr->header_raw('In-Reply-To');
-
- return mid_clean($irt) if defined $irt && $irt ne '';
+ 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)));
@@ -117,34 +202,55 @@ sub nr_to_s ($$$) {
# this is already inside a
sub index_entry {
- my ($mime, $ctx, $more) = @_;
+ my ($smsg, $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 $subj = $smsg->subject;
+ my $mid_raw = $smsg->mid;
my $id = id_compress($mid_raw, 1);
my $id_m = 'm'.$id;
my $root_anchor = $ctx->{root_anchor} || '';
my $irt;
+ my $obfs_ibx = $ctx->{-obfs_ibx};
my $rv = "* ";
$subj = ''.ascii_html($subj).'';
+ obfuscate_addrs($obfs_ibx, $subj) if $obfs_ibx;
$subj = "$subj" if $root_anchor eq $id_m;
$rv .= $subj . "\n";
$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_escape($mid_raw) . '/';
$rv .= qq{ (permalink / };
$rv .= qq{raw)\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 .= '; +Cc: '.$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)))) {
@@ -156,7 +262,8 @@ sub index_entry {
$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^ ".
@@ -169,17 +276,17 @@ sub index_entry {
$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 = "$flat";
} else {
- $threaded = "$threaded";
+ $nested = "$nested";
}
$rv .= "\t[$flat";
- $rv .= "|$threaded]";
+ $rv .= "|$nested]";
$rv .= " $ctx->{s_nr}";
} else {
$hr = $ctx->{-hr};
@@ -201,13 +308,19 @@ sub _th_index_lite {
my $rv = '';
my $mapping = $ctx->{mapping} or return $rv;
my $pad = ' ';
- my ($attr, $node, $idx, $level) = @{$mapping->{$mid_raw}};
+ 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 $siblings;
if (my $smsg = $node->{smsg}) {
- ($$irt) = (($smsg->{references} || '') =~ m/<([^>]+)>\z/);
+ # 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 $$irt;
if (defined $irt_map) {
@@ -260,12 +373,12 @@ sub _th_index_lite {
}
sub walk_thread {
- my ($th, $ctx, $cb) = @_;
- my @q = map { (0, $_, -1) } @{$th->{rootset}};
+ my ($rootset, $ctx, $cb) = @_;
+ my @q = map { (0, $_, -1) } @$rootset;
while (@q) {
my ($level, $node, $i) = splice(@q, 0, 3);
defined $node or next;
- $cb->($ctx, $level, $node, $i);
+ $cb->($ctx, $level, $node, $i) or return;
++$level;
$i = 0;
unshift @q, map { ($level, $_, $i++) } @{$node->{children}};
@@ -279,29 +392,30 @@ sub pre_thread {
}
sub thread_index_entry {
- my ($ctx, $level, $mime) = @_;
+ my ($ctx, $level, $smsg) = @_;
my ($beg, $end) = thread_adj_level($ctx, $level);
- $beg . '' . index_entry($mime, $ctx, 0) . '
' . $end;
+ $beg . '' . index_entry($smsg, $ctx, 0) . '
' . $end;
}
sub stream_thread ($$) {
- 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;
my $cl = $level + 1;
unshift @q, map { ($cl, $_) } @{$node->{children}};
- $mime = $inbox->msg_by_smsg($node->{smsg}) and last;
+ $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) {
@@ -309,10 +423,8 @@ sub stream_thread ($$) {
my $node = shift @q or next;
my $cl = $level + 1;
unshift @q, map { ($cl, $_) } @{$node->{children}};
- my $mid = $node->{id};
- if ($mime = $inbox->msg_by_smsg($node->{smsg})) {
- $mime = Email::MIME->new($mime);
- return thread_index_entry($ctx, $level, $mime);
+ if ($smsg = $ibx->smsg_mime($node->{smsg})) {
+ return thread_index_entry($ctx, $level, $smsg);
} else {
return ghost_index_entry($ctx, $level, $node);
}
@@ -327,9 +439,8 @@ sub stream_thread ($$) {
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 $srch = $ctx->{srch};
+ my ($nr, $msgs) = $srch->get_thread($mid);
return missing_thread($ctx) if $nr == 0;
my $skel = '
';
$skel .= $nr == 1 ? 'only message in thread' : 'end of thread';
@@ -345,49 +456,46 @@ sub thread_html {
$ctx->{prev_attr} = '';
$ctx->{prev_level} = 0;
$ctx->{root_anchor} = anchor_for($mid);
- $ctx->{seen} = {};
$ctx->{mapping} = {};
$ctx->{s_nr} = "$nr+ messages in thread";
- my $th = thread_results($msgs);
- walk_thread($th, $ctx, *pre_thread);
+ my $rootset = thread_results($ctx, $msgs);
+
+ # reduce hash lookups in pre_thread->skel_dump
+ my $ibx = $ctx->{-inbox};
+ $ctx->{-obfs_ibx} = $ibx->{obfuscate} ? $ibx : undef;
+ walk_thread($rootset, $ctx, *pre_thread);
+
$skel .= '
';
- return stream_thread($th, $ctx) unless $ctx->{flat};
+ return stream_thread($rootset, $ctx) unless $ctx->{flat};
# flat display: lazy load the full message from smsg
- my $inbox = $ctx->{-inbox};
- my $mime;
- while ($mime = shift @$msgs) {
- $mime = $inbox->msg_by_smsg($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;
+ 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) {
- $mime = $inbox->msg_by_smsg($mime) 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;
$skel;
});
}
sub multipart_text_as_html {
- my ($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;
}
@@ -396,12 +504,12 @@ 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;
$$s .= qq() . $rv . ''
}
@@ -434,109 +542,160 @@ sub attach_link ($$$$;$) {
}
$ret .= "[-- Attachment #$idx: ";
my $ts = "Type: $ct, Size: $size bytes";
+ $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 || 'text/plain';
my $fn = $part->filename;
-
- 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);
+ 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;
+ };
+
+ # 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 || $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 ($hdr, $ctx, $more, $nr) = @_;
my $srch = $ctx->{srch} if $ctx;
my $atom = '';
- my $rv = ""; # anchor for body start
-
+ 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 ($srch) {
$ctx->{-upfx} = '../';
}
my @title;
- my $mid = mid_clean($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: );
- $rv .= $v->as_html . "\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 ($srch) {
+ $rv .= qq(Subject: $v\n);
+ } else {
+ $rv .= "Subject: $v\n";
+ }
+ $title[0] = $v;
+ } else { # dummy anchor for thread skeleton at bottom of page
+ $rv .= qq() 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);
- $rv .= 'Message-ID: <' . $mid->as_html . '> ';
- $rv .= "(raw)\n";
+ 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, $srch);
$rv .= "\n";
}
@@ -544,37 +703,43 @@ sub _msg_html_prepare {
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 ($nr, $msgs) = $srch->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 };
+ my $subj = $hdr->header('Subject');
+ defined $subj or $subj = '';
+ $ctx->{prev_subj} = [ split(/ /, $srch->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
+ my $ibx = $ctx->{-inbox};
+ $ctx->{-obfs_ibx} = $ibx->{obfuscate} ? $ibx : undef;
+ walk_thread(thread_results($ctx, $msgs), $ctx, *skel_dump);
+
$ctx->{parent_msg} = $parent;
}
@@ -582,7 +747,8 @@ sub _parent_headers {
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;
@@ -595,69 +761,13 @@ sub _parent_headers {
# 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;
}
-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='.squote_maybe(mid_clean($mid));
- my $irt = mid_escape($mid);
- delete $cc{$to};
- push @arg, "--to=$to";
- $to = uri_escape_utf8($to);
- $subj = uri_escape_utf8($subj);
- my @cc = sort values %cc;
- push(@arg, map { "--cc=$_" } @cc);
- my $cc = uri_escape_utf8(join(',', @cc));
- my $href = "mailto:$to?In-Reply-To=$irt&Cc=${cc}&Subject=$subj";
-
- (\@arg, ascii_html($href));
-}
-
sub html_footer {
my ($hdr, $standalone, $ctx, $rhref) = @_;
@@ -729,28 +839,59 @@ sub ghost_parent {
sub indent_for {
my ($level) = @_;
- INDENT x ($level - 1);
+ $level ? INDENT x ($level - 1) : '';
}
-sub load_results {
- my ($sres) = @_;
-
- [ map { $_->ensure_metadata; $_ } @{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) = @_;
+ my ($ctx, $msgs) = @_;
require PublicInbox::SearchThread;
- my $th = PublicInbox::SearchThread->new($msgs);
- $th->thread;
- $th->order(*sort_ts);
- $th
+ 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 {
@@ -759,14 +900,27 @@ 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 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);
@@ -774,8 +928,16 @@ sub skel_dump {
my $dst = $ctx->{dst};
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 $d = fmt_ts($smsg->{ts}) . ' ' . indent_for($level) . th_pfx($level);
+ 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;
@@ -789,7 +951,7 @@ sub skel_dump {
delete $ctx->{cur};
$$dst .= "$d".
"$attr [this message]\n";
- return;
+ return 1;
} else {
$ctx->{prev_msg} = $mid;
}
@@ -800,19 +962,25 @@ sub skel_dump {
# Subject is never undef, this mail was loaded from
# our Xapian which would've resulted in '' if it were
# really missing (and Filter rejects empty subjects)
- my $s = $smsg->subject;
- my $h = $ctx->{srch}->subject_path($s);
- if ($ctx->{seen}->{$h}) {
- $s = undef;
+ my @subj = split(/ /, $ctx->{srch}->subject_normalized($smsg->subject));
+
+ # remove common suffixes from the subject if it matches the previous,
+ # so we do not show redundant text at the end.
+ my $prev_subj = $ctx->{prev_subj} || [];
+ $ctx->{prev_subj} = [ @subj ];
+ my $omit = dedupe_subject($prev_subj, \@subj, '" ');
+ my $end;
+ if (@subj) {
+ my $subj = join(' ', @subj);
+ $subj = ascii_html($subj);
+ obfuscate_addrs($obfs_ibx, $subj) if $obfs_ibx;
+ $end = "$subj $omit$f\n"
} else {
- $ctx->{seen}->{$h} = 1;
- $s = PublicInbox::Hval->new($s);
- $s = $s->as_html;
+ $end = "$f\n";
}
my $m;
my $id = '';
my $mapping = $ctx->{mapping};
- my $end = defined($s) ? "$s $f\n" : "$f\n";
if ($mapping) {
my $map = $mapping->{$mid};
$id = id_compress($mid, 1);
@@ -823,6 +991,7 @@ sub skel_dump {
$m = $ctx->{-upfx}.mid_escape($mid).'/';
}
$$dst .= $d . "" . $end;
+ 1;
}
sub _skel_ghost {
@@ -848,12 +1017,13 @@ sub _skel_ghost {
}
my $dst = $ctx->{dst};
$$dst .= $d;
+ 1;
}
-sub sort_ts {
+sub sort_ds {
[ sort {
- (eval { $a->topmost->{smsg}->ts } || 0) <=>
- (eval { $b->topmost->{smsg}->ts } || 0)
+ (eval { $a->topmost->{smsg}->ds } || 0) <=>
+ (eval { $b->topmost->{smsg}->ds } || 0)
} @{$_[0]} ];
}
@@ -863,22 +1033,22 @@ sub acc_topic {
my ($ctx, $level, $node) = @_;
my $srch = $ctx->{srch};
my $mid = $node->{id};
- my $x = $node->{smsg} || $srch->lookup_mail($mid);
- my ($subj, $ts);
+ my $x = $node->{smsg} || $ctx->{-inbox}->smsg_by_mid($mid);
+ my ($subj, $ds);
my $topic;
if ($x) {
$subj = $x->subject;
$subj = $srch->subject_normalized($subj);
- $ts = $x->ts;
+ $ds = $x->ds;
if ($level == 0) {
- $topic = [ $ts, 1, { $subj => $mid }, $subj ];
+ $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
- $topic->[0] = $ts if $ts > $topic->[0];
+ $topic->[0] = $ds if $ds > $topic->[0];
$topic->[1]++;
my $seen = $topic->[2];
if (scalar(@$topic) == 3) { # parent was a ghost
@@ -888,32 +1058,37 @@ sub acc_topic {
}
$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 {
my ($ctx) = @_;
- my $order = delete $ctx->{order}; # [ ts, subj1, subj2, subj3, ... ]
+ my $order = delete $ctx->{order}; # [ ds, subj1, subj2, subj3, ... ]
if (!@$order) {
$ctx->{-html_tip} = '[No topics in range]
';
return 404;
}
my @out;
+ my $ibx = $ctx->{-inbox};
+ my $obfs_ibx = $ibx->{obfuscate} ? $ibx : undef;
+ my $srch = $ctx->{srch};
# sort by recency, this allows new posts to "bump" old topics...
foreach my $topic (sort { $b->[0] <=> $a->[0] } @$order) {
- my ($ts, $n, $seen, $top, @ex) = @$topic;
+ 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;
- $ts = fmt_ts($ts);
+ $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
@@ -928,16 +1103,21 @@ sub dump_topics {
my $mbox = qq(mbox.gz);
my $atom = qq(Atom);
- my $s = "$top\n" .
- " $ts UTC $n - $mbox / $atom\n";
+ my $s = "$top\n" .
+ " $ds UTC $n - $mbox / $atom\n";
for (my $i = 0; $i < scalar(@ex); $i += 2) {
my $level = $ex[$i];
- my $sub = $ex[$i + 1];
- $mid = delete $seen->{$sub};
- $sub = PublicInbox::Hval->new($sub)->as_html;
+ my $subj = $ex[$i + 1];
+ $mid = delete $seen->{$subj};
+ my @subj = split(/ /, $srch->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 .= "$sub\n";
+ $s .= "$subj$omit\n";
}
push @out, $s;
}
@@ -945,44 +1125,81 @@ sub dump_topics {
200;
}
+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) = @_;
- delete $ctx->{qp} or return;
- my ($next, $prev);
- $next = $prev = ' ';
- my $latest = '';
-
- my $next_o = $ctx->{-next_o};
- if ($next_o) {
- $next = qq!next!;
- }
- if (my $cur_o = $ctx->{-cur_o}) {
- $latest = qq! latest!;
-
- my $o = $cur_o - ($next_o - $cur_o);
- if ($o > 0) {
- $prev = qq!prev!;
- } elsif ($o == 0) {
- $prev = qq!prev!;
+ pagination_footer($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(\d{8,14})-// and $after = str2ts($1);
+ $t =~ /\A(\d{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;
}
}
- "
page: $next $prev$latest
";
+ if (defined($oldest) && $more) {
+ my $s = ts2str($oldest);
+ $ctx->{next_page} = qq!next!;
+ }
+ if (defined($newest) && (defined($before) || defined($after))) {
+ my $s = ts2str($newest);
+ $ctx->{prev_page} = qq!prev!;
+ }
+ $msgs;
}
sub index_topics {
my ($ctx) = @_;
- my ($off) = (($ctx->{qp}->{o} || '0') =~ /(\d+)/);
- my $opts = { offset => $off, limit => 200 };
-
- $ctx->{order} = [];
- my $sres = $ctx->{srch}->query('', $opts);
- my $nr = scalar @{$sres->{msgs}};
- if ($nr) {
- $sres = load_results($sres);
- walk_thread(thread_results($sres), $ctx, *acc_topic);
+ my $msgs = paginate_recent($ctx, 200); # 200 is our window
+ if (@$msgs) {
+ walk_thread(thread_results($ctx, $msgs), $ctx, *acc_topic);
}
- $ctx->{-next_o} = $off+ $nr;
- $ctx->{-cur_o} = $off;
PublicInbox::WwwStream->response($ctx, dump_topics($ctx), *index_nav);
}