backup/warcprox
1
0
Fork 0
mirror of https://github.com/internetarchive/warcprox.git synced 2025-01-18 13:22:09 +01:00
Code Issues Projects Releases Activity

renaming scope rule "host" to "domain" to make it a less confusing, since rules apply to subdomains as well

Browse Source
This commit is contained in:
Noah Levitt 2016-06-28 15:35:02 -05:00
parent 04c21408d7
commit 04c4b63f03
3 changed files with 17 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
setup.py
View File

@ -51,7 +51,7 @@ except:
setuptools.setup(
name='warcprox',
version='2.0.dev15',
version='2.0.dev16',
description='WARC writing MITM HTTP/S proxy',
url='https://github.com/internetarchive/warcprox',
author='Noah Levitt',

28
tests/test_warcprox.py
View File

@ -732,7 +732,7 @@ def test_dedup_buckets(https_daemon, http_daemon, warcprox_, archiving_proxies,
def test_block_rules(http_daemon, https_daemon, warcprox_, archiving_proxies):
rules = [
{
"host": "localhost",
"domain": "localhost",
"url_match": "STRING_MATCH",
"value": "bar",
},
@ -746,7 +746,7 @@ def test_block_rules(http_daemon, https_daemon, warcprox_, archiving_proxies):
"value": "http://(localhost:%s,)/fuh/" % (https_daemon.server_port),
},
{
"host": "badhost.com",
"domain": "bad.domain.com",
},
]
request_meta = {"blocks":rules}
@ -790,16 +790,16 @@ def test_block_rules(http_daemon, https_daemon, warcprox_, archiving_proxies):
verify=False)
assert response.status_code == 200
# blocked by blanket host block
url = 'http://badhost.com/'
# blocked by blanket domain block
url = 'http://bad.domain.com/'
response = requests.get(
url, proxies=archiving_proxies, headers=headers, stream=True)
assert response.status_code == 403
assert response.content.startswith(b"request rejected by warcprox: blocked by rule found in Warcprox-Meta header:")
assert json.loads(response.headers['warcprox-meta']) == {"blocked-by-rule":rules[3]}
# blocked by blanket host block
url = 'https://badhost.com/'
# blocked by blanket domain block
url = 'https://bad.domain.com/'
response = requests.get(
url, proxies=archiving_proxies, headers=headers, stream=True,
verify=False)
@ -807,24 +807,24 @@ def test_block_rules(http_daemon, https_daemon, warcprox_, archiving_proxies):
assert response.content.startswith(b"request rejected by warcprox: blocked by rule found in Warcprox-Meta header:")
assert json.loads(response.headers['warcprox-meta']) == {"blocked-by-rule":rules[3]}
# blocked by blanket host block
url = 'http://badhost.com:1234/'
# blocked by blanket domain block
url = 'http://bad.domain.com:1234/'
response = requests.get(
url, proxies=archiving_proxies, headers=headers, stream=True)
assert response.status_code == 403
assert response.content.startswith(b"request rejected by warcprox: blocked by rule found in Warcprox-Meta header:")
assert json.loads(response.headers['warcprox-meta']) == {"blocked-by-rule":rules[3]}
# blocked by blanket host block
url = 'http://foo.bar.badhost.com/'
# blocked by blanket domain block
url = 'http://foo.bar.bad.domain.com/'
response = requests.get(
url, proxies=archiving_proxies, headers=headers, stream=True)
assert response.status_code == 403
assert response.content.startswith(b"request rejected by warcprox: blocked by rule found in Warcprox-Meta header:")
assert json.loads(response.headers['warcprox-meta']) == {"blocked-by-rule":rules[3]}
# host block also applies to subdomains
url = 'https://foo.bar.badhost.com/'
# domain block also applies to subdomains
url = 'https://foo.bar.bad.domain.com/'
response = requests.get(
url, proxies=archiving_proxies, headers=headers, stream=True,
verify=False)
@ -832,8 +832,8 @@ def test_block_rules(http_daemon, https_daemon, warcprox_, archiving_proxies):
assert response.content.startswith(b"request rejected by warcprox: blocked by rule found in Warcprox-Meta header:")
assert json.loads(response.headers['warcprox-meta']) == {"blocked-by-rule":rules[3]}
# blocked by blanket host block
url = 'http://foo.bar.badhost.com:1234/'
# blocked by blanket domain block
url = 'http://foo.bar.bad.domain.com:1234/'
response = requests.get(
url, proxies=archiving_proxies, headers=headers, stream=True)
assert response.status_code == 403

4
warcprox/warcproxy.py
View File

@ -120,7 +120,7 @@ class WarcProxyHandler(warcprox.mitmproxy.MitmProxyHandler):
def _scope_rule_applies(self, rule):
u = Url(self.url)
if "host" in rule and not u.matches_ip_or_domain(rule["host"]):
if "domain" in rule and not u.matches_ip_or_domain(rule["domain"]):
return False
if "url_match" in rule:
if rule["url_match"] == "STRING_MATCH":
@ -139,7 +139,7 @@ class WarcProxyHandler(warcprox.mitmproxy.MitmProxyHandler):
self.logger.warn("invalid rule.url_match=%s", rule.url_match)
return False
else:
if "host" in rule:
if "domain" in rule:
# we already know that it matches from earlier check
return True
else: