Skip to content

Release 0.5.0

Choose a tag to compare

@github-actions github-actions released this 01 Dec 03:29
· 207 commits to main since this release
355c4ff

Release 0.5.0

Installation

pip install baselinr==0.5.0

See the full changelog for details.

Release Notes for v0.5.0

New Features

  • Smart Table Selection: Automatically recommend tables for monitoring based on database usage patterns, query frequency, and metadata analysis
  • Usage-based table discovery: Analyze query history, table sizes, and write activity to identify high-value tables worth monitoring
  • New CLI command: baselinr recommend command to generate, review, and apply table recommendations
  • Lineage visualization: Export data lineage diagrams in multiple formats (ASCII, Mermaid, Graphviz DOT, SVG, PNG, PDF, JSON)
  • Column-level lineage: Track and visualize data flow at the column level, not just table level
  • Lineage sync command: Bulk sync lineage from query history providers with incremental updates
  • Lineage cleanup command: Automatically remove stale lineage edges based on expiration policies
  • Lineage providers command: List and check availability of lineage data sources (dbt, SQL parser, query history, Dagster)
  • Multi-database metadata collection: Support for PostgreSQL, Snowflake, BigQuery, Redshift, MySQL, and SQLite metadata collection
  • Configurable scoring algorithm: Customize recommendation weights for query frequency, recency, write activity, and table size
  • Recommendation modes: Choose between "recommend" (review before applying) or "auto" (apply high-confidence recommendations automatically)
  • Drift-aware lineage visualization: Highlight tables affected by data drift in lineage diagrams

Improvements

  • Reduced configuration overhead: Automatically discover and recommend tables instead of manual configuration
  • Transparent recommendations: Detailed explanations for why tables are recommended or excluded
  • Existing table filtering: Automatically skip tables already configured to avoid duplicates
  • Enhanced lineage querying: Improved upstream, downstream, and path queries with column-level support
  • Multiple export formats: Lineage can be exported for documentation (Mermaid), visualization tools (Cytoscape, D3), or static images

Maintenance

  • Enhanced PostgreSQL metadata collection: Improved query performance and timezone handling for metadata collection
  • Windows compatibility: Fixed Unicode encoding issues for CLI output on Windows terminals