SQLLineage

Supports multiple SQL dialects like ANSI, Hive, and SparkSQL, enabling accurate parsing of non-standard syntax, as shown in the example where SparkSQL correctly handles 'INSERT OVERWRITE' while others fail.

Traces column dependencies across queries for granular data flow analysis, evidenced by the detailed column lineage output that maps paths from source to target columns.

Enhances lineage accuracy by connecting to databases via SQLAlchemy, resolving ambiguities like wildcards and unqualified columns, as demonstrated with the sqlite example improving results.

Generates interactive DAG visualizations in a web browser, making complex lineage graphs easy to interpret visually without manual diagramming.