1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
|
import base64
import gzip
import io
import unittest
from lxml.tests.common_imports import make_doctest
import lxml.html
from lxml.html.clean import Cleaner, clean_html
class CleanerTest(unittest.TestCase):
def test_allow_tags(self):
html = """
<html>
<head>
</head>
<body>
<p>some text</p>
<table>
<tr>
<td>hello</td><td>world</td>
</tr>
<tr>
<td>hello</td><td>world</td>
</tr>
</table>
<img>
</body>
</html>
"""
html_root = lxml.html.document_fromstring(html)
cleaner = Cleaner(
remove_unknown_tags = False,
allow_tags = ['table', 'tr', 'td'])
result = cleaner.clean_html(html_root)
self.assertEqual(12-5+1, len(list(result.iter())))
def test_allow_and_remove(self):
with self.assertRaises(ValueError):
Cleaner(allow_tags=['a'], remove_unknown_tags=True)
def test_remove_unknown_tags(self):
html = """<div><bun>lettuce, tomato, veggie patty</bun></div>"""
clean_html = """<div>lettuce, tomato, veggie patty</div>"""
cleaner = Cleaner(remove_unknown_tags=True)
result = cleaner.clean_html(html)
self.assertEqual(
result,
clean_html,
msg="Unknown tags not removed. Got: %s" % result,
)
def test_safe_attrs_included(self):
html = """<p><span style="color: #00ffff;">Cyan</span></p>"""
safe_attrs=set(lxml.html.defs.safe_attrs)
safe_attrs.add('style')
cleaner = Cleaner(
safe_attrs_only=True,
safe_attrs=safe_attrs)
result = cleaner.clean_html(html)
self.assertEqual(html, result)
def test_safe_attrs_excluded(self):
html = """<p><span style="color: #00ffff;">Cyan</span></p>"""
expected = """<p><span>Cyan</span></p>"""
safe_attrs=set()
cleaner = Cleaner(
safe_attrs_only=True,
safe_attrs=safe_attrs)
result = cleaner.clean_html(html)
self.assertEqual(expected, result)
def test_clean_invalid_root_tag(self):
# only testing that cleaning with invalid root tags works at all
s = lxml.html.fromstring('parent <invalid tag>child</another>')
self.assertEqual('parent child', clean_html(s).text_content())
s = lxml.html.fromstring('<invalid tag>child</another>')
self.assertEqual('child', clean_html(s).text_content())
def test_clean_with_comments(self):
html = """<p><span style="color: #00ffff;">Cy<!-- xx -->an</span><!-- XXX --></p>"""
s = lxml.html.fragment_fromstring(html)
self.assertEqual(
b'<p><span>Cyan</span></p>',
lxml.html.tostring(clean_html(s)))
self.assertEqual(
'<p><span>Cyan</span></p>',
clean_html(html))
cleaner = Cleaner(comments=False)
result = cleaner.clean_html(s)
self.assertEqual(
b'<p><span>Cy<!-- xx -->an</span><!-- XXX --></p>',
lxml.html.tostring(result))
self.assertEqual(
'<p><span>Cy<!-- xx -->an</span><!-- XXX --></p>',
cleaner.clean_html(html))
def test_sneaky_noscript_in_style(self):
# This gets parsed as <noscript> -> <style>"...</noscript>..."</style>
# thus passing the </noscript> through into the output.
html = '<noscript><style><a title="</noscript><img src=x onerror=alert(1)>">'
s = lxml.html.fragment_fromstring(html)
self.assertEqual(
b'<noscript><style>/* deleted */</style></noscript>',
lxml.html.tostring(clean_html(s)))
def test_sneaky_js_in_math_style(self):
# This gets parsed as <math> -> <style>"..."</style>
# thus passing any tag/script/whatever content through into the output.
html = '<math><style><img src=x onerror=alert(1)></style></math>'
s = lxml.html.fragment_fromstring(html)
self.assertEqual(
b'<math><style>/* deleted */</style></math>',
lxml.html.tostring(clean_html(s)))
def test_sneaky_import_in_style(self):
# Prevent "@@importimport" -> "@import" replacement etc.
style_codes = [
"@@importimport(extstyle.css)",
"@ @ import import(extstyle.css)",
"@ @ importimport(extstyle.css)",
"@@ import import(extstyle.css)",
"@ @import import(extstyle.css)",
"@@importimport()",
"@@importimport() ()",
"@/* ... */import()",
"@im/* ... */port()",
"@ @import/* ... */import()",
"@ /* ... */ import()",
]
for style_code in style_codes:
html = '<style>%s</style>' % style_code
s = lxml.html.fragment_fromstring(html)
cleaned = lxml.html.tostring(clean_html(s))
self.assertEqual(
b'<style>/* deleted */</style>',
cleaned,
"%s -> %s" % (style_code, cleaned))
def test_sneaky_schemes_in_style(self):
style_codes = [
"javasjavascript:cript:",
"javascriptjavascript::",
"javascriptjavascript:: :",
"vbjavascript:cript:",
]
for style_code in style_codes:
html = '<style>%s</style>' % style_code
s = lxml.html.fragment_fromstring(html)
cleaned = lxml.html.tostring(clean_html(s))
self.assertEqual(
b'<style>/* deleted */</style>',
cleaned,
"%s -> %s" % (style_code, cleaned))
def test_sneaky_urls_in_style(self):
style_codes = [
"url(data:image/svg+xml;base64,...)",
"url(javasjavascript:cript:)",
"url(javasjavascript:cript: ::)",
"url(vbjavascript:cript:)",
"url(vbjavascript:cript: :)",
]
for style_code in style_codes:
html = '<style>%s</style>' % style_code
s = lxml.html.fragment_fromstring(html)
cleaned = lxml.html.tostring(clean_html(s))
self.assertEqual(
b'<style>url()</style>',
cleaned,
"%s -> %s" % (style_code, cleaned))
def test_svg_data_links(self):
# Remove SVG images with potentially insecure content.
svg = b'<svg onload="alert(123)" />'
gzout = io.BytesIO()
f = gzip.GzipFile(fileobj=gzout, mode='wb')
f.write(svg)
f.close()
svgz = gzout.getvalue()
svg_b64 = base64.b64encode(svg).decode('ASCII')
svgz_b64 = base64.b64encode(svgz).decode('ASCII')
urls = [
"data:image/svg+xml;base64," + svg_b64,
"data:image/svg+xml-compressed;base64," + svgz_b64,
]
for url in urls:
html = '<img src="%s">' % url
s = lxml.html.fragment_fromstring(html)
cleaned = lxml.html.tostring(clean_html(s))
self.assertEqual(
b'<img src="">',
cleaned,
"%s -> %s" % (url, cleaned))
def test_image_data_links(self):
data = b'123'
data_b64 = base64.b64encode(data).decode('ASCII')
urls = [
"data:image/jpeg;base64," + data_b64,
"data:image/apng;base64," + data_b64,
"data:image/png;base64," + data_b64,
"data:image/gif;base64," + data_b64,
"data:image/webp;base64," + data_b64,
"data:image/bmp;base64," + data_b64,
"data:image/tiff;base64," + data_b64,
"data:image/x-icon;base64," + data_b64,
]
for url in urls:
html = '<img src="%s">' % url
s = lxml.html.fragment_fromstring(html)
cleaned = lxml.html.tostring(clean_html(s))
self.assertEqual(
html.encode("UTF-8"),
cleaned,
"%s -> %s" % (url, cleaned))
def test_image_data_links_in_style(self):
data = b'123'
data_b64 = base64.b64encode(data).decode('ASCII')
urls = [
"data:image/jpeg;base64," + data_b64,
"data:image/apng;base64," + data_b64,
"data:image/png;base64," + data_b64,
"data:image/gif;base64," + data_b64,
"data:image/webp;base64," + data_b64,
"data:image/bmp;base64," + data_b64,
"data:image/tiff;base64," + data_b64,
"data:image/x-icon;base64," + data_b64,
]
for url in urls:
html = '<style> url(%s) </style>' % url
s = lxml.html.fragment_fromstring(html)
cleaned = lxml.html.tostring(clean_html(s))
self.assertEqual(
html.encode("UTF-8"),
cleaned,
"%s -> %s" % (url, cleaned))
def test_formaction_attribute_in_button_input(self):
# The formaction attribute overrides the form's action and should be
# treated as a malicious link attribute
html = ('<form id="test"><input type="submit" formaction="javascript:alert(1)"></form>'
'<button form="test" formaction="javascript:alert(1)">X</button>')
expected = ('<div><form id="test"><input type="submit" formaction=""></form>'
'<button form="test" formaction="">X</button></div>')
cleaner = Cleaner(
forms=False,
safe_attrs_only=False,
)
self.assertEqual(
expected,
cleaner.clean_html(html))
def test_suite():
suite = unittest.TestSuite()
suite.addTests([make_doctest('test_clean.txt')])
suite.addTests([make_doctest('test_clean_embed.txt')])
suite.addTests(unittest.makeSuite(CleanerTest))
return suite
|
