Fix: Add version constraints to schemaregistry dependencies

[aywengo]

Description

This PR fixes the "No module named 'attrs'" error that occurs when installing confluent-kafka with the schemaregistry extra.

Issue

Fixes #1972

Problem

When installing confluent-kafka[protobuf,schemaregistry]==2.10.0, the attrs and cachetools dependencies are not properly installed because they lack version constraints in the requirements file.

Solution

Added explicit version constraints to ensure these dependencies are properly resolved:

• attrs>=21.2.0
• cachetools>=5.5.0

These version constraints are based on the versions that were successfully resolved in the pip-compile output from the issue report.

Testing

To test this fix:

pip install git+https://github.com/aywengo/confluent-kafka-python.git@fix-attrs-dependency#egg=confluent-kafka[protobuf,schemaregistry]

Then verify:

from confluent_kafka.schema_registry import RegisteredSchema
# Should work without "ModuleNotFoundError: No module named 'attrs'"

[Copilot]

Pull Request Overview

A fix to ensure attrs and cachetools are properly installed by adding explicit version constraints to the schemaregistry extra, addressing the ModuleNotFoundError: No module named 'attrs'.

• Replace unbounded attrs and cachetools entries with minimum version requirements.
• Align dependency pins with versions resolved from pip-compile output.

[confluent-cla-assistant]

🎉 All Contributor License Agreements have been signed. Ready to merge.
✅ aywengo
_{Please push an empty commit if you would like to re-run the checks to verify CLA status for all contributors.}

[aywengo]

Testing Confirmation

I've tested this fix locally and it resolves the issue. Here's my testing process:

Environment

• Python 3.9.6
• pip 24.0
• OS: [Your OS]

Test Steps

1. Created a fresh virtual environment

2. Installed from the fix branch:

   pip install git+https://github.com/aywengo/confluent-kafka-python.git@fix-attrs-dependency#egg=confluent-kafka[protobuf,schemaregistry]

3. Verified the fix:

   # test_import.py
   from confluent_kafka.schema_registry import RegisteredSchema
   print("✅ Import successful - attrs module found")

Results

• ✅ Installation completed without errors
• ✅ All dependencies properly installed (attrs, cachetools, httpx, authlib)
• ✅ Import statement works without "ModuleNotFoundError"

This fix ensures that the required dependencies are properly installed when using the schemaregistry extra, preventing the import error reported in #1972.