v0.14.7-alpha.2

• Relax version requirement for rich dependency by @elephantum in #367

v0.14.7-alpha.1

• Make CLI accept multiple --name values

v0.15.0-alpha.4

• Deprecate Python 3.9

v0.15.0-alpha.3

InputSpec and key mapping

See key-mapping.md for motivation

• Renamed JoinSpec to InputSpec
• Added keys parameter to InputSpec and ComputeInput to support joining tables with different key names
• Added DataField accessor for InputSpec.keys

v0.15.0-alpha.2

• Introduced MetaPlane/TableMeta/TransformMeta interfaces to decouple
  metadata management from the compute plane
• Added SQL reference implementation (SQLMetaPlane, SQLTableMeta,
  SQLTransformMeta) and rewired DataStore, DataTable, and batch transform
  steps to consume the new meta plane API
• Added meta-plane design doc and removed legacy MetaTable plumbing in lints,
  migrations, and tests
• Make CLI accept multiple --name values

v0.14.6

• Allow DataTable objects directly in ComputeStep.input_dts parameter (auto-wraps in ComputeInput)
• Add backward compatibility for DataTable API in BatchTransformStep
• Add optional force_vectors_to_ram parameter to QdrantStore
• Moved calculating checksum from MetaTable to TableStore
• Added PandasParquetFile adapter for TableStoreFiledir to store pandas DataFrames in parquet
• Added hash_row for BytesFile and PILFile
• Added use_adapter_hash parameter in TableStoreFiledir for backward compatibility
• Support for Python 3.13 and Python 3.14
• Support for Numpy 2.0
• Removed dependency on iteration-utilities

v0.14.5

• Fix for getting existing idx of empy meta-table
• Add keep_existing=True to TransformMetaTable to prevent table metadata conflicts (temporary fix)
• Enhance RedisStore with multi-node cluster support and password authentication

v0.14.4

• Relax SQLAlchemy requirement to allow 1.4.x support

v0.14.3

• New table store: datapipe.store.neo4j.Neo4JStore for graph databases

v0.14.2

New stuff:

• New table store: datapipe.store.elastic.ElasticStore

Internal changes:

• Convert dependency management to UV
• Fix datapipe.types.OrmTable type
• Add TableStore.caps to introspect what features does specific table store supports
• Implement abstract test suite for TableStore: datapipe.store.tests.abstract.AbstractBaseStoreTests
• Refactor test_table_store to use AbstractBaseStoreTests