Benjamin_Loison/waybackpy
1
0
Fork 0
Code Issues 1 Pull Requests Packages Projects Releases Wiki Activity

add: type annotation to test modules

Browse Source
This commit is contained in:
eggplants
2022-02-04 04:25:18 +09:00
parent 38088fa0d8
commit e8b8bf853d
7 changed files with 45 additions and 38 deletions
Download Patch File Download Diff File Expand all files Collapse all files

0
tests/__init__.py
View File

19
tests/test_availability_api.py
View File

@@ -15,13 +15,13 @@ url = "https://example.com/"
user_agent = "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/97.0.4692.99 Safari/537.36"
def rndstr(n):
def rndstr(n: int) -> str:
return "".join(
random.choice(string.ascii_uppercase + string.digits) for _ in range(n)
)
def test_oldest():
def test_oldest() -> None:
"""
Test the oldest archive of Google.com and also checks the attributes.
"""
@@ -33,12 +33,15 @@ def test_oldest():
assert "2002" in oldest_archive_url
oldest_timestamp = oldest.timestamp()
assert abs(oldest_timestamp - now) > timedelta(days=7000) # More than 19 years
assert availability_api.JSON["archived_snapshots"]["closest"]["available"] is True
assert (
availability_api.JSON is not None
and availability_api.JSON["archived_snapshots"]["closest"]["available"] is True
)
assert repr(oldest).find("example.com") != -1
assert "2002" in str(oldest)
def test_newest():
def test_newest() -> None:
"""
Assuming that the recent most Google Archive was made no more earlier than
last one day which is 86400 seconds.
@@ -54,7 +57,7 @@ def test_newest():
assert abs(newest_timestamp - now) < timedelta(seconds=86400 * 3)
def test_invalid_json():
def test_invalid_json() -> None:
"""
When the API is malfunctioning or we don't pass a URL it may return invalid JSON data.
"""
@@ -63,7 +66,7 @@ def test_invalid_json():
_ = availability_api.archive_url
def test_no_archive():
def test_no_archive() -> None:
"""
ArchiveNotInAvailabilityAPIResponse may be raised if Wayback Machine did not
replied with the archive despite the fact that we know the site has million
@@ -79,7 +82,7 @@ def test_no_archive():
_ = availability_api.archive_url
def test_no_api_call_str_repr():
def test_no_api_call_str_repr() -> None:
"""
Some entitled users maybe want to see what is the string representation
if they don’t make any API requests.
@@ -92,7 +95,7 @@ def test_no_api_call_str_repr():
assert "" == str(availability_api)
def test_no_call_timestamp():
def test_no_call_timestamp() -> None:
"""
If no API requests were made the bound timestamp() method returns
the datetime.max as a default value.

4
tests/test_cdx_api.py
View File

@@ -1,7 +1,7 @@
from waybackpy.cdx_api import WaybackMachineCDXServerAPI
def test_a():
def test_a() -> None:
user_agent = "Mozilla/5.0 (MacBook Air; M1 Mac OS X 11_4) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/14.1.1 Safari/604.1"
url = "https://twitter.com/jack"
@@ -21,7 +21,7 @@ def test_a():
assert snapshot.timestamp.startswith("2010")
def test_b():
def test_b() -> None:
user_agent = "Mozilla/5.0 (MacBook Air; M1 Mac OS X 11_4) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/14.1.1 Safari/604.1"
url = "https://www.google.com"

2
tests/test_cdx_snapshot.py
View File

@@ -3,7 +3,7 @@ from datetime import datetime
from waybackpy.cdx_snapshot import CDXSnapshot
def test_CDXSnapshot():
def test_CDXSnapshot() -> None:
sample_input = "org,archive)/ 20080126045828 http://github.com text/html 200 Q4YULN754FHV2U6Q5JUT6Q2P57WEWNNY 1415"
prop_values = sample_input.split(" ")
properties = {}

30
tests/test_cdx_utils.py
View File

@@ -1,3 +1,5 @@
from typing import Any, Dict, List
import pytest
from waybackpy.cdx_utils import (
@@ -11,15 +13,15 @@ from waybackpy.cdx_utils import (
from waybackpy.exceptions import WaybackError
def test_get_total_pages():
def test_get_total_pages() -> None:
url = "twitter.com"
user_agent = "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_6) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/14.0.2 Safari/605.1.15"
assert get_total_pages(url=url, user_agent=user_agent) >= 56
def test_full_url():
params = {}
def test_full_url() -> None:
endpoint = "https://web.archive.org/cdx/search/cdx"
params: Dict[str, Any] = {}
assert endpoint == full_url(endpoint, params)
params = {"a": "1"}
@@ -39,36 +41,36 @@ def test_full_url():
)
def test_get_response():
def test_get_response() -> None:
url = "https://github.com"
user_agent = (
"Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:78.0) Gecko/20100101 Firefox/78.0"
)
headers = {"User-Agent": "%s" % user_agent}
response = get_response(url, headers=headers)
assert response.status_code == 200
assert not isinstance(response, Exception) and response.status_code == 200
url = "http/wwhfhfvhvjhmom"
with pytest.raises(WaybackError):
get_response(url, headers=headers)
def test_check_filters():
filters = []
def test_check_filters() -> None:
filters: List[str] = []
check_filters(filters)
filters = ["statuscode:200", "timestamp:20215678901234", "original:https://url.com"]
check_filters(filters)
with pytest.raises(WaybackError):
check_filters("not-list")
check_filters("not-list") # type: ignore[arg-type]
with pytest.raises(WaybackError):
check_filters(["invalid"])
def test_check_collapses():
collapses = []
def test_check_collapses() -> None:
collapses: List[str] = []
check_collapses(collapses)
collapses = ["timestamp:10"]
@@ -77,7 +79,7 @@ def test_check_collapses():
collapses = ["urlkey"]
check_collapses(collapses)
collapses = "urlkey" # NOT LIST
collapses = "urlkey" # type: ignore[assignment]
with pytest.raises(WaybackError):
check_collapses(collapses)
@@ -86,11 +88,11 @@ def test_check_collapses():
check_collapses(collapses)
def test_check_match_type():
assert check_match_type(None, "url") is None
def test_check_match_type() -> None:
assert check_match_type(None, "url")
match_type = "exact"
url = "test_url"
assert check_match_type(match_type, url) is None
assert check_match_type(match_type, url)
url = "has * in it"
with pytest.raises(WaybackError):

24
tests/test_save_api.py
View File

@@ -9,13 +9,13 @@ from waybackpy.exceptions import MaximumSaveRetriesExceeded
from waybackpy.save_api import WaybackMachineSaveAPI
def rndstr(n):
def rndstr(n: int) -> str:
return "".join(
random.choice(string.ascii_uppercase + string.digits) for _ in range(n)
)
def test_save():
def test_save() -> None:
url = "https://github.com/akamhy/waybackpy"
user_agent = "Mozilla/5.0 (MacBook Air; M1 Mac OS X 11_4) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/14.1.1 Safari/604.1"
save_api = WaybackMachineSaveAPI(url, user_agent)
@@ -31,7 +31,7 @@ def test_save():
assert isinstance(save_api.timestamp(), datetime)
def test_max_redirect_exceeded():
def test_max_redirect_exceeded() -> None:
with pytest.raises(MaximumSaveRetriesExceeded):
url = "https://%s.gov" % rndstr
user_agent = "Mozilla/5.0 (MacBook Air; M1 Mac OS X 11_4) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/14.1.1 Safari/604.1"
@@ -39,7 +39,7 @@ def test_max_redirect_exceeded():
save_api.save()
def test_sleep():
def test_sleep() -> None:
"""
sleeping is actually very important for SaveAPI
interface stability.
@@ -60,7 +60,7 @@ def test_sleep():
assert (e_time - s_time) >= 5
def test_timestamp():
def test_timestamp() -> None:
url = "https://example.com"
user_agent = "Mozilla/5.0 (MacBook Air; M1 Mac OS X 11_4) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/14.1.1 Safari/604.1"
save_api = WaybackMachineSaveAPI(url, user_agent)
@@ -75,7 +75,7 @@ def test_timestamp():
assert save_api.cached_save is True
def test_archive_url_parser():
def test_archive_url_parser() -> None:
"""
Testing three regex for matches and also tests the response URL.
"""
@@ -83,7 +83,7 @@ def test_archive_url_parser():
user_agent = "Mozilla/5.0 (MacBook Air; M1 Mac OS X 11_4) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/14.1.1 Safari/604.1"
save_api = WaybackMachineSaveAPI(url, user_agent)
save_api.headers = """
save_api.headers_str = """
START
Content-Location: /web/20201126185327/https://www.scribbr.com/citing-sources/et-al
END
@@ -94,7 +94,7 @@ def test_archive_url_parser():
== "https://web.archive.org/web/20201126185327/https://www.scribbr.com/citing-sources/et-al"
)
save_api.headers = """
save_api.headers_str = """
{'Server': 'nginx/1.15.8', 'Date': 'Sat, 02 Jan 2021 09:40:25 GMT', 'Content-Type': 'text/html; charset=UTF-8', 'Transfer-Encoding': 'chunked', 'Connection': 'keep-alive', 'X-Archive-Orig-Server': 'nginx', 'X-Archive-Orig-Date': 'Sat, 02 Jan 2021 09:40:09 GMT', 'X-Archive-Orig-Transfer-Encoding': 'chunked', 'X-Archive-Orig-Connection': 'keep-alive', 'X-Archive-Orig-Vary': 'Accept-Encoding', 'X-Archive-Orig-Last-Modified': 'Fri, 01 Jan 2021 12:19:00 GMT', 'X-Archive-Orig-Strict-Transport-Security': 'max-age=31536000, max-age=0;', 'X-Archive-Guessed-Content-Type': 'text/html', 'X-Archive-Guessed-Charset': 'utf-8', 'Memento-Datetime': 'Sat, 02 Jan 2021 09:40:09 GMT', 'Link': '<https://www.scribbr.com/citing-sources/et-al/>; rel="original", <https://web.archive.org/web/timemap/link/https://www.scribbr.com/citing-sources/et-al/>; rel="timemap"; type="application/link-format", <https://web.archive.org/web/https://www.scribbr.com/citing-sources/et-al/>; rel="timegate", <https://web.archive.org/web/20200601082911/https://www.scribbr.com/citing-sources/et-al/>; rel="first memento"; datetime="Mon, 01 Jun 2020 08:29:11 GMT", <https://web.archive.org/web/20201126185327/https://www.scribbr.com/citing-sources/et-al/>; rel="prev memento"; datetime="Thu, 26 Nov 2020 18:53:27 GMT", <https://web.archive.org/web/20210102094009/https://www.scribbr.com/citing-sources/et-al/>; rel="memento"; datetime="Sat, 02 Jan 2021 09:40:09 GMT", <https://web.archive.org/web/20210102094009/https://www.scribbr.com/citing-sources/et-al/>; rel="last memento"; datetime="Sat, 02 Jan 2021 09:40:09 GMT"', 'Content-Security-Policy': "default-src 'self' 'unsafe-eval' 'unsafe-inline' data: blob: archive.org web.archive.org analytics.archive.org pragma.archivelab.org", 'X-Archive-Src': 'spn2-20210102092956-wwwb-spn20.us.archive.org-8001.warc.gz', 'Server-Timing': 'captures_list;dur=112.646325, exclusion.robots;dur=0.172010, exclusion.robots.policy;dur=0.158205, RedisCDXSource;dur=2.205932, esindex;dur=0.014647, LoadShardBlock;dur=82.205012, PetaboxLoader3.datanode;dur=70.750239, CDXLines.iter;dur=24.306278, load_resource;dur=26.520179', 'X-App-Server': 'wwwb-app200', 'X-ts': '200', 'X-location': 'All', 'X-Cache-Key': 'httpsweb.archive.org/web/20210102094009/https://www.scribbr.com/citing-sources/et-al/IN', 'X-RL': '0', 'X-Page-Cache': 'MISS', 'X-Archive-Screenname': '0', 'Content-Encoding': 'gzip'}
"""
@@ -103,7 +103,7 @@ def test_archive_url_parser():
== "https://web.archive.org/web/20210102094009/https://www.scribbr.com/citing-sources/et-al/"
)
save_api.headers = """
save_api.headers_str = """
START
X-Cache-Key: https://web.archive.org/web/20171128185327/https://www.scribbr.com/citing-sources/et-al/US
END
@@ -114,7 +114,9 @@ def test_archive_url_parser():
== "https://web.archive.org/web/20171128185327/https://www.scribbr.com/citing-sources/et-al/"
)
save_api.headers = "TEST TEST TEST AND NO MATCH - TEST FOR RESPONSE URL MATCHING"
save_api.headers_str = (
"TEST TEST TEST AND NO MATCH - TEST FOR RESPONSE URL MATCHING"
)
save_api.response_url = "https://web.archive.org/web/20171128185327/https://www.scribbr.com/citing-sources/et-al"
assert (
save_api.archive_url_parser()
@@ -122,7 +124,7 @@ def test_archive_url_parser():
)
def test_archive_url():
def test_archive_url() -> None:
"""
Checks the attribute archive_url's value when the save method was not
explicitly invoked by the end-user but the save method was invoked implicitly

4
tests/test_utils.py
View File

@@ -6,13 +6,13 @@ from waybackpy.utils import (
)
def test_default_user_agent():
def test_default_user_agent() -> None:
assert (
DEFAULT_USER_AGENT
== "waybackpy %s - https://github.com/akamhy/waybackpy" % __version__
)
def test_latest_version():
def test_latest_version() -> None:
package_name = "waybackpy"
assert latest_version_github(package_name) == latest_version_pypi(package_name)