X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FGitHTTPBackend.pm;h=974a1831340c3581a3fc55a5e0b335b7055c22f3;hb=9ea503ef65426070303fe1929f456b3591d74d93;hp=d1132fb7492c8b5039532fdc224ac9d348abcceb;hpb=02f9b34f398bef722159cd54a629441f861d37b7;p=public-inbox.git

diff --git a/lib/PublicInbox/GitHTTPBackend.pm b/lib/PublicInbox/GitHTTPBackend.pm
index d1132fb7..974a1831 100644
--- a/lib/PublicInbox/GitHTTPBackend.pm
+++ b/lib/PublicInbox/GitHTTPBackend.pm
@@ -1,17 +1,18 @@
-# Copyright (C) 2016-2019 all contributors <meta@public-inbox.org>
+# Copyright (C) all contributors <meta@public-inbox.org>
 # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
 
 # when no endpoints match, fallback to this and serve a static file
 # or smart HTTP.  This is our wrapper for git-http-backend(1)
 package PublicInbox::GitHTTPBackend;
 use strict;
-use warnings;
+use v5.10.1;
 use Fcntl qw(:seek);
-use IO::Handle;
+use IO::Handle; # ->flush
 use HTTP::Date qw(time2str);
 use PublicInbox::Qspawn;
 use PublicInbox::Tmpfile;
 use PublicInbox::WwwStatic qw(r @NO_CACHE);
+use Carp ();
 
 # 32 is same as the git-daemon connection limit
 my $default_limiter = PublicInbox::Qspawn::Limiter->new(32);
@@ -22,13 +23,10 @@ my @text = qw[HEAD info/refs info/attributes
 	objects/info/(?:http-alternates|alternates|packs)
 	cloneurl description];
 
-my @binary = qw!
-	objects/[a-f0-9]{2}/[a-f0-9]{38}
-	objects/pack/pack-[a-f0-9]{40}\.(?:pack|idx)
-	!;
+my @binary = ('objects/[a-f0-9]{2}/[a-f0-9]{38,62}',
+	'objects/pack/pack-[a-f0-9]{40,64}\.(?:pack|idx)');
 
 our $ANY = join('|', @binary, @text, 'git-upload-pack');
-my $BIN = join('|', @binary);
 my $TEXT = join('|', @text);
 
 sub serve {
@@ -45,10 +43,7 @@ sub serve {
 	serve_dumb($env, $git, $path);
 }
 
-sub err ($@) {
-	my ($env, @msg) = @_;
-	$env->{'psgi.errors'}->print(@msg, "\n");
-}
+sub ucarp { Carp::carp(@_); undef }
 
 my $prev = 0;
 my $exp;
@@ -64,13 +59,13 @@ sub serve_dumb {
 
 	my $h = [];
 	my $type;
-	if ($path =~ m!\Aobjects/[a-f0-9]{2}/[a-f0-9]{38}\z!) {
+	if ($path =~ m!\Aobjects/[a-f0-9]{2}/[a-f0-9]{38,62}\z!) {
 		$type = 'application/x-git-loose-object';
 		cache_one_year($h);
-	} elsif ($path =~ m!\Aobjects/pack/pack-[a-f0-9]{40}\.pack\z!) {
+	} elsif ($path =~ m!\Aobjects/pack/pack-[a-f0-9]{40,64}\.pack\z!) {
 		$type = 'application/x-git-packed-objects';
 		cache_one_year($h);
-	} elsif ($path =~ m!\Aobjects/pack/pack-[a-f0-9]{40}\.idx\z!) {
+	} elsif ($path =~ m!\Aobjects/pack/pack-[a-f0-9]{40,64}\.idx\z!) {
 		$type = 'application/x-git-packed-objects-toc';
 		cache_one_year($h);
 	} elsif ($path =~ /\A(?:$TEXT)\z/o) {
@@ -98,6 +93,7 @@ sub serve_smart {
 	foreach my $name (qw(QUERY_STRING
 				REMOTE_USER REMOTE_ADDR
 				HTTP_CONTENT_ENCODING
+				HTTP_GIT_PROTOCOL
 				CONTENT_TYPE
 				SERVER_PROTOCOL
 				REQUEST_METHOD)) {
@@ -117,42 +113,23 @@ sub input_prepare {
 
 	my $input = $env->{'psgi.input'};
 	my $fd = eval { fileno($input) };
-	if (defined $fd && $fd >= 0) {
-		return { 0 => $fd };
-	}
+	return { 0 => $fd } if (defined $fd && $fd >= 0);
 	my $id = "git-http.input.$env->{REMOTE_ADDR}:$env->{REMOTE_PORT}";
-	my $in = tmpfile($id);
-	unless (defined $in) {
-		err($env, "could not open temporary file: $!");
-		return;
-	}
+	my $in = tmpfile($id) // return ucarp("tmpfile: $!");
 	my $buf;
 	while (1) {
-		my $r = $input->read($buf, 8192);
-		unless (defined $r) {
-			err($env, "error reading input: $!");
-			return;
-		}
+		my $r = $input->read($buf, 8192) // return ucarp("read $!");
 		last if $r == 0;
-		unless (print $in $buf) {
-			err($env, "error writing temporary file: $!");
-			return;
-		}
+		print $in $buf // return ucarp("print: $!");
 	}
 	# ensure it's visible to git-http-backend(1):
-	unless ($in->flush) {
-		err($env, "error writing temporary file: $!");
-		return;
-	}
-	unless (defined(sysseek($in, 0, SEEK_SET))) {
-		err($env, "error seeking temporary file: $!");
-		return;
-	}
+	$in->flush // return ucarp("flush: $!");
+	sysseek($in, 0, SEEK_SET) // return ucarp($env, "seek: $!");
 	{ 0 => $in };
 }
 
-sub parse_cgi_headers {
-	my ($r, $bref) = @_;
+sub parse_cgi_headers { # {parse_hdr} for Qspawn
+	my ($r, $bref, $ctx) = @_;
 	return r(500) unless defined $r && $r >= 0;
 	$$bref =~ s/\A(.*?)\r?\n\r?\n//s or return $r == 0 ? r(500) : undef;
 	my $h = $1;
@@ -166,7 +143,26 @@ sub parse_cgi_headers {
 			push @h, $k, $v;
 		}
 	}
-	[ $code, \@h ]
+
+	# fallback to WwwCoderepo if cgit 404s.  Duplicating $ctx prevents
+	# ->finalize from the current Qspawn from using qspawn.wcb.
+	# This makes qspawn skip ->async_pass and causes
+	# PublicInbox::HTTPD::Async::event_step to close shortly after
+	if ($code == 404 && $ctx->{www} && !$ctx->{_coderepo_tried}++) {
+		my %ctx = %$ctx;
+		$ctx{env} = +{ %{$ctx->{env}} };
+		delete $ctx->{env}->{'qspawn.wcb'};
+		$ctx->{env}->{'plack.skip-deflater'} = 1; # prevent 2x gzip
+		my $res = $ctx->{www}->coderepo->srv(\%ctx);
+		if (ref($res) eq 'CODE') {
+			$res->(delete $ctx{env}->{'qspawn.wcb'});
+		} else { # ref($res) eq 'ARRAY'
+			$ctx->{env}->{'qspawn.wcb'} = $ctx{env}->{'qspawn.wcb'};
+		}
+		$res; # non ARRAY ref for ->psgi_return_init_cb
+	} else {
+		[ $code, \@h ]
+	}
 }
 
 1;