[Erp5-report] r16821 - /erp5/trunk/products/ERP5Form/ScribusUtils.py

[nobody at svn.erp5.org]

Author: jerome
Date: Thu Oct  4 14:12:59 2007
New Revision: 16821

URL: http://svn.erp5.org?rev=16821&view=rev
Log:
doing str() on DOM parser results is dangerous, because it fails if the value contains non ascii characters.


Modified:
    erp5/trunk/products/ERP5Form/ScribusUtils.py

Modified: erp5/trunk/products/ERP5Form/ScribusUtils.py
URL: http://svn.erp5.org/erp5/trunk/products/ERP5Form/ScribusUtils.py?rev=16821&r1=16820&r2=16821&view=diff
==============================================================================
--- erp5/trunk/products/ERP5Form/ScribusUtils.py (original)
+++ erp5/trunk/products/ERP5Form/ScribusUtils.py Thu Oct  4 14:12:59 2007
@@ -1310,8 +1310,8 @@
             #iterating PAGEOBJECT attributes
             #old parsing method employed also here
             for node_id in page_object.attributes.keys():
-              node_name = str(node_id)
-              node_value = str(page_object.attributes[node_id].value)
+              node_name = node_id.encode('utf8')
+              node_value = page_object.attributes[node_id].value.encode('utf8')
 
               returned_page_object[node_name] = node_value
 
@@ -1367,8 +1367,8 @@
           field_OwnPage = ''
           # iterating field attributes
           for node_id in page_object.attributes.keys():
-            node_name = str(node_id)
-            node_value = str(page_object.attributes[node_id].value)
+            node_name = node_id.encode('utf8')
+            node_value = page_object.attributes[node_id].value.encode('utf8')
 
             if node_name == 'ANNAME':
               if node_value != '':