Removed reset_log_levels fixture

[jamesbraza]

#311 added some log parsing as a way of checking call ordering, and used a complicated fixture reset_log_levels to do so.

This code was confusing me a bit, so I just went ahead and cleaned it up to better use pytest's built-in caplog, with tighter assertions.

[Copilot]

Pull Request Overview

This PR removes a custom logging fixture in favor of pytest’s built-in caplog, scopes log-level settings to the specific paperqa.clients logger, and tightens assertions around log record ordering.

• Deleted the reset_log_levels fixture and unused logging import.
• Updated tests to call caplog.set_level(logging.DEBUG, logger="paperqa.clients") instead of a global reset.
• Changed record_indices to track multiple occurrences and added clearer assertions.

Reviewed Changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
tests/test_clients.py	Removed reset_log_levels usage, scoped caplog settings, and updated index-tracking logic for sequential log checks.
tests/conftest.py	Removed the reset_log_levels fixture and the now-unused import logging.

Comments suppressed due to low confidence (1)

tests/test_clients.py:563

• The assertion message is incorrect for semantic_scholar; it should read something like "Semantic Scholar should run" to match the key being checked.

        assert record_indices["semantic_scholar"], "Crossref should run"