From cb73535c8eae02092df984bafbecabbce8049cd0 Mon Sep 17 00:00:00 2001
From: Andrew Dolgov <fox@madoka.volgo-balt.ru>
Date: Tue, 5 Jun 2012 21:52:37 +0400
Subject: Revert "Update HTML Purifier to version 4.4.0."

This reverts commit dd205fbad642ace6d0e33c8553f7d73404f140b4.
---
 .../library/HTMLPurifier/AttrDef/URI/Host.php      | 45 +++-------------------
 1 file changed, 6 insertions(+), 39 deletions(-)

(limited to 'lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI')

diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/Host.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/Host.php
index 125decb2d..feca469d7 100644
--- a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/Host.php
+++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/Host.php
@@ -44,8 +44,9 @@ class HTMLPurifier_AttrDef_URI_Host extends HTMLPurifier_AttrDef
 
         // A regular domain name.
 
-        // This doesn't match I18N domain names, but we don't have proper IRI support,
-        // so force users to insert Punycode.
+        // This breaks I18N domain names, but we don't have proper IRI support,
+        // so force users to insert Punycode. If there's complaining we'll
+        // try to fix things into an international friendly form.
 
         // The productions describing this are:
         $a   = '[a-z]';     // alpha
@@ -56,44 +57,10 @@ class HTMLPurifier_AttrDef_URI_Host extends HTMLPurifier_AttrDef
         // toplabel    = alpha | alpha *( alphanum | "-" ) alphanum
         $toplabel      = "$a($and*$an)?";
         // hostname    = *( domainlabel "." ) toplabel [ "." ]
-        if (preg_match("/^($domainlabel\.)*$toplabel\.?$/i", $string)) {
-            return $string;
-        }
-
-        // If we have Net_IDNA2 support, we can support IRIs by
-        // punycoding them. (This is the most portable thing to do,
-        // since otherwise we have to assume browsers support
-
-        if ($config->get('Core.EnableIDNA')) {
-            $idna = new Net_IDNA2(array('encoding' => 'utf8', 'overlong' => false, 'strict' => true));
-            // we need to encode each period separately
-            $parts = explode('.', $string);
-            try {
-                $new_parts = array();
-                foreach ($parts as $part) {
-                    $encodable = false;
-                    for ($i = 0, $c = strlen($part); $i < $c; $i++) {
-                        if (ord($part[$i]) > 0x7a) {
-                            $encodable = true;
-                            break;
-                        }
-                    }
-                    if (!$encodable) {
-                        $new_parts[] = $part;
-                    } else {
-                        $new_parts[] = $idna->encode($part);
-                    }
-                }
-                $string = implode('.', $new_parts);
-                if (preg_match("/^($domainlabel\.)*$toplabel\.?$/i", $string)) {
-                    return $string;
-                }
-            } catch (Exception $e) {
-                // XXX error reporting
-            }
-        }
+        $match = preg_match("/^($domainlabel\.)*$toplabel\.?$/i", $string);
+        if (!$match) return false;
 
-        return false;
+        return $string;
     }
 
 }
-- 
cgit v1.2.3