feat: add normalized CSV output option for runner minutes report

Ricardo Amarillarequested to merge
improve-collection-and-solve-bug into main
  • Add --normalized flag to generate two separate CSV files instead of denormalized single file
  • Implement flatten_normalized() function to separate group and project data
  • Generate *_groups.csv with unique group-month records (no duplication)
  • Generate *_projects.csv with detailed project-level usage data
  • Update documentation with normalized format examples and benefits
  • Maintain backward compatibility with existing denormalized format as default
  • Add comprehensive test coverage for both output formats
  • Improve data integrity and analysis capabilities for large datasets

Benefits:

  • Eliminates data duplication in group totals
  • Enables easier database imports with proper relationships
  • Facilitates independent analysis of group vs project metrics
  • Maintains all existing functionality and performance optimizations

Example usage: python3 shared_runner_minutes_report.py $TOKEN --normalized python3 shared_runner_minutes_report.py $TOKEN --normalized --parallel --workers 4

Tested successfully on:

  • GitLab.com (gl-demo-ultimate-ramarilla group)
  • Self-managed instance (cs.gitlabdemo.cloud with 648 groups

Merge request reports