v1.2.0

• Add support for dbt with materialized views #25
• Add support for pg_rewrite table
• Allow renaming materialized views

v1.1.0

• Add support for materialized views
• Assign default column names for json_column->'value' or json_column->>'value' with AS json_column_value
• Fix ISO datetime parsing with microseconds without timezone

v1.0.1

• Compact Iceberg tables with multiple Parquet files per manifest file
• Improve Parquet file statistics collection
• Handle unexpected Amplitude 404 errors
• Convert timestamptz to timestamp in UTC data type for DuckDB (DuckDB C API limitation)

v1.0.0

• Add support for jsonb_agg() and jsonb_object_agg()
• Fix column->>'key' and column->'key' operator precedence
• Get rid of Trino as a dependency added for Iceberg compaction
• Upgrade pg_query_go to v6 #52

v1.0.0-beta.2

• Speed up full-refresh syncing up to 50x times
• Remove Trino as a dependency when full-refresh syncing
• Improve memory usage to avoid OOM

v1.0.0-beta.1

• Re-architect the code to support data syncing from different sources
• Add support for Amplitude data syncing
• Use Trino for writing and compacting data

v0.51.1

• Reduce the data file chunk size to have smaller Parquet files

v0.51.0

• Add bemidb_last_synced_at('schema.table') for checking data freshness

v0.50.3

• Fix compatibility with old BemiDB versions w/o internal metadata

v0.50.2

• Upgrade golang.org/x/crypto to fix CVE-2025-22869 (not affected)