GeoIP2 Behavior Change in Django 5.1

Based on the information provided, it appears that there has been a significant change in how Django 5.1 handles GeoIP2 database selection compared to Django 5.0.

Analysis of the Issue

The behavior change you're experiencing is consistent with what's mentioned in the Django 5.1 release notes. In Django 5.1, when a directory path is provided to GeoIP2(), it now prefers the city database over the country database if both are available. Since the city database is more comprehensive (containing country information as well), Django now treats the country database as a subset and doesn't automatically fall back to it.

This explains why your test is failing with AddressNotFoundError - it seems that the IP address 8.8.8.8 exists in the country database but not in the city database that's being prioritized in Django 5.1.

Recommendations

1. Is this behavior expected? Yes, this appears to be an intentional change in Django 5.1 based on the release notes you referenced.

2. What adjustments should be made to the code? You have a few options:

   a. Explicitly specify the country database when you need country-level data:

   PYTHONCopied
   class TestGeoIP(TestCase):
       def test_geoip(self):
           # Explicitly use the country database
           geoip = GeoIP2(country='path/to/GeoLite2-Country.mmdb')
           # Or if using GEOIP_PATH setting:
           # geoip = GeoIP2(country=True)
           result = geoip.country('8.8.8.8')
           self.assertEqual(result, {'country_code': 'US', 'country_name': 'United States'})

   b. Update your test data to use IP addresses that exist in both databases.

   c. Mock the GeoIP2 functionality in your tests to avoid dependency on actual database content.

Additional Considerations

• Make sure both your country and city databases are up-to-date, as MaxMind regularly updates their databases.
• If you're developing a library that needs to be compatible with both Django 5.0 and 5.1, you might need conditional logic based on the Django version.
• For testing purposes, you might want to consider using a fixed test dataset to ensure consistent results across different environments.

This change in behavior highlights the importance of checking release notes when upgrading Django versions, especially for features like GeoDjango that interact with external data sources.