diff --git a/product/ERP5OOo/Document/DMSFile.py b/product/ERP5OOo/Document/DMSFile.py
index 405d805b79e0ef1360d68fb8918449063350bb16..8441196ee29e65349a07e1fc204f2628c702feca 100644
--- a/product/ERP5OOo/Document/DMSFile.py
+++ b/product/ERP5OOo/Document/DMSFile.py
@@ -41,8 +41,8 @@ mimetypes.init()
 
 
 rs=[]
-rs.append(re.compile('<!.*>'))
 rs.append(re.compile('<HEAD>.*</HEAD>',re.DOTALL|re.MULTILINE|re.IGNORECASE))
+rs.append(re.compile('<!DOCTYPE[^>]*>'))
 rs.append(re.compile('<.?(HTML|BODY)[^>]*>',re.DOTALL|re.MULTILINE|re.IGNORECASE))
 
 def stripHtml(txt):