From 8bff2606d8c5fc375353399c8297a204a31fe623 Mon Sep 17 00:00:00 2001
From: "Edward Z. Yang" <edwardzyang@thewritingpot.com>
Date: Thu, 17 Jul 2008 00:06:24 -0400
Subject: [PATCH] Factor out parse() from process(); remove unnecessary xmlns.

process() is a super-method that does many things; parse() now performs
the parsing of text into DOM form, stopping prior to the DOM filters. This
is useful for filters that need to compose multiple XHTML documents into one.

The post-processing of the HTML string now removes unnecessary xmlns from
non root elements in order to ease validation.

Signed-off-by: Edward Z. Yang <edwardzyang@thewritingpot.com>
---
 XHTMLCompiler/FilterManager.php | 29 +++++++++++++++++++++++++----
 1 file changed, 25 insertions(+), 4 deletions(-)

diff --git a/XHTMLCompiler/FilterManager.php b/XHTMLCompiler/FilterManager.php
index d61f21f..ea3f514 100644
--- a/XHTMLCompiler/FilterManager.php
+++ b/XHTMLCompiler/FilterManager.php
@@ -100,12 +100,13 @@ class XHTMLCompiler_FilterManager
     }
     
     /**
-     * Accepts a page's text (usually XHTML) and processes it.
+     * Accepts a page's text and turns it into its DOM representation. 
+     * Text, initial validation and XIncludes will be processed before
+     * returning. DOM filters will *not* be processed.
      * @param $text String text to be processed
-     * @param $page XHTMLCompiler_Page representing currently processed page
+     * @param 
      */
-    public function process($text, $page) {
-        
+    public function parse($text, $page) {
         // do pre-text processing
         foreach ($this->preTextFilters as $filter) {
             $text = $filter->process($text, $page, $this);
@@ -126,6 +127,18 @@ class XHTMLCompiler_FilterManager
         $this->analyzeXIncludes($dom);
         $dom->xinclude();
         
+        return $dom;
+    }
+    
+    /**
+     * Accepts a page's text (usually XHTML) and processes it.
+     * @param $text String text to be processed
+     * @param $page XHTMLCompiler_Page representing currently processed page
+     */
+    public function process($text, $page) {
+        
+        $dom = $this->parse($text, $page);
+        
         // run DOM filters
         foreach ($this->DOMFilters as $filter) {
             $filter->setup($dom);
@@ -150,6 +163,14 @@ class XHTMLCompiler_FilterManager
             $text
         );
         
+        // replace any redundant xmlns sections, although they are
+        // valid they interfere with DTD validation
+        $text = preg_replace(
+            '#(<(?!html)[^>]+) xmlns="http://www.w3.org/1999/xhtml"#',
+            '$1',
+            $text
+        );
+        
         // okay, now finally do validation, and let the errors get
         // spit out if there are some collect parse errors
         set_error_handler(array($this, 'validationErrorHandler'));
-- 
2.11.4.GIT