#
# Used by the web interface to link to messages outside of the our
@@ -8,16 +8,17 @@
package PublicInbox::ExtMsg;
use strict;
use warnings;
-use PublicInbox::Hval;
-use PublicInbox::MID qw/mid2path/;
-use PublicInbox::WwwStream;
+use PublicInbox::Hval qw(ascii_html prurl mid_href);
+use PublicInbox::WwwStream qw(html_oneshot);
+use PublicInbox::Smsg;
+our $MIN_PARTIAL_LEN = 16;
# TODO: user-configurable
-our @EXT_URL = (
+our @EXT_URL = map { ascii_html($_) } (
# leading "//" denotes protocol-relative (http:// or https://)
'//marc.info/?i=%s',
'//www.mail-archive.com/search?l=mid&q=%s',
- 'http://mid.gmane.org/%s',
+ 'nntp://news.gmane.io/%s',
'https://lists.debian.org/msgid-search/%s',
'//docs.FreeBSD.org/cgi/mid.cgi?db=mid&id=%s',
'https://www.w3.org/mid/%s',
@@ -26,94 +27,160 @@ our @EXT_URL = (
'doc-url=/lurker&format=en.html&query=id:%s'
);
-sub ext_msg {
- my ($ctx) = @_;
- my $cur = $ctx->{-inbox};
- my $mid = $ctx->{mid};
+sub PARTIAL_MAX () { 100 }
- eval { require PublicInbox::Msgmap };
- my $have_mm = $@ ? 0 : 1;
- my (@ibx, @found);
+sub search_partial ($$) {
+ my ($ibx, $mid) = @_;
+ return if length($mid) < $MIN_PARTIAL_LEN;
+ my $srch = $ibx->search or return;
+ my $opt = { limit => PARTIAL_MAX, mset => 2 };
+ my @try = ("m:$mid*");
+ my $chop = $mid;
+ if ($chop =~ s/(\W+)(\w*)\z//) {
+ my ($delim, $word) = ($1, $2);
+ if (length($word)) {
+ push @try, "m:$chop$delim";
+ push @try, "m:$chop$delim*";
+ }
+ push @try, "m:$chop";
+ push @try, "m:$chop*";
+ }
- $ctx->{www}->{pi_config}->each_inbox(sub {
- my ($other) = @_;
- return if $other->{name} eq $cur->{name} || !$other->base_url;
+ # break out long words individually to search for, because
+ # too many messages begin with "Pine.LNX." (or "alpine" or "nycvar")
+ if ($mid =~ /\w{9,}/) {
+ my @long = ($mid =~ m!(\w{3,})!g);
+ push(@try, join(' ', map { "m:$_" } @long));
+
+ # is the last element long enough to not trigger excessive
+ # wildcard matches?
+ if (length($long[-1]) > 8) {
+ $long[-1] .= '*';
+ push(@try, join(' ', map { "m:$_" } @long));
+ }
+ }
+
+ foreach my $m (@try) {
+ # If Xapian can't handle the wildcard since it
+ # has too many results. $@ can be
+ # Search::Xapian::QueryParserError or even:
+ # "something terrible happened at ../Search/Xapian/Enquire.pm"
+ my $mset = eval { $srch->mset($m, $opt) } or next;
+ my @mids = map {
+ $_->{mid}
+ } @{$srch->mset_to_smsg($ibx, $mset)};
+ return \@mids if scalar(@mids);
+ }
+}
+
+sub ext_msg_i {
+ my ($other, $ctx) = @_;
+
+ return if $other->{name} eq $ctx->{-inbox}->{name} || !$other->base_url;
+
+ my $mm = $other->mm or return;
+
+ # try to find the URL with Msgmap to avoid forking
+ my $num = $mm->num_for($ctx->{mid});
+ if (defined $num) {
+ push @{$ctx->{found}}, $other;
+ } else {
+ # no point in trying the fork fallback if we
+ # know Xapian is up-to-date but missing the
+ # message in the current repo
+ push @{$ctx->{again}}, $other;
+ }
+}
+
+sub ext_msg_step {
+ my ($pi_cfg, $section, $ctx) = @_;
+ if (defined($section)) {
+ return if $section !~ m!\Apublicinbox\.([^/]+)\z!;
+ my $ibx = $pi_cfg->lookup_name($1) or return;
+ ext_msg_i($ibx, $ctx);
+ } else { # undef == "EOF"
+ finalize_exact($ctx);
+ }
+}
- my $mm = $other->mm or return;
+sub ext_msg {
+ my ($ctx) = @_;
+ sub {
+ $ctx->{-wcb} = $_[0]; # HTTP server write callback
- # try to find the URL with Msgmap to avoid forking
- my $num = $mm->num_for($mid);
- if (defined $num) {
- push @found, $other;
+ if ($ctx->{env}->{'pi-httpd.async'}) {
+ require PublicInbox::ConfigIter;
+ my $iter = PublicInbox::ConfigIter->new(
+ $ctx->{www}->{pi_config},
+ \&ext_msg_step, $ctx);
+ $iter->event_step;
} else {
- # no point in trying the fork fallback if we
- # know Xapian is up-to-date but missing the
- # message in the current repo
- push @ibx, $other;
+ $ctx->{www}->{pi_config}->each_inbox(\&ext_msg_i, $ctx);
+ finalize_exact($ctx);
}
- });
+ };
+}
- return exact($ctx, \@found, $mid) if @found;
+# called via PublicInbox::DS->EventLoop
+sub event_step {
+ my ($ctx, $sync) = @_;
+ # can't find a partial match in current inbox, try the others:
+ my $ibx = shift @{$ctx->{again}} or return finalize_partial($ctx);
+ my $mids = search_partial($ibx, $ctx->{mid}) or
+ return ($sync ? undef : PublicInbox::DS::requeue($ctx));
+ $ctx->{n_partial} += scalar(@$mids);
+ push @{$ctx->{partial}}, [ $ibx, $mids ];
+ $ctx->{n_partial} >= PARTIAL_MAX ? finalize_partial($ctx)
+ : ($sync ? undef : PublicInbox::DS::requeue($ctx));
+}
+
+sub finalize_exact {
+ my ($ctx) = @_;
+
+ return $ctx->{-wcb}->(exact($ctx)) if $ctx->{found};
# fall back to partial MID matching
- my $n_partial = 0;
- my @partial;
-
- if (my $mm = $cur->mm) {
- my $tmp_mid = $mid;
- my $res = $mm->mid_prefixes($tmp_mid, 100);
- if ($res && scalar(@$res)) {
- $n_partial += scalar(@$res);
- push @partial, [ $cur, $res ];
- # fixup common errors:
- } elsif ($tmp_mid =~ s,/[tTf],,) {
- $res = $mm->mid_prefixes($tmp_mid, 100);
- if ($res && scalar(@$res)) {
- $n_partial += scalar(@$res);
- push @partial, [ $cur, $res ];
- }
+ my $mid = $ctx->{mid};
+ my $cur = $ctx->{-inbox};
+ my $mids = search_partial($cur, $mid);
+ if ($mids) {
+ $ctx->{n_partial} = scalar(@$mids);
+ push @{$ctx->{partial}}, [ $cur, $mids ];
+ } elsif ($ctx->{again} && length($mid) >= $MIN_PARTIAL_LEN) {
+ bless $ctx, __PACKAGE__;
+ if ($ctx->{env}->{'pi-httpd.async'}) {
+ $ctx->event_step;
+ return;
}
- }
- # can't find a partial match in current inbox, try the others:
- if (!$n_partial && length($mid) >= 16) {
- my $tmp_mid = $mid;
-again:
- foreach my $ibx (@ibx) {
- my $mm = $ibx->mm or next;
- my $res = $mm->mid_prefixes($tmp_mid, 100);
- if ($res && scalar(@$res)) {
- $n_partial += scalar(@$res);
- push @partial, [ $ibx, $res ];
- last if $n_partial >= 100;
- }
- }
- # fixup common errors:
- if (!$n_partial && $tmp_mid =~ s,/[tTf],,) {
- goto again;
- }
+ # synchronous fall-through
+ $ctx->event_step while @{$ctx->{again}};
}
+ finalize_partial($ctx);
+}
+sub finalize_partial {
+ my ($ctx) = @_;
+ my $mid = $ctx->{mid};
my $code = 404;
- my $h = PublicInbox::Hval->new_msgid($mid);
- my $href = $h->{href};
- my $html = $h->as_html;
+ my $href = mid_href($mid);
+ my $html = ascii_html($mid);
my $title = "<$html> not found";
my $s = "Message-ID <$html>\nnot found\n";
- if ($n_partial) {
+ if (my $n_partial = $ctx->{n_partial}) {
$code = 300;
my $es = $n_partial == 1 ? '' : 'es';
+ $n_partial .= '+' if ($n_partial == PARTIAL_MAX);
$s .= "\n$n_partial partial match$es found:\n\n";
- my $cur_name = $cur->{name};
- foreach my $pair (@partial) {
+ my $cur_name = $ctx->{-inbox}->{name};
+ foreach my $pair (@{$ctx->{partial}}) {
my ($ibx, $res) = @$pair;
my $env = $ctx->{env} if $ibx->{name} eq $cur_name;
my $u = $ibx->base_url($env) or next;
foreach my $m (@$res) {
- my $p = PublicInbox::Hval->new_msgid($m);
- my $r = $p->{href};
- my $t = $p->as_html;
- $s .= qq{$u$t/\n};
+ my $href = mid_href($m);
+ my $html = ascii_html($m);
+ $s .= qq{$u$html/\n};
}
}
}
@@ -125,7 +192,7 @@ again:
$ctx->{-html_tip} = $s .= '
';
$ctx->{-title_html} = $title;
$ctx->{-upfx} = '../';
- PublicInbox::WwwStream->response($ctx, $code);
+ $ctx->{-wcb}->(html_oneshot($ctx, $code));
}
sub ext_urls {
@@ -136,7 +203,7 @@ sub ext_urls {
my $env = $ctx->{env};
my $e = "\nPerhaps try an external site:\n\n";
foreach my $url (@EXT_URL) {
- my $u = PublicInbox::Hval::prurl($env, $url);
+ my $u = prurl($env, $url);
my $r = sprintf($u, $href);
my $t = sprintf($u, $html);
$e .= qq{$t\n};
@@ -147,10 +214,11 @@ sub ext_urls {
}
sub exact {
- my ($ctx, $found, $mid) = @_;
- my $h = PublicInbox::Hval->new_msgid($mid);
- my $href = $h->{href};
- my $html = $h->as_html;
+ my ($ctx) = @_;
+ my $mid = $ctx->{mid};
+ my $found = $ctx->{found};
+ my $href = mid_href($mid);
+ my $html = ascii_html($mid);
my $title = "<$html> found in ";
my $end = @$found == 1 ? 'another inbox' : 'other inboxes';
$ctx->{-title_html} = $title . $end;
@@ -164,7 +232,7 @@ sub exact {
qq($u$html/\n)
} @$found),
$ext_urls, '');
- PublicInbox::WwwStream->response($ctx, $code);
+ html_oneshot($ctx, $code);
}
1;