[mugenkyou]

Summary

Implements QuerySet.select_related() for single-level relationships.

Previously, select_related() returned a cloned QuerySet, but the requested relationship was not applied when the query was executed. This adds the query planning and row reconstruction needed to eagerly load related objects.

Changes

• Preserve select_related() state when cloning and chaining a QuerySet
• Add extra_alias support to the Python plan builder
• Generate qualified columns and LEFT OUTER JOINs for related models
• Alias related columns using the relationship name
• Reconstruct related model instances from joined rows
• Handle nullable foreign-key relationships
• Keep count() and exists() independent of select_related()
• Add integration tests for cloning, SQL generation, eager loading, nullable relationships, and count()/exists()
• Update the select_related() documentation

Scope

This implementation is limited to single-level relationships. Nested lookups such as author__company are intentionally not included.

Testing

• pytest ryx-python/tests/integration/test_select_related.py
• cargo check

Both pass successfully.

Closes #64

[Wgoeh]

Great job 👍🏾 thanks.
Everything looks cleaner.

[Einswilli]

Thank you for your contribution.