Fix TextDocument class so it can work consistently with abstract Document class

git-svn-id: https://svn.erp5.org/repos/public/erp5/trunk@14907 20353a03-c40f-0410-a6d1-a30d3c3de9de

product/ERP5/Document/Document.py

@@ -1090,7 +1090,7 @@ class Document(XMLObject, UrlMixIn, ConversionCacheMixin, SnapshotMixin):
     else:
       stripped_html = html
     # find charset and convert to utf-8
-    charset_list = self.charset_parser.findall(html)
+    charset_list = self.charset_parser.findall(str(html))
     if charset_list:
       stripped_html = unicode(stripped_html, charset_list[0]).encode('utf-8')
     return stripped_html

product/ERP5/Document/TextDocument.py

@@ -164,4 +164,14 @@ class TextDocument(Document, TextContent):
       base_list = re.findall(self.base_parser, str(html))
       if base_list:
         return base_list[0]
-      return Document.getContentBaseURL(self)
\ No newline at end of file
+      return Document.getContentBaseURL(self)
+      
+    def hasBaseData(self):
+      """ 
+        This method is an override of dynamically generated method for Document class.
+        We need to manually override it because for some backwards compatibility 
+        instances of TextDocument as 'Web Page' doesn't use 'base_data' to store raw 
+        data information. Instead they use 'text-content'
+        This makes results and login of abstract Document class inconsistent.
+      """
+      return self.hasTextContent()