Issue 6701 - Implement CSP support in python-abp

tests/test_parser.py

 # This file is part of Adblock Plus <https://adblockplus.org/>,
 # Copyright (C) 2006-present eyeo GmbH
 #
 # Adblock Plus is free software: you can redistribute it and/or modify
 # it under the terms of the GNU General Public License version 3 as
 # published by the Free Software Foundation.
 #
 # Adblock Plus is distributed in the hope that it will be useful,
 # but WITHOUT ANY WARRANTY; without even the implied warranty of
 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
@@ skipping 49 matching lines @@
     },
     '@@http://bla$~script,~other,sitekey=foo|bar': {
         'selector': {'type': ST.URL_PATTERN, 'value': 'http://bla'},
         'action': FA.ALLOW,
         'options': [
             (OPT.SCRIPT, False),
             (OPT.OTHER, False),
             (OPT.SITEKEY, ['foo', 'bar']),
         ],
     },
-    "||foo.com^$csp=script-src 'self' * 'unsafe-inline'": {
+    "||foo.com^$csp=script-src 'self' * 'unsafe-inline',script,sitekey=foo," +
+    'other,match-case,domain=foo.com': {
         'selector': {'type': ST.URL_PATTERN, 'value': '||foo.com^'},
         'action': FA.BLOCK,
         'options': [
             (OPT.CSP, "script-src 'self' * 'unsafe-inline'"),
-        ],
-    },
-    'bla$match-case,csp=first csp,script,other,domain=foo.com,sitekey=foo': {

[Vasily Kuznetsov, 2018/06/08 16:41:48]

It seems like this test does more or less the sane as the previous one. That one
has a more tricky CSP option, this one has other options -- maybe we could
combine them into one test.

[rhowell, 2018/06/12 21:13:18]

Done.

-        'selector': {'type': ST.URL_PATTERN, 'value': 'bla'},
-        'action': FA.BLOCK,
-        'options': [
+            ('script', True),
+            ('sitekey', ['foo']),
+            ('other', True),
             ('match-case', True),
-            ('csp', 'first csp'),
-            ('script', True),
-            ('other', True),
             ('domain', [('foo.com', True)]),
-            ('sitekey', ['foo']),
-        ],
-    },
-    'bla$~match-case,~csp=csp,~script,~other,~third-party,domain=~bar.com': {
-        'selector': {'type': ST.URL_PATTERN, 'value': 'bla'},
-        'action': FA.BLOCK,
-        'options': [
-            ('match-case', False),
-            ('csp=csp', False),

[Vasily Kuznetsov, 2018/06/08 16:41:48]

Hm, I think this should be ('csp', False). I guess the simple swapping trick in
`parser.py` is not sufficient and we'll need to be a bit more clever about this.

[rhowell, 2018/06/12 21:13:18]

Done.

-            ('script', False),
-            ('other', False),
-            ('third-party', False),
-            ('domain', [('bar.com', False)]),
         ],
     },
     '@@bla$script,other,domain=foo.com|~bar.foo.com,csp=c s p': {

[Vasily Kuznetsov, 2018/06/08 16:41:48]

This one adds testing, of a CSP option in an exception rule. I guess it's useful
too, but perhaps it could be combined with the test of line 77 or the test of
line 101. All in all, 2 tests for csp=xxx and one for ~csp should probably be
enough.

[rhowell, 2018/06/12 21:13:18]

I already combined the first two csp tests, so now we have one test for csp
block, one for csp allow, and one for ~csp. Good enough? Or, I could combine the
two csp tests, but then it would only test the exception.

         'selector': {'type': ST.URL_PATTERN, 'value': 'bla'},
         'action': FA.ALLOW,
         'options': [
             ('script', True),
             ('other', True),
             ('domain', [('foo.com', True), ('bar.foo.com', False)]),
             ('csp', 'c s p'),
         ],
     },
     # Element hiding filters and exceptions.
@@ skipping 87 matching lines @@
 def test_exception_timing():
     result = parse_filterlist(['! good line', '%bad line%'])
     assert next(result) == Comment('good line')
     with pytest.raises(ParseError):
         next(result)
 
 
 def test_parse_line_bytes():
     line = parse_line(b'! \xc3\xbc')
     assert line.text == '\xfc'