From 218eb67167c8d9c1c7ea8d6d9a8764a5cd4cd140 Mon Sep 17 00:00:00 2001
From: "Edward Z. Yang" <edwardzyang@thewritingpot.com>
Date: Mon, 14 Aug 2006 21:06:57 +0000
Subject: [PATCH] Remove legacy required code from AttrDef_URI, also explicitly
 disallow < and > in URIs.

git-svn-id: http://htmlpurifier.org/svnroot/htmlpurifier/trunk@253 48356398-32a2-884e-a903-53898d9a118a
---
 docs/config.txt                        |  8 ++++++--
 library/HTMLPurifier/AttrDef/URI.php   | 21 ++++++++++-----------
 library/HTMLPurifier/Definition.php    |  3 +--
 tests/HTMLPurifier/AttrDef/URITest.php |  4 ++++
 4 files changed, 21 insertions(+), 15 deletions(-)

diff --git a/docs/config.txt b/docs/config.txt
index a4cdd965..0562c0af 100644
--- a/docs/config.txt
+++ b/docs/config.txt
@@ -3,5 +3,9 @@ Configuration
 
 Configuration is documented on a per-use case: if a class uses a certain
 value from the configuration object, it has to define its name and what the
-value is used for.  This means decentralized configuration declaration that
-is nevertheless error checking.
+value is used for.  This means decentralized configuration declarations that
+are nevertheless error checking and a centralized configuration object.
+
+Directives are divided into namespaces, indicating the major portion of
+functionality they cover (although there may be overlaps.  Please consult
+the documentation in ConfigDef for more information on these namespaces.
diff --git a/library/HTMLPurifier/AttrDef/URI.php b/library/HTMLPurifier/AttrDef/URI.php
index 6e2c7449..b457efdc 100644
--- a/library/HTMLPurifier/AttrDef/URI.php
+++ b/library/HTMLPurifier/AttrDef/URI.php
@@ -12,12 +12,6 @@ HTMLPurifier_ConfigDef::define(
 class HTMLPurifier_AttrDef_URI extends HTMLPurifier_AttrDef
 {
     
-    var $required = false;
-    
-    function HTMLPurifier_AttrDef_URI($required = false) {
-        $this->required = $required;
-    }
-    
     function validate($uri, $config, &$context) {
         
         // We'll write stack-based parsers later, for now, use regexps to
@@ -30,18 +24,23 @@ class HTMLPurifier_AttrDef_URI extends HTMLPurifier_AttrDef
         // for HTTP and thus won't work for our generic URI parsing
         
         // according to the RFC... (but this cuts corners, i.e. non-validating)
-        $r_URI = '!^(([^:/?#]+):)?(//([^/?#]*))?([^?#]*)(\?([^#]*))?(#(.*))?!';
-        //           12            3  4          5       6  7        8 9
+        $r_URI = '!^'.
+            '(([^:/?#<>]+):)?'. // 2. Scheme
+            '(//([^/?#<>]*))?'. // 4. Authority
+            '([^?#<>]*)'.       // 5. Path
+            '(\?([^#<>]*))?'.   // 7. Query
+            '(#([^<>]*))?'.     // 8. Fragment
+            '$!';
         
         $matches = array();
         $result = preg_match($r_URI, $uri, $matches);
         
-        if (!$result)  return '';
+        if (!$result) return false; // invalid URI
         
         // seperate out parts
         $scheme     = !empty($matches[1]) ? $matches[2] : null;
         $authority  = !empty($matches[3]) ? $matches[4] : null;
-        $path       = $matches[5]; // always present
+        $path       = $matches[5]; // always present, can be empty
         $query      = !empty($matches[6]) ? $matches[7] : null;
         $fragment   = !empty($matches[8]) ? $matches[9] : null;
         
@@ -53,7 +52,7 @@ class HTMLPurifier_AttrDef_URI extends HTMLPurifier_AttrDef
             // retrieve the specific scheme object from the registry
             $scheme = ctype_lower($scheme) ? $scheme : strtolower($scheme);
             $scheme_obj =& $registry->getScheme($scheme, $config);
-            if (!$scheme_obj) return $this->required ? '' : false; // invalid scheme, clean it out
+            if (!$scheme_obj) return false; // invalid scheme, clean it out
         } else {
             $scheme_obj =& $registry->getScheme(
                 $config->get('URI', 'DefaultScheme'), $config
diff --git a/library/HTMLPurifier/Definition.php b/library/HTMLPurifier/Definition.php
index 2670401e..e2e852e4 100644
--- a/library/HTMLPurifier/Definition.php
+++ b/library/HTMLPurifier/Definition.php
@@ -317,13 +317,12 @@ class HTMLPurifier_Definition
         $e_URI = new HTMLPurifier_AttrDef_URI();
         $this->info['a']->attr['href'] =
         $this->info['img']->attr['longdesc'] =
+        $this->info['img']->attr['src'] =
         $this->info['del']->attr['cite'] =
         $this->info['ins']->attr['cite'] =
         $this->info['blockquote']->attr['cite'] =
         $this->info['q']->attr['cite'] = $e_URI;
         
-        $this->info['img']->attr['src'] = new HTMLPurifier_AttrDef_URI(true);
-        
         //////////////////////////////////////////////////////////////////////
         // UNIMP : info_tag_transform : transformations of tags
         
diff --git a/tests/HTMLPurifier/AttrDef/URITest.php b/tests/HTMLPurifier/AttrDef/URITest.php
index d59b8e35..921aa910 100644
--- a/tests/HTMLPurifier/AttrDef/URITest.php
+++ b/tests/HTMLPurifier/AttrDef/URITest.php
@@ -153,6 +153,10 @@ class HTMLPurifier_AttrDef_URITest extends HTMLPurifier_AttrDefHarness
         $uri[18] = '/a/b';
         $components[18] = array(null, null, null, '/a/b', null);
         
+        // it's not allowed, so generic URI should get it
+        $uri[19] = '<';
+        $expect_uri[19] = false;
+        
         foreach ($uri as $i => $value) {
             
             // setUpAssertDef