Refactored count_bugs to use libmozdata

[rafey1104]

Fix #3676 @suhaibmujahid

Used Bugzilla class from libmozdata.bugzilla instead of doing a direct request

[suhaibmujahid]

The libmozdata API is not used correctly. You could have a look at the example in #3676 as guidance.

Please test locally whether the function works correctly and make sure that your changes pass the CI tests.

I will cover the PR to draft; feel free to convert it back when it is ready for review.

[rafey1104]

Before:-
def count_bugs(bug_query_params):
bug_query_params["count_only"] = 1
r = utils.get_session("bugzilla").get(
"https://bugzilla.mozilla.org/rest/bug", params=bug_query_params
)
r.raise_for_status()
count = r.json()["bug_count"]
return count

After:-
def count_bugs(bug_query_params):
count = Bugzilla(
queries={"bug_query_parameter": "1"},
bug_query_params={"count_only": 1}
).get_data().wait()
return count

I think now the API has been used correctly. How should I run the tests locally can you please guide me to some resources @suhaibmujahid ?

[suhaibmujahid]

I think now the API has been used correctly.

Unfortunately, it is not :( See that example in the issue. Also, you could explore other places where that class is used.

How should I run the tests locally can you please guide me to some resources?

You do not need to run all existing tests locally to test the function. You could write a simple script/test to run it. Anyway, you could run the existing tests locally by running the following command:

python -m pytest tests/test_*.py

However, any time you push a commit to the PR, the tests will automatically run, and you can see the results. For example, the following are the results of your last comment: https://community-tc.services.mozilla.com/tasks/groups/Yf7oAmmcQJ-b4f7MOWFLGg

The commit failed on the lining task, not the tests. You could prevent that by using pre-commit. See https://github.com/mozilla/bugbug/tree/master#auto-formatting