test_validation.py — 10Web Booster – Website speed optimization, Cache & Page Speed optimizer trunk

tenweb-speed-optimizer / vendor / monperrus / crawler-user-agents / test_validation.py

tenweb-speed-optimizer / vendor / monperrus / crawler-user-agents Last commit date

.github 1 year ago .gitignore 1 year ago .travis.yml 2 years ago LICENSE 3 years ago MANIFEST.in 2 years ago README.md 1 year ago __init__.py 1 year ago composer.json 3 years ago crawler-user-agents.json 1 year ago format.js 1 year ago go.mod 2 years ago go.sum 2 years ago index.d.ts 3 years ago main.php 3 years ago package-lock.json 2 years ago package.json 2 years ago pyproject.toml 1 year ago test_harness.py 1 year ago test_validation.py 3 years ago validate.go 1 year ago validate.php 3 years ago validate.py 2 years ago validate_test.go 1 year ago

test_validation.py

189 lines

1	"""
2	Break crawler-user-agents.json in ways that validate.py should detect
3
4	Usage:
5	$ pytest test_validation.py
6
7	"""
8	from __future__ import print_function
9
10	import json
11	import subprocess
12	from copy import deepcopy
13
14	import pytest
15
16
17	def update_json_file(data):
18	with open('crawler-user-agents.json', 'w') as f:
19	json.dump(data, f, indent=2)
20
21	@pytest.fixture
22	def restore_original_json():
23	# Load original version of crawler-user-agents.json
24	with open('crawler-user-agents.json') as f:
25	original_json = json.load(f)
26
27	# By using a yield statement instead of return, all the code after
28	# the yield statement serves as the teardown code:
29	yield None
30
31	# tear down code: restore original version of crawler-user-agents.json
32	update_json_file(original_json)
33
34
35	def assert_validate_failed():
36	assert subprocess.call(['python', 'validate.py']) != 0
37
38
39	def assert_validate_passed():
40	assert subprocess.call(['python', 'validate.py']) == 0
41
42
43	def test_simple_pass(restore_original_json):
44	# the json must be an array of objects containing "pattern"
45	# there must be more than 10 instances to pass
46	user_agent_list = [{'pattern': 'foo',
47	'instances': ['foo',
48	'afoo',
49	'foob',
50	'cfood',
51	'/foo/',
52	'\\foo\\',
53	':foo:',
54	'foo.',
55	'!foo',
56	'/foo',
57	'foo\\',
58	'FoofooFoo',
59	'foot']}]
60	update_json_file(user_agent_list)
61	assert_validate_passed()
62
63	def test_simplest_pass(restore_original_json):
64	# the simplest crawler file passes
65	user_agent_list = [{'pattern': 'foo', 'instances': []}]
66	update_json_file(user_agent_list)
67	assert_validate_passed()
68
69	def test_schema_violation_dict1(restore_original_json):
70	# contract: the json must be an array
71	user_agent_list = {'foo':None}
72	update_json_file(user_agent_list)
73	assert_validate_failed()
74
75	def test_schema_violation_dict2(restore_original_json):
76	# contract: the json must be an array of objects containing "pattern"
77	user_agent_list = [{'foo':None}]
78	update_json_file(user_agent_list)
79	assert_validate_failed()
80
81	def test_schema_violation_dict3(restore_original_json):
82	# contract: the json must be an array of objects containing "pattern" and valid properties
83	user_agent_list = [{'pattern':'foo', 'foo':3}]
84	update_json_file(user_agent_list)
85	assert_validate_failed()
86
87	def test_simple_duplicate_detection(restore_original_json):
88	# contract: if we have the same pattern twice, it fails
89	user_agent_list = [{'pattern': 'foo',
90	'instances': ['foo',
91	'afoo',
92	'foob',
93	'cfood',
94	'/foo/',
95	'\\foo\\',
96	':foO:',
97	'foo.',
98	'!foo',
99	'/foo',
100	'foo\\',
101	'FoofooFoo',
102	'foot']},
103	{'pattern': 'foo'}]
104	update_json_file(user_agent_list)
105	assert_validate_failed()
106
107
108	def test_simple_duplicate_detection2(restore_original_json):
109	# contract: if we have the same pattern twice, it fails (even w/o instances)
110	user_agent_list = [{'pattern': 'foo',
111	'instances': ['foo',
112	'afoo',
113	'foob',
114	'cfood',
115	'/foo/',
116	'\\foo\\',
117	':foo:',
118	'foo.',
119	'!foo',
120	'/foo',
121	'foo\\',
122	'FoofooFoo',
123	'foot']},
124	{'pattern': 'bar'},
125	{'pattern': 'bar'}]
126	update_json_file(user_agent_list)
127	assert_validate_failed()
128
129
130	def test_subset_duplicate_detection(restore_original_json):
131	# contract: if a pattern matches another pattern, it fails
132	user_agent_list = [{'pattern': 'foo',
133	'instances': ['foo',
134	'afoo',
135	'foob',
136	'cfood',
137	'/foo/',
138	'\\foo\\',
139	':foo:',
140	'foo.',
141	'!foo',
142	'/foo',
143	'foo\\',
144	'FoofooFoo',
145	'foot']},
146	{'pattern': 'afoot'}]
147	update_json_file(user_agent_list)
148	assert_validate_failed()
149
150
151	def test_case_sensitivity(restore_original_json):
152	# contract: the patterns are case sensitive
153	user_agent_list = [{'pattern': 'foo',
154	'instances': ['FOO',
155	'aFoo',
156	'foob',
157	'cfood',
158	'/foo/',
159	'\\foo\\',
160	':foo:',
161	'fOo.',
162	'!FOO',
163	'/foo',
164	'foo\\',
165	'FoofooFoo',
166	'foot']}]
167	update_json_file(user_agent_list)
168	assert_validate_failed()
169
170	def test_duplicate_case_insensitive_detection(restore_original_json):
171	# contract: fail if we have patterns that differ only in capitailization
172	user_agent_list = [{'pattern': 'foo',
173	'instances': ['foo',
174	'afoo',
175	'foob',
176	'cfood',
177	'/foo/',
178	'\\foo\\',
179	':foo:',
180	'foo.',
181	'!foo',
182	'/foo',
183	'foo\\',
184	'FoofooFoo',
185	'foot']},
186	{'pattern': 'fOo'}]
187	update_json_file(user_agent_list)
188	assert_validate_failed()
189