X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FExtMsg.pm;h=51e7799de8640929be9538713d3f2ee7e1b1081d;hb=a46893a2b5dabfdbcf7b593ac19967daecfb1772;hp=019d50a457aa98a609e4bbca0da0c43213d7c0ce;hpb=af70d4db0d539887d1688befff3f2faf0eeda30d;p=public-inbox.git diff --git a/lib/PublicInbox/ExtMsg.pm b/lib/PublicInbox/ExtMsg.pm index 019d50a4..51e7799d 100644 --- a/lib/PublicInbox/ExtMsg.pm +++ b/lib/PublicInbox/ExtMsg.pm @@ -1,5 +1,5 @@ -# Copyright (C) 2015 all contributors -# License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt) +# Copyright (C) 2015-2018 all contributors +# License: AGPL-3.0+ # # Used by the web interface to link to messages outside of the our # public-inboxes. Mail threads may cross projects/threads; so @@ -14,44 +14,82 @@ use PublicInbox::WwwStream; # TODO: user-configurable our @EXT_URL = ( - 'http://mid.gmane.org/%s', - 'https://lists.debian.org/msgid-search/%s', # leading "//" denotes protocol-relative (http:// or https://) - '//mid.mail-archive.com/%s', '//marc.info/?i=%s', + '//www.mail-archive.com/search?l=mid&q=%s', + 'nntp://news.gmane.org/%s', + 'https://lists.debian.org/msgid-search/%s', + '//docs.FreeBSD.org/cgi/mid.cgi?db=mid&id=%s', + 'https://www.w3.org/mid/%s', + 'http://www.postgresql.org/message-id/%s', + 'https://lists.debconf.org/cgi-lurker/keyword.cgi?'. + 'doc-url=/lurker&format=en.html&query=id:%s' ); +sub PARTIAL_MAX () { 100 } + +sub search_partial ($$) { + my ($srch, $mid) = @_; + 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*"; + } + + # 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) { + my $mset = eval { $srch->query($m, $opt) }; + if (ref($@) eq 'Search::Xapian::QueryParserError') { + # If Xapian can't handle the wildcard since it + # has too many results. + next; + } + my @mids = map { + my $doc = $_->get_document; + PublicInbox::SearchMsg->load_doc($doc)->mid; + } $mset->items; + return \@mids if scalar(@mids); + } +} + sub ext_msg { my ($ctx) = @_; - my $pi_config = $ctx->{pi_config}; my $cur = $ctx->{-inbox}; my $mid = $ctx->{mid}; - my $env = $ctx->{env}; - eval { require PublicInbox::Search }; - my $have_xap = $@ ? 0 : 1; - my (@nox, @ibx, @found); + eval { require PublicInbox::Msgmap }; + my $have_mm = $@ ? 0 : 1; + my (@ibx, @found); - $pi_config->each_inbox(sub { + $ctx->{www}->{pi_config}->each_inbox(sub { my ($other) = @_; return if $other->{name} eq $cur->{name} || !$other->base_url; - my $s = $other->search; - if (!$s) { - push @nox, $other; - return; - } - - # try to find the URL with Xapian to avoid forking - my $doc_id = eval { $s->find_unique_doc_id('mid', $mid) }; - if ($@) { - # xapian not configured properly for this repo - push @nox, $other; - return; - } + my $mm = $other->mm or return; - # maybe we found it! - if (defined $doc_id) { + # try to find the URL with Msgmap to avoid forking + my $num = $mm->num_for($mid); + if (defined $num) { push @found, $other; } else { # no point in trying the fork fallback if we @@ -63,59 +101,46 @@ sub ext_msg { return exact($ctx, \@found, $mid) if @found; - # Xapian not installed or configured for some repos, - # do a full MID check (this is expensive...): - if (@nox) { - my $path = mid2path($mid); - foreach my $other (@nox) { - my (undef, $type, undef) = $other->path_check($path); - - if ($type && $type eq 'blob') { - push @found, $other; - } - } - } - return exact($ctx, \@found, $mid) if @found; - # fall back to partial MID matching - my $n_partial = 0; my @partial; + my $n_partial = 0; + my $srch = $cur->search; + my $mids = search_partial($srch, $mid) if $srch; + if ($mids) { + $n_partial = scalar(@$mids); + push @partial, [ $cur, $mids ]; + } - eval { require PublicInbox::Msgmap }; - my $have_mm = $@ ? 0 : 1; - if ($have_mm) { - my $tmp_mid = $mid; -again: - unshift @ibx, $cur; + # can't find a partial match in current inbox, try the others: + if (!$n_partial && length($mid) >= 16) { foreach my $ibx (@ibx) { - my $mm = $ibx->mm or next; - if (my $res = $mm->mid_prefixes($tmp_mid)) { - $n_partial += scalar(@$res); - push @partial, [ $ibx, $res ]; - } - } - # fixup common errors: - if (!$n_partial && $tmp_mid =~ s,/[tTf],,) { - goto again; + $srch = $ibx->search or next; + $mids = search_partial($srch, $mid) or next; + $n_partial += scalar(@$mids); + push @partial, [ $ibx, $mids]; + last if $n_partial >= PARTIAL_MAX; } } my $code = 404; my $h = PublicInbox::Hval->new_msgid($mid); - my $href = $h->as_href; + my $href = $h->{href}; my $html = $h->as_html; my $title = "<$html> not found"; my $s = "
Message-ID <$html>\nnot found\n";
 	if ($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 $pair (@partial) {
 			my ($ibx, $res) = @$pair;
-			my $u = $ibx->base_url or next;
+			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->as_href;
+				my $r = $p->{href};
 				my $t = $p->as_html;
 				$s .= qq{$u$t/\n};
 			}
@@ -153,7 +178,7 @@ sub ext_urls {
 sub exact {
 	my ($ctx, $found, $mid) = @_;
 	my $h = PublicInbox::Hval->new_msgid($mid);
-	my $href = $h->as_href;
+	my $href = $h->{href};
 	my $html = $h->as_html;
 	my $title = "<$html> found in ";
 	my $end = @$found == 1 ? 'another inbox' : 'other inboxes';