[BugzillaREST] requests.exceptions.HTTPError: 400 Client Error: Bad Request for url

System information

• Python 3.12.8
• perceval==1.3.1

Issue description

When downloading Bugzilla data, the request fails with the error:

requests.exceptions.HTTPError: 400 Client Error: Bad Request for url: https://bugzilla.mozilla.org/rest/bug?last_change_time=2023-02-09T03%3A28%3A55Z&limit=25&order=changeddate&include_fields=_all&offset=25&api_key=<your-bugzilla-api-key>

When executing the URL in the browser, it returns an error as well, however if you increase the offset by amount x (in this example offset=40), it works again.

Steps to reproduce the issue

from perceval.backends.core.bugzillarest import BugzillaREST
from datetime import datetime

if __name__ == '__main__':
    init_args={
        'url': 'https://bugzilla.mozilla.org',
        'api_key': '<your-bugzilla-api-key>',
        'max_bugs': 25,
    }

    fetch_args={
        'from_date': datetime(2023, 1, 1)
    }

    backend = BugzillaREST(**init_args)
    iterator = backend.fetch(**fetch_args)
    for i, issue in enumerate(iterator):
        print('i')

What's the expected result?

• The fetch function should increment offset by amount x until the request returns data again.
• Or take offset as initial kwarg in the fetch function, so the users can skip it themselves.

What's the actual result?

• The fetch function stops with an errro

Start in perceval/backends/core/bugzillarest.py at BugzillaREST.fetch and reproduce the failing request with the issue's Python example. Trace how the offset is passed to Bugzilla and determine how the requested offset behavior should be handled. Done means fetching continues past a 400 response or provides the requested way to skip the problematic offset, with the behavior verified against the reproduction.