Switch build_dbs.py to boto3, add Pfam-A.clans.tsv auto-download, replace MySQLdb with mysql-connector-python

[Copilot]

• Add boto3 and mysql-connector-python to requirements-dev.txt
• Update bin/build_dbs.py to use boto3 for S3 operations directly (no wrapper functions), credentials from environment
• Update bin/build_dbs.py to download Pfam-A.clans.tsv from S3 if it doesn't exist locally
• Update bin/build_dbs.py to use logging instead of print statements
• Update agfusion/database.py to use mysql-connector-python instead of MySQLdb, with error handling

Original prompt

Update the #file:build_dbs.py file with the following requirements:

• Have it use the boto3 aws library when download/uploading any files from s3. it can assume credentials come from environemnt variables. boto3 should only be a dev dependency in #file:requirements-dev.txt
• Have it download download the 'Pfam-A.clans.tsv' file from the s3 bucket if it does not exist locally
• Update #file:database.py to use mysql-connector-python instead. It should have that dependency installed in #file:requirements-dev.txt
  Create a pull request.

The user has attached the following file paths as relevant context:

• requirements-dev.txt
• agfusion/database.py
• bin/build_dbs.py
• .github/instructions/python.instructions.md
• .github/copilot-instructions.md

Created from VS Code.

[murphycj]

@copilot get rid of the wrapper functions around the boto download and upload functions. Just use boto directly. Also, use the logging library instead of print statements for logging

[Copilot]

@copilot get rid of the wrapper functions around the boto download and upload functions. Just use boto directly. Also, use the logging library instead of print statements for logging

Done in ebbee3e. Removed the s3_key_exists, s3_upload, and s3_download wrapper functions — boto3 is now called directly inline. Replaced all print statements with logger.info calls using the standard logging module.