diff --git a/product/ERP5Form/ScribusUtils.py b/product/ERP5Form/ScribusUtils.py
index b6aa11769f862819081984284a21cdaad285f7f9..dfc3a2636ff6bd3411ddad15677d7f11fd3c3e72 100644
--- a/product/ERP5Form/ScribusUtils.py
+++ b/product/ERP5Form/ScribusUtils.py
@@ -1235,10 +1235,9 @@ class ScribusParser:
     This function is used to get attributes'values in an object_dict and 
     to be sure a compatible value is returned (for that use default value)
     """
-    #return object_dict.get(check_key, None) or default_value
     if object_dict.has_key(check_key):
       # 'check_key' exists
-      if len(object_dict[check_key]) != 0:
+      if len(object_dict[check_key]):
         # check_key corresponding value is not null
         # returning this value
         return object_dict[check_key]
@@ -1261,6 +1260,15 @@ class ScribusParser:
     containing a dict of all the relative attributes
     """
 
+    # XXX this is a hack to correct a scribus problem
+    # actualy scribus (version 1.3.3.12svn) use some invali xml caracters
+    # like the '' caracter wich is a carriage return caratere, used for
+    # example in a text frame with many lines
+    # this problem is well knowed by scribus community and perhaps will be
+    # correct in futur scribus version
+    xml_string = xml_string.replace('', '\n')
+    xml_string = xml_string.replace('', '\t')
+
     # Create DOM tree from the xml string
     print " > create DOM tree"
     dom_tree = minidom.parseString(xml_string)
@@ -1351,7 +1359,7 @@ class ScribusParser:
         # end parsing document version 1.2.*
 
       else:
-        print " found Scribus Doucment format 1.3"
+        print " found Scribus Doucment format 1.3 or higher"
         # assuming version is compliant with 1.3.* specifications
 
         keep_page = 1