Skip to content
Projects
Groups
Snippets
Help
Loading...
Help
Support
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
S
surykatka
Project overview
Project overview
Details
Activity
Releases
Repository
Repository
Files
Commits
Branches
Tags
Contributors
Graph
Compare
Issues
0
Issues
0
List
Boards
Labels
Milestones
Merge Requests
0
Merge Requests
0
Analytics
Analytics
Repository
Value Stream
Wiki
Wiki
Snippets
Snippets
Members
Members
Collapse sidebar
Close sidebar
Activity
Graph
Create a new issue
Commits
Issue Boards
Open sidebar
Romain Courteaud
surykatka
Commits
e335e8fb
Commit
e335e8fb
authored
Jun 14, 2022
by
Romain Courteaud
Browse files
Options
Browse Files
Download
Email Patches
Plain Diff
Add CONTACT configuration parameter.
Surcharge the bot User-Agent contact url/email.
parent
c3769d8b
Changes
4
Hide whitespace changes
Inline
Side-by-side
Showing
4 changed files
with
54 additions
and
26 deletions
+54
-26
src/surykatka/bot.py
src/surykatka/bot.py
+2
-0
src/surykatka/configuration.py
src/surykatka/configuration.py
+4
-0
src/surykatka/http.py
src/surykatka/http.py
+21
-9
tests/test_http.py
tests/test_http.py
+27
-17
No files found.
src/surykatka/bot.py
View file @
e335e8fb
...
...
@@ -328,6 +328,7 @@ class WebBot:
timeout
=
int
(
self
.
config
[
"TIMEOUT"
])
elapsed_fast
=
float
(
self
.
config
[
"ELAPSED_FAST"
])
elapsed_moderate
=
float
(
self
.
config
[
"ELAPSED_MODERATE"
])
contact
=
self
.
config
[
"CONTACT"
]
# logPlatform(self._db, __version__, status_id)
# Get list of all domains
...
...
@@ -420,6 +421,7 @@ class WebBot:
url
,
ip
,
__version__
,
contact
,
timeout
,
elapsed_fast
,
elapsed_moderate
,
...
...
src/surykatka/configuration.py
View file @
e335e8fb
...
...
@@ -64,6 +64,10 @@ def createConfiguration(
config
[
CONFIG_SECTION
][
"ELAPSED_FAST"
]
=
"-1"
if
"ELAPSED_MODERATE"
not
in
config
[
CONFIG_SECTION
]:
config
[
CONFIG_SECTION
][
"ELAPSED_MODERATE"
]
=
"-1"
if
"CONTACT"
not
in
config
[
CONFIG_SECTION
]:
config
[
CONFIG_SECTION
][
"CONTACT"
]
=
"https://lab.nexedi.com/nexedi/surykatka"
if
"RELOAD"
not
in
config
[
CONFIG_SECTION
]:
config
[
CONFIG_SECTION
][
"RELOAD"
]
=
str
(
False
)
...
...
src/surykatka/http.py
View file @
e335e8fb
...
...
@@ -40,15 +40,22 @@ def getRootUrl(url):
return
"%s://%s/"
%
(
parsed_url
.
scheme
,
parsed_url
.
hostname
)
def
getUserAgent
(
version
):
return
"%s/%s (+%s)"
%
(
"SURYKATKA"
,
version
,
"https://lab.nexedi.com/nexedi/surykatka"
,
)
def
getUserAgent
(
version
,
contact
=
None
):
if
contact
:
contact
=
" (+%s)"
%
contact
else
:
contact
=
""
return
"%s/%s%s"
%
(
"SURYKATKA"
,
version
,
contact
)
def
request
(
url
,
timeout
=
TIMEOUT
,
headers
=
None
,
session
=
requests
,
version
=
0
):
def
request
(
url
,
timeout
=
TIMEOUT
,
headers
=
None
,
session
=
requests
,
contact
=
None
,
version
=
0
,
):
if
headers
is
None
:
headers
=
{}
...
...
@@ -56,7 +63,7 @@ def request(url, timeout=TIMEOUT, headers=None, session=requests, version=0):
headers
[
"Accept"
]
=
"%s;q=0.9,*/*;q=0.8"
%
PREFERRED_TYPE
if
"User-Agent"
not
in
headers
:
# XXX user agent
headers
[
"User-Agent"
]
=
getUserAgent
(
version
)
headers
[
"User-Agent"
]
=
getUserAgent
(
version
,
contact
)
kwargs
=
{}
kwargs
[
"stream"
]
=
False
...
...
@@ -188,6 +195,7 @@ def checkHttpStatus(
url
,
ip
,
bot_version
,
contact
,
timeout
=
TIMEOUT
,
elapsed_fast
=
ELAPSED_FAST
,
elapsed_moderate
=
ELAPSED_MODERATE
,
...
...
@@ -213,7 +221,11 @@ def checkHttpStatus(
raise
NotImplementedError
(
"Unhandled url: %s"
%
url
)
response
=
request
(
ip_url
,
headers
=
{
"Host"
:
hostname
},
version
=
bot_version
,
**
request_kw
ip_url
,
headers
=
{
"Host"
:
hostname
},
contact
=
contact
,
version
=
bot_version
,
**
request_kw
,
)
# Blacklisted, because of non stability
...
...
tests/test_http.py
View file @
e335e8fb
...
...
@@ -58,11 +58,12 @@ class SurykatkaHttpTestCase(unittest.TestCase):
# getUserAgent
################################################
def
test_getUserAgent_version
(
self
):
result
=
getUserAgent
(
"0.0.3"
)
assert
(
result
==
"SURYKATKA/0.0.3 (+https://lab.nexedi.com/nexedi/surykatka)"
)
result
=
getUserAgent
(
"0.0.3"
,
"foocontact"
)
assert
result
==
"SURYKATKA/0.0.3 (+foocontact)"
def
test_getUserAgent_no_contact
(
self
):
result
=
getUserAgent
(
"0.0.4"
)
assert
result
==
"SURYKATKA/0.0.4"
################################################
# request
...
...
@@ -78,7 +79,7 @@ class SurykatkaHttpTestCase(unittest.TestCase):
allow_redirects
=
False
,
headers
=
{
"Accept"
:
"text/html;q=0.9,*/*;q=0.8"
,
"User-Agent"
:
"SURYKATKA/0
(+https://lab.nexedi.com/nexedi/surykatka)
"
,
"User-Agent"
:
"SURYKATKA/0"
,
},
stream
=
False
,
timeout
=
2
,
...
...
@@ -97,10 +98,7 @@ class SurykatkaHttpTestCase(unittest.TestCase):
assert
last_request
.
headers
[
"Accept-Encoding"
]
==
"gzip, deflate"
assert
last_request
.
headers
[
"Connection"
]
==
"keep-alive"
assert
last_request
.
headers
[
"Host"
]
==
"example.org"
assert
(
last_request
.
headers
[
"User-Agent"
]
==
"SURYKATKA/0 (+https://lab.nexedi.com/nexedi/surykatka)"
)
assert
last_request
.
headers
[
"User-Agent"
]
==
"SURYKATKA/0"
assert
len
(
last_request
.
body
)
==
0
assert
response
.
status_code
==
418
...
...
@@ -898,6 +896,7 @@ class SurykatkaHttpTestCase(unittest.TestCase):
ip
=
"127.0.0.1"
url
=
"http://example.org/foo?bar=1"
bot_version
=
1
bot_contact
=
"http://example.org/contact10"
httpretty
.
register_uri
(
httpretty
.
GET
,
"http://127.0.0.1/foo?bar=1"
,
...
...
@@ -906,7 +905,7 @@ class SurykatkaHttpTestCase(unittest.TestCase):
)
status_id
=
logStatus
(
self
.
db
,
"foo"
)
checkHttpStatus
(
self
.
db
,
status_id
,
url
,
ip
,
bot_version
)
checkHttpStatus
(
self
.
db
,
status_id
,
url
,
ip
,
bot_version
,
bot_contact
)
last_request
=
httpretty
.
last_request
()
assert
len
(
last_request
.
headers
)
==
5
,
last_request
.
headers
.
keys
()
...
...
@@ -916,7 +915,7 @@ class SurykatkaHttpTestCase(unittest.TestCase):
assert
last_request
.
headers
[
"Host"
]
==
"example.org"
assert
(
last_request
.
headers
[
"User-Agent"
]
==
"SURYKATKA/1 (+http
s://lab.nexedi.com/nexedi/surykatka
)"
==
"SURYKATKA/1 (+http
://example.org/contact10
)"
)
assert
len
(
last_request
.
body
)
==
0
...
...
@@ -935,6 +934,7 @@ class SurykatkaHttpTestCase(unittest.TestCase):
ip
=
"127.0.0.1"
url
=
"https://example.org/foo?bar=1"
bot_version
=
2
bot_contact
=
"http://example.org/contact3"
status_id
=
logStatus
(
self
.
db
,
"foo"
)
with
mock
.
patch
(
"surykatka.http.request"
)
as
mock_request
:
...
...
@@ -943,20 +943,26 @@ class SurykatkaHttpTestCase(unittest.TestCase):
"Cache-Control"
:
"public"
,
}
checkHttpStatus
(
self
.
db
,
status_id
,
url
,
ip
,
bot_version
)
checkHttpStatus
(
self
.
db
,
status_id
,
url
,
ip
,
bot_version
,
bot_contact
)
assert
mock_request
.
call_count
==
1
assert
mock_request
.
call_args
.
args
==
(
"https://example.org/foo?bar=1"
,
)
assert
(
len
(
mock_request
.
call_args
.
kwargs
)
==
4
len
(
mock_request
.
call_args
.
kwargs
)
==
5
),
mock_request
.
call_args
.
kwargs
assert
mock_request
.
call_args
.
kwargs
[
"headers"
]
==
{
"Host"
:
"example.org"
}
assert
mock_request
.
call_args
.
kwargs
[
"session"
]
is
not
None
assert
mock_request
.
call_args
.
kwargs
[
"version"
]
==
2
assert
(
mock_request
.
call_args
.
kwargs
[
"contact"
]
==
"http://example.org/contact3"
)
assert
mock_request
.
call_args
.
kwargs
[
"timeout"
]
==
2
assert
self
.
db
.
HttpCodeChange
.
select
().
count
()
==
1
...
...
@@ -974,10 +980,13 @@ class SurykatkaHttpTestCase(unittest.TestCase):
ip
=
"127.0.0.1"
url
=
"foo?bar=1"
bot_version
=
1
bot_contact
=
"http://example.org/contact"
status_id
=
logStatus
(
self
.
db
,
"foo"
)
try
:
checkHttpStatus
(
self
.
db
,
status_id
,
url
,
ip
,
bot_version
)
checkHttpStatus
(
self
.
db
,
status_id
,
url
,
ip
,
bot_version
,
bot_contact
)
except
NotImplementedError
as
err
:
assert
str
(
err
)
==
"Unhandled url: foo?bar=1"
else
:
...
...
@@ -997,6 +1006,7 @@ class SurykatkaHttpTestCase(unittest.TestCase):
ip
=
"127.0.0.1"
url
=
"http://example.org/foo?bar=1"
bot_version
=
1
bot_contact
=
"http://example.org/contact"
whitelist_header_list
=
[
# Redirect
"Location"
,
...
...
@@ -1046,7 +1056,7 @@ class SurykatkaHttpTestCase(unittest.TestCase):
)
status_id
=
logStatus
(
self
.
db
,
"foo"
)
checkHttpStatus
(
self
.
db
,
status_id
,
url
,
ip
,
bot_version
)
checkHttpStatus
(
self
.
db
,
status_id
,
url
,
ip
,
bot_version
,
bot_contact
)
last_request
=
httpretty
.
last_request
()
assert
len
(
last_request
.
headers
)
==
5
,
last_request
.
headers
.
keys
()
...
...
@@ -1056,7 +1066,7 @@ class SurykatkaHttpTestCase(unittest.TestCase):
assert
last_request
.
headers
[
"Host"
]
==
"example.org"
assert
(
last_request
.
headers
[
"User-Agent"
]
==
"SURYKATKA/1 (+http
s://lab.nexedi.com/nexedi/surykatka
)"
==
"SURYKATKA/1 (+http
://example.org/contact
)"
)
assert
len
(
last_request
.
body
)
==
0
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment