From 582ffc414320fd2cd895791d99b421d50236fd61 Mon Sep 17 00:00:00 2001
From: "Edward Z. Yang" <edwardzyang@thewritingpot.com>
Date: Sun, 14 Jan 2007 15:54:05 +0000
Subject: [PATCH] [1.4.0] Implemented AttrDef_CSSURI for url(http://google.com)
 style declarations - 1.3.3 release downgraded to "not likely"

git-svn-id: http://htmlpurifier.org/svnroot/htmlpurifier/trunk@639 48356398-32a2-884e-a903-53898d9a118a
---
 NEWS                                      |  4 +-
 library/HTMLPurifier/AttrDef/CSS.php      |  3 ++
 library/HTMLPurifier/AttrDef/CSSURI.php   | 58 +++++++++++++++++++++++
 tests/HTMLPurifier/AttrDef/CSSURITest.php | 37 +++++++++++++++
 tests/index.php                           |  1 +
 5 files changed, 101 insertions(+), 2 deletions(-)
 create mode 100644 library/HTMLPurifier/AttrDef/CSSURI.php
 create mode 100644 tests/HTMLPurifier/AttrDef/CSSURITest.php

diff --git a/NEWS b/NEWS
index 9ef2fe26..3365bc70 100644
--- a/NEWS
+++ b/NEWS
@@ -10,9 +10,9 @@ NEWS ( CHANGELOG and HISTORY )                                     HTMLPurifier
 ==========================
 
 1.4.0, unknown release date
-(major feature release)
+. Implemented AttrDef_CSSURI for url(http://google.com) style declarations
 
-1.3.3, unknown release date, may be dropped
+1.3.3, unknown release date, likely to be dropped
 ! Moved SLOW to docs/enduser-slow.html and added code examples
 
 1.3.2, released 2006-12-25
diff --git a/library/HTMLPurifier/AttrDef/CSS.php b/library/HTMLPurifier/AttrDef/CSS.php
index 404c7000..ca1b0224 100644
--- a/library/HTMLPurifier/AttrDef/CSS.php
+++ b/library/HTMLPurifier/AttrDef/CSS.php
@@ -20,6 +20,9 @@ class HTMLPurifier_AttrDef_CSS extends HTMLPurifier_AttrDef
         
         // we're going to break the spec and explode by semicolons.
         // This is because semicolon rarely appears in escaped form
+        // Doing this is generally flaky but fast
+        // IT MIGHT APPEAR IN URIs, see HTMLPurifier_AttrDef_CSSURI
+        // for details
         
         $declarations = explode(';', $css);
         $propvalues = array();
diff --git a/library/HTMLPurifier/AttrDef/CSSURI.php b/library/HTMLPurifier/AttrDef/CSSURI.php
new file mode 100644
index 00000000..a92b6263
--- /dev/null
+++ b/library/HTMLPurifier/AttrDef/CSSURI.php
@@ -0,0 +1,58 @@
+<?php
+
+require_once 'HTMLPurifier/AttrDef/URI.php';
+
+/**
+ * Validates a URI in CSS syntax, which uses url('http://example.com')
+ * @note While theoretically speaking we a URI in a CSS document could
+ *       be non-embedded, as of CSS2 there is no such usage so we're
+ *       generalizing it. This may need to be changed in the future.
+ * @warning Since HTMLPurifier_AttrDef_CSS blindly uses semicolons as
+ *          the separator, you cannot put a literal semicolon in
+ *          in the URI. Try percent encoding it, in that case.
+ */
+class HTMLPurifier_AttrDef_CSSURI extends HTMLPurifier_AttrDef_URI
+{
+    
+    function HTMLPurifier_AttrDef_CSSURI() {
+        $this->HTMLPurifier_AttrDef_URI(true); // always embedded
+    }
+    
+    function validate($uri_string, $config, &$context) {
+        // parse the URI out of the string and then pass it onto
+        // the parent object
+        
+        $uri_string = $this->parseCDATA($uri_string);
+        if (strpos($uri_string, 'url(') !== 0) return false;
+        $uri_string = substr($uri_string, 4);
+        $new_length = strlen($uri_string) - 1;
+        if ($uri_string[$new_length] != ')') return false;
+        $uri = trim(substr($uri_string, 0, $new_length));
+        
+        if (isset($uri[0]) && ($uri[0] == "'" || $uri[0] == '"')) {
+            $quote = $uri[0];
+            $new_length = strlen($uri) - 1;
+            if ($uri[$new_length] !== $quote) return false;
+            $uri = substr($uri, 1, $new_length - 1);
+        }
+        
+        $keys   = array(  '(',   ')',   ',',   ' ',   '"',   "'");
+        $values = array('\\(', '\\)', '\\,', '\\ ', '\\"', "\\'");
+        $uri = str_replace($values, $keys, $uri);
+        
+        $result = parent::validate($uri, $config, $context);
+        
+        if ($result === false) return false;
+        
+        // escape necessary characters according to CSS spec
+        // except for the comma, none of these should appear in the
+        // URI at all
+        $result = str_replace($keys, $values, $result);
+        
+        return "url($result)";
+        
+    }
+    
+}
+
+?>
\ No newline at end of file
diff --git a/tests/HTMLPurifier/AttrDef/CSSURITest.php b/tests/HTMLPurifier/AttrDef/CSSURITest.php
new file mode 100644
index 00000000..1fe1a3dc
--- /dev/null
+++ b/tests/HTMLPurifier/AttrDef/CSSURITest.php
@@ -0,0 +1,37 @@
+<?php
+
+require_once 'HTMLPurifier/AttrDef/CSSURI.php';
+require_once 'HTMLPurifier/AttrDefHarness.php';
+
+class HTMLPurifier_AttrDef_CSSURITest extends HTMLPurifier_AttrDefHarness
+{
+    
+    function test() {
+        
+        $this->def = new HTMLPurifier_AttrDef_CSSURI();
+        
+        $this->assertDef('', false);
+        
+        // we could be nice but we won't be
+        $this->assertDef('http://www.example.com/', false);
+        
+        // no quotes are used, since that's the most widely supported
+        // syntax
+        $this->assertDef('url(', false);
+        $this->assertDef('url()', true);
+        $result = "url(http://www.example.com/)";
+        $this->assertDef('url(http://www.example.com/)', $result);
+        $this->assertDef('url("http://www.example.com/")', $result);
+        $this->assertDef("url('http://www.example.com/')", $result);
+        $this->assertDef(
+            '  url(  "http://www.example.com/" )   ', $result);
+        
+        // escaping
+        $this->assertDef("url(http://www.example.com/foo,bar\))", 
+            "url(http://www.example.com/foo\,bar\))");
+        
+    }
+    
+}
+
+?>
\ No newline at end of file
diff --git a/tests/index.php b/tests/index.php
index 92c845fe..3b8b87b1 100644
--- a/tests/index.php
+++ b/tests/index.php
@@ -84,6 +84,7 @@ $test_files[] = 'AttrDef/FontTest.php';
 $test_files[] = 'AttrDef/BorderTest.php';
 $test_files[] = 'AttrDef/ListStyleTest.php';
 $test_files[] = 'AttrDef/Email/SimpleCheckTest.php';
+$test_files[] = 'AttrDef/CSSURITest.php';
 $test_files[] = 'IDAccumulatorTest.php';
 $test_files[] = 'TagTransformTest.php';
 $test_files[] = 'AttrTransform/LangTest.php';