[Erp5-report] r41898 kazuhiko - /erp5/trunk/products/ERP5/Document/TextDocument.py

[nobody at svn.erp5.org]

Author: kazuhiko
Date: Fri Dec 31 00:33:26 2010
New Revision: 41898

URL: http://svn.erp5.org?rev=41898&view=rev
Log:
since XML may contain encoding, changing encoding only is not good. so we try to parse and reoutput utf-8 XML with a valid encoding information for XML.

Modified:
    erp5/trunk/products/ERP5/Document/TextDocument.py

Modified: erp5/trunk/products/ERP5/Document/TextDocument.py
URL: http://svn.erp5.org/erp5/trunk/products/ERP5/Document/TextDocument.py?rev=41898&r1=41897&r2=41898&view=diff
==============================================================================
--- erp5/trunk/products/ERP5/Document/TextDocument.py [utf8] (original)
+++ erp5/trunk/products/ERP5/Document/TextDocument.py [utf8] Fri Dec 31 00:33:26 2010
@@ -47,6 +47,7 @@ except ImportError:
 from Products.ERP5Type.Utils import guessEncodingFromText
 
 from lxml import html as etree_html
+from lxml import etree
 
 class TextDocument(CachedConvertableMixin, BaseConvertableFileMixin,
                                                             TextContent, File):
@@ -306,7 +307,15 @@ class TextDocument(CachedConvertableMixi
 
       content_type = self.getContentType() or DEFAULT_CONTENT_TYPE
       text_content = self.getData()
-      if content_type == 'text/html':
+      if content_type.endswith('xml'):
+        try:
+          tree = etree.fromstring(text_content)
+          text_content = etree.tostring(tree, encoding='utf-8', xml_declaration=True)
+          content_type = 'application/xml'
+          message = 'Conversion to base format succeeds'
+        except etree.XMLSyntaxError:
+          message = 'Conversion to base format without codec fails'
+      elif content_type == 'text/html':
         re_match = self.charset_parser.search(text_content)
         message = 'Conversion to base format succeeds'
         if re_match is not None: