cmlenz@113: # -*- coding: utf-8 -*-
cmlenz@113: #
cmlenz@113: # Copyright (C) 2006 Edgewall Software
cmlenz@113: # All rights reserved.
cmlenz@113: #
cmlenz@113: # This software is licensed as described in the file COPYING, which
cmlenz@113: # you should have received as part of this distribution. The terms
cmlenz@113: # are also available at http://markup.edgewall.org/wiki/License.
cmlenz@113: #
cmlenz@113: # This software consists of voluntary contributions made by many
cmlenz@113: # individuals. For the exact contribution history, see the revision
cmlenz@113: # history and logs, available at http://markup.edgewall.org/log/.
cmlenz@113: 
cmlenz@113: import doctest
cmlenz@113: import unittest
cmlenz@113: 
cmlenz@113: from markup.core import Stream
cmlenz@113: from markup.input import HTML, ParseError
cmlenz@113: from markup.filters import HTMLSanitizer
cmlenz@113: 
cmlenz@113: 
cmlenz@113: class HTMLSanitizerTestCase(unittest.TestCase):
cmlenz@113: 
cmlenz@113:     def test_sanitize_unchanged(self):
cmlenz@113:         html = HTML('<a href="#">fo<br />o</a>')
cmlenz@113:         self.assertEquals('<a href="#">fo<br/>o</a>',
cmlenz@113:                           str(html.filter(HTMLSanitizer()))
cmlenz@113: 
cmlenz@113:     def test_sanitize_escape_text(self):
cmlenz@113:         html = HTML('<a href="#">fo&amp;</a>')
cmlenz@113:         self.assertEquals('<a href="#">fo&amp;</a>',
cmlenz@113:                           str(html.filter(HTMLSanitizer()))
cmlenz@113:         html = HTML('<a href="#">&lt;foo&gt;</a>')
cmlenz@113:         self.assertEquals('<a href="#">&lt;foo&gt;</a>',
cmlenz@113:                           str(html.filter(HTMLSanitizer()))
cmlenz@113: 
cmlenz@113:     def test_sanitize_entityref_text(self):
cmlenz@113:         html = HTML('<a href="#">fo&ouml;</a>')
cmlenz@113:         self.assertEquals(u'<a href="#">fo&ouml;</a>',
cmlenz@113:                           str(html.filter(HTMLSanitizer()))
cmlenz@113: 
cmlenz@113:     def test_sanitize_escape_attr(self):
cmlenz@113:         html = HTML('<div title="&lt;foo&gt;"></div>')
cmlenz@113:         self.assertEquals('<div title="&lt;foo&gt;"/>',
cmlenz@113:                           str(html.filter(HTMLSanitizer())))
cmlenz@113: 
cmlenz@113:     def test_sanitize_close_empty_tag(self):
cmlenz@113:         html = HTML('<a href="#">fo<br>o</a>')
cmlenz@113:         self.assertEquals('<a href="#">fo<br/>o</a>',
cmlenz@113:                           str(html.filter(HTMLSanitizer()))
cmlenz@113: 
cmlenz@113:     def test_sanitize_invalid_entity(self):
cmlenz@113:         html = HTML('&junk;')
cmlenz@113:         self.assertEquals('&junk;', str(html.filter(HTMLSanitizer()))
cmlenz@113: 
cmlenz@113:     def test_sanitize_remove_script_elem(self):
cmlenz@113:         html = HTML('<script>alert("Foo")</script>')
cmlenz@113:         self.assertEquals('', str(html.filter(HTMLSanitizer()))
cmlenz@113:         html = HTML('<SCRIPT SRC="http://example.com/"></SCRIPT>')
cmlenz@113:         self.assertEquals('', str(html.filter(HTMLSanitizer()))
cmlenz@113:         self.assertRaises(ParseError, HTML, '<SCR\0IPT>alert("foo")</SCR\0IPT>')
cmlenz@113:         self.assertRaises(ParseError, HTML,
cmlenz@113:                           '<SCRIPT&XYZ SRC="http://example.com/"></SCRIPT>')
cmlenz@113: 
cmlenz@113:     def test_sanitize_remove_onclick_attr(self):
cmlenz@113:         html = HTML('<div onclick=\'alert("foo")\' />')
cmlenz@113:         self.assertEquals('<div/>', str(html.filter(HTMLSanitizer()))
cmlenz@113: 
cmlenz@113:     def test_sanitize_remove_style_scripts(self):
cmlenz@113:         # Inline style with url() using javascript: scheme
cmlenz@113:         html = HTML('<DIV STYLE=\'background: url(javascript:alert("foo"))\'>')
cmlenz@113:         self.assertEquals('<div/>', str(html.filter(HTMLSanitizer()))
cmlenz@113:         # Inline style with url() using javascript: scheme, using control char
cmlenz@113:         html = HTML('<DIV STYLE=\'background: url(&#1;javascript:alert("foo"))\'>')
cmlenz@113:         self.assertEquals('<div/>', str(html.filter(HTMLSanitizer()))
cmlenz@113:         # Inline style with url() using javascript: scheme, in quotes
cmlenz@113:         html = HTML('<DIV STYLE=\'background: url("javascript:alert(foo)")\'>')
cmlenz@113:         self.assertEquals('<div/>', str(html.filter(HTMLSanitizer()))
cmlenz@113:         # IE expressions in CSS not allowed
cmlenz@113:         html = HTML('<DIV STYLE=\'width: expression(alert("foo"));\'>')
cmlenz@113:         self.assertEquals('<div/>', str(html.filter(HTMLSanitizer()))
cmlenz@113:         html = HTML('<DIV STYLE=\'background: url(javascript:alert("foo"));'
cmlenz@113:                                  'color: #fff\'>')
cmlenz@113:         self.assertEquals('<div style="color: #fff"/>',
cmlenz@113:                           str(html.filter(HTMLSanitizer()))
cmlenz@113: 
cmlenz@113:     def test_sanitize_remove_src_javascript(self):
cmlenz@113:         html = HTML('<img src=\'javascript:alert("foo")\'>')
cmlenz@113:         self.assertEquals('<img/>', str(html.filter(HTMLSanitizer()))
cmlenz@113:         # Case-insensitive protocol matching
cmlenz@113:         html = HTML('<IMG SRC=\'JaVaScRiPt:alert("foo")\'>')
cmlenz@113:         self.assertEquals('<img/>', str(html.filter(HTMLSanitizer()))
cmlenz@113:         # Grave accents (not parsed)
cmlenz@113:         self.assertRaises(ParseError, HTML,
cmlenz@113:                           '<IMG SRC=`javascript:alert("RSnake says, \'foo\'")`>')
cmlenz@113:         # Protocol encoded using UTF-8 numeric entities
cmlenz@113:         html = HTML('<IMG SRC=\'&#106;&#97;&#118;&#97;&#115;&#99;&#114;&#105;'
cmlenz@113:                     '&#112;&#116;&#58;alert("foo")\'>')
cmlenz@113:         self.assertEquals('<img/>', str(html.filter(HTMLSanitizer()))
cmlenz@113:         # Protocol encoded using UTF-8 numeric entities without a semicolon
cmlenz@113:         # (which is allowed because the max number of digits is used)
cmlenz@113:         html = HTML('<IMG SRC=\'&#0000106&#0000097&#0000118&#0000097'
cmlenz@113:                     '&#0000115&#0000099&#0000114&#0000105&#0000112&#0000116'
cmlenz@113:                     '&#0000058alert("foo")\'>')
cmlenz@113:         self.assertEquals('<img/>', str(html.filter(HTMLSanitizer()))
cmlenz@113:         # Protocol encoded using UTF-8 numeric hex entities without a semicolon
cmlenz@113:         # (which is allowed because the max number of digits is used)
cmlenz@113:         html = HTML('<IMG SRC=\'&#x6A&#x61&#x76&#x61&#x73&#x63&#x72&#x69'
cmlenz@113:                     '&#x70&#x74&#x3A;alert("foo")\'>')
cmlenz@113:         self.assertEquals('<img/>', str(html.filter(HTMLSanitizer()))
cmlenz@113:         # Embedded tab character in protocol
cmlenz@113:         html = HTML('<IMG SRC=\'jav\tascript:alert("foo");\'>')
cmlenz@113:         self.assertEquals('<img/>', str(html.filter(HTMLSanitizer()))
cmlenz@113:         # Embedded tab character in protocol, but encoded this time
cmlenz@113:         html = HTML('<IMG SRC=\'jav&#x09;ascript:alert("foo");\'>')
cmlenz@113:         self.assertEquals('<img/>', str(html.filter(HTMLSanitizer()))
cmlenz@113: 
cmlenz@113: 
cmlenz@113: def suite():
cmlenz@113:     suite = unittest.TestSuite()
cmlenz@113:     suite.addTest(unittest.makeSuite(HTMLSanitizerTestCase, 'test'))
cmlenz@113:     return suite
cmlenz@113: 
cmlenz@113: if __name__ == '__main__':
cmlenz@113:     unittest.main(defaultTest='suite')