From 610e07776ea454ef5b86714f14718540ab0b4dae Mon Sep 17 00:00:00 2001
From: Eric Wong <e@80x24.org>
Date: Thu, 10 Apr 2014 19:26:40 +0000
Subject: cgi: be strict about UTF-8 encoding in HTML and XML

Hopefully this forces us to generate valid UTF-8 data.
---
 public-inbox-cgi | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)

(limited to 'public-inbox-cgi')

diff --git a/public-inbox-cgi b/public-inbox-cgi
index 91314f06..cd79a4a7 100755
--- a/public-inbox-cgi
+++ b/public-inbox-cgi
@@ -26,11 +26,10 @@ BEGIN {
 	}
 }
 
-binmode STDOUT, ':utf8';
-
 my $ret = main();
 
 my ($status, $headers, $body) = @$ret;
+set_binmode($headers);
 if (@ARGV && $ARGV[0] eq 'static') {
 	print $body;
 } else { # CGI
@@ -143,3 +142,13 @@ sub get_mid_txt {
 	my $x = mid2blob($ctx);
 	$x ? [ "200 OK", {'Content-Type' => 'text/plain'}, $$x ] : r404();
 }
+
+sub set_binmode {
+	my ($headers) = @_;
+	if ($headers->{'Content-Type'} eq 'text/plain') {
+		# no way to validate raw messages, mixed encoding is possible.
+		binmode STDOUT;
+	} else { # strict encoding for HTML and XML
+		binmode STDOUT, ':encoding(UTF-8)';
+	}
+}
-- 
cgit v1.2.3-24-ge0c7