Retry Plausible stats using backoff

[m-aciek]

Better version of #51.

---

📚 Documentation preview 📚: https://pydocs-translation-dashboard--52.org.readthedocs.build/

[willingc]

Thanks!

[mattwang44]

Note that backoff is not actively maintained (ref). The retry mechanism from urllib3 (which has exponential backoff too) could be a good alternative.

from requests.adapters import HTTPAdapter
from requests.sessions import Session
from urllib3.util import Retry

def make_session():
    session = Session()
    retry = Retry(
        total=5,
        backoff_factor=1,
        allowed_methods=None,  # all methods are allowed
        status_forcelist={502, 503, 504},  # add the status code of Plausible error response here
    )
    session.mount('https://', HTTPAdapter(max_retries=retry))
    return session

session = make_session()
session.get(..., timeout=40)

[m-aciek]

Please note that BadZipFile is raised later, after successfully receiving the request. Hm, or not, actually I now realized we don't check the response status code. Let me investigate and try with urllib3.Retry 👍