f990b93f8a
* add sort param support in CDX API class see https://nla.github.io/outbackcdx/api.html#operation/query sort takes string input which must be one of the follwoing: - default - closest - reverse This commit shall help in closing issue at https://github.com/akamhy/waybackpy/issues/155 * add BlockedSiteError for cases when archiving is blocked by site's robots.txt * create check_for_blocked_site for handling the BlockedSiteError for sites that are blocking wayback machine by their robots.txt policy * add attrs use_pagination and closest, which are can be used to use the pagination API and lookup archive close to a timestamp respectively. And now to get out of infinte blank pages loop just check for two succesive black and not total two blank pages while using the CDX server API. * added cli support for sort, use-pagination and closest * added tests * fix codeql warnings, nothing to worry about here. * fix save test for archive_url
93 lines
2.5 KiB
Python
93 lines
2.5 KiB
Python
from waybackpy.cdx_api import WaybackMachineCDXServerAPI
|
|
|
|
|
|
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"
|
|
|
|
wayback = WaybackMachineCDXServerAPI(
|
|
url=url,
|
|
user_agent=user_agent,
|
|
match_type="prefix",
|
|
collapses=["urlkey"],
|
|
start_timestamp="201001",
|
|
end_timestamp="201002",
|
|
)
|
|
# timeframe bound prefix matching enabled along with active urlkey based collapsing
|
|
|
|
snapshots = wayback.snapshots() # <class 'generator'>
|
|
|
|
for snapshot in snapshots:
|
|
assert snapshot.timestamp.startswith("2010")
|
|
|
|
|
|
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"
|
|
|
|
wayback = WaybackMachineCDXServerAPI(
|
|
url=url,
|
|
user_agent=user_agent,
|
|
start_timestamp="202101",
|
|
end_timestamp="202112",
|
|
collapses=["urlkey"],
|
|
)
|
|
# timeframe bound prefix matching enabled along with active urlkey based collapsing
|
|
|
|
snapshots = wayback.snapshots() # <class 'generator'>
|
|
|
|
for snapshot in snapshots:
|
|
assert snapshot.timestamp.startswith("2021")
|
|
|
|
|
|
def test_c() -> 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"
|
|
|
|
cdx = WaybackMachineCDXServerAPI(
|
|
url=url,
|
|
user_agent=user_agent,
|
|
closest="201010101010",
|
|
sort="closest",
|
|
limit="1",
|
|
)
|
|
snapshots = cdx.snapshots()
|
|
for snapshot in snapshots:
|
|
archive_url = snapshot.archive_url
|
|
timestamp = snapshot.timestamp
|
|
break
|
|
|
|
assert str(archive_url).find("google.com")
|
|
assert "20101010" in timestamp
|
|
|
|
|
|
def test_d() -> 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"
|
|
)
|
|
|
|
cdx = WaybackMachineCDXServerAPI(
|
|
url="akamhy.github.io",
|
|
user_agent=user_agent,
|
|
match_type="prefix",
|
|
use_pagination=True,
|
|
filters=["statuscode:200"],
|
|
)
|
|
snapshots = cdx.snapshots()
|
|
|
|
count = 0
|
|
for snapshot in snapshots:
|
|
count += 1
|
|
assert str(snapshot.archive_url).find("akamhy.github.io")
|
|
assert count > 50
|