Skip to content

CHANGELOG.md

Latest commit

 

History

History
97 lines (81 loc) · 5.2 KB

CHANGELOG.md

File metadata and controls

97 lines (81 loc) · 5.2 KB

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

Unreleased

0.15.0 - 2026-02-22

Added

  • All connectors enabled: Kafka, WebSocket, PostgreSQL CDC, PostgreSQL sink, MySQL CDC, Delta Lake, Iceberg, Parquet Lookup
  • DuckDB-style API on QueryResult:
    • df(), pl(lazy=False), arrow() — DataFrame conversion aliases
    • fetchall(), fetchone(), fetchmany(size) — row-oriented fetch methods
    • show(max_rows=20) — terminal preview
    • _repr_html_() — Jupyter notebook HTML rendering
    • __len__(), __iter__() — len(result) and for row in result
  • DuckDB-style API on Connection:
    • sql(query) — alias for query()
    • tables() — alias for list_tables()
    • materialized_views() — alias for list_streams()
    • explain(query) — query execution plan
    • stats(table) — table statistics as dict
  • New exceptions: StreamError, CheckpointError, ConnectorError (all subclass LaminarError)
  • Python wrapper types (laminardb.types):
    • Schema, Column, TableStats, Watermark, CheckpointStatus, Metrics
    • ChangeRow, ChangeEvent — structured change data capture
    • MaterializedView — high-level stream wrapper with query(), subscribe(), schema()
  • Module-level functions: laminardb.sql(), laminardb.execute(), laminardb.mv() with thread-safe default connection
  • Aliases: Config = LaminarConfig, BatchWriter = Writer
  • Benchmarks: ingestion, query, and streaming throughput benchmarks
  • Examples: quickstart.py (DuckDB-style API), streaming_analytics.py (MaterializedView pattern)

Fixed

  • CI workflows now clone the laminardb monorepo for local path dependencies
  • CI test matrix updated to Python 3.11+ (matches abi3-py311 build target)
  • maturin develop in CI now runs inside a virtualenv
  • schema() and query() now work correctly for streams

Changed

  • Minimum Python version is now 3.11 (was 3.10) to match the abi3-py311 stable ABI target

0.1.0 - 2026-02-08

Added

  • Connection management: open(path) and connect(uri) with context manager support
  • Configuration: LaminarConfig with buffer_size, storage_dir, checkpoint_interval_ms
  • Schema operations: create_table(), list_tables(), list_streams(), list_sinks(), schema(), execute()
  • Data ingestion from 10 input types:
    • Python dicts (single row, list of rows, columnar)
    • pandas DataFrame, polars DataFrame
    • pyarrow RecordBatch, pyarrow Table
    • JSON string (insert_json()), CSV string (insert_csv())
    • Any object implementing __arrow_c_stream__ (zero-copy)
  • Streaming writer with context manager, insert(), flush(), close(), watermark(), current_watermark
  • SQL queries with query() and streaming stream() iterator
  • ExecuteResult: rich DDL/DML introspection with result_type, ddl_type, ddl_object, rows_affected, int(), bool() coercion
  • Result export to 6 output formats:
    • to_arrow() → pyarrow.Table
    • to_pandas() → pandas.DataFrame
    • to_polars() → polars.DataFrame
    • to_dicts() → columnar dict
    • to_df() → auto-detect best library
    • __arrow_c_stream__ → zero-copy Arrow PyCapsule
  • Subscriptions: sync (subscribe()) and async (subscribe_async()) continuous queries
    • Iterator/async iterator protocols
    • try_next() non-blocking poll (sync)
    • is_active property, cancel() method
  • Named stream subscriptions: subscribe_stream() / subscribe_stream_async() with schema access, next_timeout(ms), try_next()
  • Catalog introspection: sources(), sinks(), streams(), queries() returning typed info objects
  • Pipeline control: start(), shutdown(), checkpoint(), cancel_query()
  • Pipeline metrics: metrics(), topology(), source_metrics(), stream_metrics(), per-source/stream observability
  • Connection properties: is_closed, pipeline_state, pipeline_watermark, total_events_processed, source_count, sink_count, active_query_count, is_checkpoint_enabled
  • Async context manager: async with conn as db: ...
  • Exception hierarchy: LaminarError base with ConnectionError, QueryError, IngestionError, SchemaError, SubscriptionError
  • Error codes: numeric .code attribute on all exceptions, laminardb.codes module with 44 constants
  • PEP 561 type stubs with py.typed marker (800+ lines, passes mypy --strict)
  • GIL release on all blocking operations via py.allow_threads()
  • Free-threaded Python support (all pyclasses are Send + Sync)
  • CI/CD pipeline: lint, test (Python 3.11-3.13), type check, multi-platform wheel builds
  • PyPI trusted publishing via OIDC

Platform Support

  • Linux x86_64, aarch64
  • macOS x86_64, aarch64 (Apple Silicon)
  • Windows x86_64