v0.1.1

What's Changed

• Expanded tags and packets to accept any Python data type and use Arrow-based datastore by @eywalker in #22
• Rename orcabridge to orcapod by @eywalker in #25
• Add working pipeline system implementation by @eywalker in #27
• Complete refactor of streams and datagram logic and use of protocols throughout by @eywalker in #34
• Prepare for 0.1.0alpha release by @eywalker in #55
• Enhance function_pod decorator to attach FunctionPod as callable attribute by @eywalker in #57
• Add execution engine options to streams and nodes by @eywalker in #60
• Fix bug to pass resource parameters to ray task options() by @brian-arnold in #62
• Enhance pipeline functionality and access to pods by @eywalker in #63
• feat(core): add packet function executor system and async channel execution by @eywalker in #67
• Integrate TODO audit findings into DESIGN_ISSUES.md with priorities by @eywalker in #68
• Add concrete asynchronous orchestrator by @eywalker in #71
• Add native async streaming overrides for core operators by @eywalker in #70
• Add async function support and fix concurrent packet execution by @eywalker in #72
• Add comprehensive specification-derived test suite by @eywalker in #73
• Fix hash_file return type and wire file_hasher through converters by @eywalker in #74
• Claude/fix utc tzdata timestamp [ENG-225] by @brian-arnold in #75
• Added GenericAliasHandler class [ENG-226] by @brian-arnold in #76
• Integrate Ray execution engine and enable passing of options [ENG-231] by @brian-arnold in #77
• Refactor source caching out of pipeline and add support for general caching for all sources by @eywalker in #80
• feat: pipeline serialization and node simplification by @eywalker in #81
• feat: add documentation site and clean up public API namespace by @eywalker in #82
• refactor: major cleanup and simplification of protocols by @eywalker in #83
• refactor(executors): improve function execution chain by @eywalker in #84
• feat: add sync pipeline orchestrator with per-packet observability by @eywalker in #85
• refactor: slim node protocols, observer injection, async orchestrator alignment (PLT-922) by @eywalker in #86
• feat: add PacketFunctionProxy for read-only pipeline loading (PLT-931) by @eywalker in #87
• Add infrastructure to capture and store logs: stdout, stderr, python logs, and tracebacks [ENG-264] by @brian-arnold in #89
• feat(hashing): integrate starfix-python for Arrow schema and data hashing by @kurodo3[bot] in #90
• feat: DBConnectorProtocol + ConnectorArrowDatabase + DBTableSource (PLT-1078) by @kurodo3[bot] in #91
• feat(status): add StatusOberver [ENG-288] by @brian-arnold in #92
• feat(observability): thread pipeline_path and pipeline_snapshot_hash through on_run_start by @kurodo3[bot] in #93
• Add Upath support [ENG-235] by @brian-arnold in #94
• Union type by @brian-arnold in #95
• fix(ray): pass runtime_env to ray.init() instead of ray.remote() [ENG-294] by @brian-arnold in #97
• feat(databases): implement SQLiteConnector — DBConnectorProtocol for SQLite (PLT-1076) by @kurodo3[bot] in #98
• fix(function_node): restore within-node concurrent packet execution by @brian-arnold in #99
• docs(logging_observer): fix and strengthen create_packet_logger Warning by @kurodo3[bot] in #102
• feat(join): unify async_execute into N-way streaming MJoin [ENG-301] by @brian-arnold in #101
• Update test docs for join operator by @brian-arnold in #103
• fix(PLT-1156): add CACHE_ONLY mode to FunctionNode for unavailable upstream sources by @kurodo3[bot] in #104
• fix(PLT-1158): uncached operator with UNAVAILABLE parents resolves to UNAVAILABLE by @kurodo3[bot] in #105
• docs(spiraldb): add pyspiral optional dep and SpiralDB exploration findings by @kurodo3[bot] in #106
• feat(spiraldb): implement SpiralDBConnector (PLT-1074) by @kurodo3[bot] in #109
• feat(sources): implement SQLiteTableSource with ROWID fallback (PLT-1077) by @kurodo3[bot] in #110
• feat(sources): SpiralDBTableSource with PK as default tag columns by @kurodo3[bot] in #111
• feat(sources): PostgreSQLTableSource + executor/channel fixes (PLT-1072) by @kurodo3[bot] in #113
• Restore Pipeline.show_graph, fix graph rendering, add tests by @eywalker in #114
• fix(operator_node): enforce non-active node semantics (PLT-1182) by @kurodo3[bot] in #115
• fix(types): Schema.repr shows clean type names instead of <class '...'> wrappers by @kurodo3[bot] in #116
• fix: narrow Stream.flow() return type from Collection to Sequence/list by @kurodo3[bot] in #118
• fix(ray): restore ray client functionality [ENG-339] by @brian-arnold in #119
• perf: reduce import orcapod startup time from >1200ms to ~150ms (PLT-1178) by @kurodo3[bot] in #117
• Observability reader [ENG-347] by @brian-arnold in #121
• feat(ENG-258): add MinIO/S3 and UPath support for DeltaTableDatabase by @kurodo3[bot] in #122
• feat(pipeline): normalized save format with DB deduplication and save levels (ENG-256/ENG-320/ENG-342) by @kurodo3[bot] in #120
• feat(databases): add at() and base_path for sub-scoped database views (ENG-341) by @kurodo3[bot] in #123
• feat(pipeline): add auto_save_path parameter to Pipeline for automatic pipeline saving by @brian-arnold in #124
• refactor: decouple pipeline_path from nodes + observer contextualization (ENG-340 / ENG-349) by @kurodo3[bot] in #125
• PLT-923: Schema ↔ Arrow nullability correspondence + logical equality tests by @kurodo3[bot] in #126
• fix(ray): propagate original function name to Ray metrics instead of generic _capture [ENG-358] by @brian-arnold in #128
• feat(ENG-373): add configurable table_scope to FunctionNode and OperatorNode by @kurodo3[bot] in #129
• fix: restore Arrow nullable flags lost in Polars round-trips (ENG-375) by @kurodo3[bot] in #130
• refactor(ENG-379): make iter_packets() read-only, consolidate DB join sites by @kurodo3[bot] in #131
• feat(ENG-374): schema-driven execution context with datagram-based storage by @eywalker in #132
• ci(PLT-1250): tag-triggered PyPI release pipeline by @kurodo3[bot] in #133
• chore: merge dev → main with repo cleanup (PLT-1253) by @kurodo3[bot] in #134
• Rename packet -> data by @eywalker in #137
• feat(pipeline): refactor Pipeline into source-agnostic DAG; introduce PipelineJob and SourceSpec by @eywalker in #138
• feat(sources): protocol-based polling source for async pipelines (PLT-1430) by @kurodo3[bot] in #139
• refactor(pipeline): Pipeline pure-descriptor refactor — ENG-493 by @kurodo3[bot] in #141
• spike(pipeline): ENG-492 OrcaDAG prototype and networkx replacement design by @kurodo3[bot] in #140
• refactor(config): rename Config → OrcapodConfig by @kurodo3[bot] in #142
• refactor(pipeline): unify recording/compile via InvocationGraph, dedupe Pipeline/PipelineJob by @kurodo3[bot] in #143
• feat(pipeline): DAG public API — dag property, OrcaDAG[NodeT], PipelineProtocol (+ copy-on-run cleanup) by @kurodo3[bot] in #146
• refactor(pipeline): address eywalker review comments (OrcaDAG[NodeT], remove JIT imports) by @kurodo3[bot] in #147
• refactor(function_node): extract _fetch_joined_records to eliminate join duplication by @kurodo3[bot] in #148
• feat(pipeline): migrate all skipped pipeline tests to PipelineJob API (ENG-491) by @kurodo3[bot] in #149
• feat(dataclass): add dataclass ↔ Arrow struct encoding with __type sentinel (ENG-555) by @kurodo3[bot] in #150
• fix(source_node): forward column_config/all_info in keys() and output_schema() by @kurodo3[bot] in #152
• fix(function_node): fix empty as_table() schema + cleanup (ENG-572) by @kurodo3[bot] in #151
• docs: remove ReST cross-reference markup from docstrings by @kurodo3[bot] in #153
• refactor(eng-376): rename CachedFunctionPod._result_database to result_database by @kurodo3[bot] in #154
• fix(type-inference): map Any to pa.null() for empty container handling (ENG-389) by @kurodo3[bot] in #155
• feat(sources): normalize tag_columns bare strings in all user-facing sources (ENG-571) by @kurodo3[bot] in #157
• feat(config): add persistent OrcaPod configuration system (PLT-964) by @kurodo3[bot] in #156
• feat(pipeline): add source_pods, function_pods, operator_pods properties by @kurodo3[bot] in #158
• refactor(function_pod): delegate _FunctionPodBase.uri to data_function.uri by @kurodo3[bot] in #159
• fix(databases): normalize string_view columns to avoid PyArrow kernel crash [ENG-601] by @brian-arnold in #160
• feat(types): UUID Arrow type migration — pa.binary(16) for all UUID storage (PLT-1162) by @kurodo3[bot] in #161
• feat(semantic-types): add native datetime ↔ Arrow timestamp support by @kurodo3[bot] in #163
• fix(hashing): remove all hash truncations from orcapod-python by @kurodo3[bot] in #164
• feat(extension_types): add ExtensionTypeConverter protocol (PLT-1652) by @kurodo3[bot] in #166
• feat(hashing): schema-metadata cleaner + starfix v0.3.0 adoption by @kurodo3[bot] in #179
• fix(spiraldb): upgrade pyspiral 0.11.7 → 0.14.9 to fix CI header-signing failure (PLT-1773) by @kurodo3[bot] in #180
• feat(extension-types): Arrow/Polars extension type semantic type system (PLT-1663) by @kurodo3[bot] in #183
• fix(hashing): pydantic/dataclass columns as pipeline columns (ITL-432) by @kurodo3[bot] in #185
• feat(operators): add Pick and Index structural projection operators (ITL-140) by @kurodo3[bot] in #186
• feat(types): add datetime.date and datetime.datetime support for Tag values (ITL-45) by @kurodo3[bot] in #188
• fix(types): support typing.Literal fields in pydantic pipeline columns (ITL-442) by @kurodo3[bot] in #189
• ci: set up Linear release system for v0.1 release cut (ITL-448) by @kurodo3[bot] in #191
• fix(types): accept union-typed function inputs (ITL-452) by @kurodo3[bot] in #192
• ci: add workflow_dispatch release workflow (ITL-449) by @kurodo3[bot] in #194
• feat(types): add orcapod.File content-identified type, refactor Path/UPath to pure-path hashing by @kurodo3[bot] in #193
• fix(hashing): make union-typed signature hash order-independent (ITL-453) by @kurodo3[bot] in #195
• feat(types): add orcapod.Directory with Merkle tree hashing + LogicalFile JSON migration (ITL-451) by @kurodo3[bot] in #196

New Contributors

• @brian-arnold made their first contribution in #62

Full Changelog: v0.0.2...v0.1.1