[IMP] [READD] [IMP] tools: mail: have a strict mode for html_sanitizer (not default...

options before website: without frames, with safe attributes only.
We will have to find a way to use the strict mode when parsing incoming
emails, but not when sending or storing openerp content.

Currently the not strict mode is the default one, to avoid side-effects with
the website.

bzr revid: tde@openerp.com-20140115141319-g15zl1kqrp8sgoa3

diff --git a/openerp/tools/mail.py b/openerp/tools/mail.py
index 55813bc..3148083 100644 (file)
--- a/openerp/tools/mail.py
+++ b/openerp/tools/mail.py
@@ -2,7 +2,7 @@
 ##############################################################################
 #
 #    OpenERP, Open Source Business Applications
-#    Copyright (C) 2012-2013 OpenERP S.A. (<http://openerp.com>).
+#    Copyright (C) 2012-TODAY OpenERP S.A. (<http://openerp.com>).
 #
 #    This program is free software: you can redistribute it and/or modify
 #    it under the terms of the GNU Affero General Public License as
@@ -45,10 +45,14 @@ tags_to_remove = ['html', 'body', 'font']
 
 # allow new semantic HTML5 tags
 allowed_tags = clean.defs.tags | frozenset('article section header footer hgroup nav aside figure main'.split())
-safe_attrs = clean.defs.safe_attrs | frozenset(['style'])
+safe_attrs = clean.defs.safe_attrs | frozenset(
+    ['style',
+     'data-oe-model', 'data-oe-id', 'data-oe-field', 'data-oe-type', 'data-oe-expression', 'data-oe-translate', 'data-oe-nodeid',
+     'data-snippet-id', 'data-publish', 'data-id', 'data-res_id', 'data-member_id', 'data-view-id'
+     ])
 
 
-def html_sanitize(src, silent=True):
+def html_sanitize(src, silent=True, strict=False):
     if not src:
         return src
     src = ustr(src, errors='replace')
@@ -62,7 +66,6 @@ def html_sanitize(src, silent=True):
     kwargs = {
         'page_structure': True,
         'style': False,             # do not remove style attributes
-        'frames': False,            # de not remove frames (embbed video in CMS blogs)
         'forms': True,              # remove form tags
         'remove_unknown_tags': False,
         'allow_tags': allowed_tags,
@@ -76,7 +79,16 @@ def html_sanitize(src, silent=True):
     else:
         kwargs['remove_tags'] = tags_to_kill + tags_to_remove
 
-    kwargs['safe_attrs_only'] = False
+    if strict:
+        if etree.LXML_VERSION >= (3, 1, 0):
+            # lxml < 3.1.0 does not allow to specify safe_attrs. We keep all attributes in order to keep "style"
+            kwargs.update({
+                'safe_attrs_only': True,
+                'safe_attrs': safe_attrs,
+            })
+    else:
+        kwargs['safe_attrs_only'] = False    # keep oe-data attributes + style
+        kwargs['frames'] = False,            # do not remove frames (embbed video in CMS blogs)
 
     try:
         # some corner cases make the parser crash (such as <SCRIPT/XSS SRC=\"http://ha.ckers.org/xss.js\"></SCRIPT> in test_mail)