Skip to content

Latest commit

 

History

History
51 lines (31 loc) · 1.51 KB

File metadata and controls

51 lines (31 loc) · 1.51 KB

BigQuery DataFrames (bigframes) release procedure

(Note: bigframes releases are marked with skip_release: true in librarian.yaml and must be kicked off manually using legacylibrarian.)

Setup (First Time Only)

  • Install legacylibrarian:

    go install github.com/googleapis/librarian/cmd/legacylibrarian@latest
    
  • Authenticate with GitHub CLI:

    gh auth login
    

Release Steps

  • Obtain GitHub token:

    export LIBRARIAN_GITHUB_TOKEN=$(gh auth token)
    
  • Stash changes (repo must be clean):

    git stash -u
    
  • Fetch and checkout base:

    git fetch origin main
    git fetch origin --tags
    git checkout origin/main
    
  • Check image updates:

    legacylibrarian update-image --push
    
  • Create release PR:

    # Option A: Push directly
    legacylibrarian release stage --repo=https://github.com/googleapis/google-cloud-python --library=bigframes --library-version=X.X.X --push
    
    # Option B: Manual edit first (omit --push, edit files in /tmp/librarian-*, commit/push from there)
    legacylibrarian release stage --repo=https://github.com/googleapis/google-cloud-python --library=bigframes --library-version=X.X.X
    # In /tmp repository:
    git commit -a -m "chore: create release" --no-verify  # keep librarian config pristine
    git push origin HEAD
    gh pr create --fill --label "release:pending"
    
  • Post-release restore:

    # Move back any stashed/relocated files (like .vscode)
    git checkout main
    git stash pop