cocoindex 0.2.13__cp311-abi3-macosx_11_0_arm64.whl → 0.2.15__cp311-abi3-macosx_11_0_arm64.whl

cocoindex/__init__.py

@@ -2,6 +2,7 @@
 Cocoindex is a framework for building and running indexing pipelines.
 """
 
+from . import _engine  # type: ignore
 from . import functions, sources, targets, cli, utils
 
 from . import targets as storages  # Deprecated: Use targets instead
@@ -21,7 +22,13 @@ from .flow import add_flow_def, remove_flow  # DEPRECATED
 from .flow import update_all_flows_async, setup_all_flows, drop_all_flows
 from .lib import settings, init, start_server, stop
 from .llm import LlmSpec, LlmApiType
-from .index import VectorSimilarityMetric, VectorIndexDef, IndexOptions
+from .index import (
+    VectorSimilarityMetric,
+    VectorIndexDef,
+    IndexOptions,
+    HnswVectorIndexMethod,
+    IvfFlatVectorIndexMethod,
+)
 from .setting import DatabaseConnectionSpec, Settings, ServerSettings
 from .setting import get_app_namespace
 from .query_handler import QueryHandlerResultFields, QueryInfo, QueryOutput
@@ -36,6 +43,8 @@ from .typing import (
     Json,
 )
 
+_engine.init_pyo3_runtime()
+
 __all__ = [
     # Submodules
     "_engine",
@@ -82,6 +91,8 @@ __all__ = [
     "VectorSimilarityMetric",
     "VectorIndexDef",
     "IndexOptions",
+    "HnswVectorIndexMethod",
+    "IvfFlatVectorIndexMethod",
     # Settings
     "DatabaseConnectionSpec",
     "Settings",

cocoindex/functions.py

@@ -2,13 +2,13 @@
 
 import dataclasses
 import functools
-from typing import Annotated, Any, Literal
+from typing import Any, Literal
 
 import numpy as np
 from numpy.typing import NDArray
 
 from . import llm, op
-from .typing import TypeAttr, Vector
+from .typing import Vector
 
 
 class ParseJson(op.FunctionSpec):
@@ -40,6 +40,24 @@ class SplitRecursively(op.FunctionSpec):
     custom_languages: list[CustomLanguageSpec] = dataclasses.field(default_factory=list)
 
 
+class SplitBySeparators(op.FunctionSpec):
+    """
+    Split text by specified regex separators only.
+    Output schema matches SplitRecursively for drop-in compatibility:
+        KTable rows with fields: location (Range), text (Str), start, end.
+    Args:
+        separators_regex: list[str]  # e.g., [r"\\n\\n+"]
+        keep_separator: Literal["NONE", "LEFT", "RIGHT"] = "NONE"
+        include_empty: bool = False
+        trim: bool = True
+    """
+
+    separators_regex: list[str] = dataclasses.field(default_factory=list)
+    keep_separator: Literal["NONE", "LEFT", "RIGHT"] = "NONE"
+    include_empty: bool = False
+    trim: bool = True
+
+
 class EmbedText(op.FunctionSpec):
     """Embed a text into a vector space."""
 

cocoindex/index.py

@@ -1,6 +1,6 @@
 from enum import Enum
 from dataclasses import dataclass
-from typing import Sequence
+from typing import Sequence, Union
 
 
 class VectorSimilarityMetric(Enum):
@@ -9,6 +9,26 @@ class VectorSimilarityMetric(Enum):
     INNER_PRODUCT = "InnerProduct"
 
 
+@dataclass
+class HnswVectorIndexMethod:
+    """HNSW vector index parameters."""
+
+    kind: str = "Hnsw"
+    m: int | None = None
+    ef_construction: int | None = None
+
+
+@dataclass
+class IvfFlatVectorIndexMethod:
+    """IVFFlat vector index parameters."""
+
+    kind: str = "IvfFlat"
+    lists: int | None = None
+
+
+VectorIndexMethod = Union[HnswVectorIndexMethod, IvfFlatVectorIndexMethod]
+
+
 @dataclass
 class VectorIndexDef:
     """
@@ -17,6 +37,7 @@ class VectorIndexDef:
 
     field_name: str
     metric: VectorSimilarityMetric
+    method: VectorIndexMethod | None = None
 
 
 @dataclass

cocoindex/targets/lancedb.py

@@ -296,6 +296,12 @@ class _Connector:
     ) -> _State:
         if len(key_fields_schema) != 1:
             raise ValueError("LanceDB only supports a single key field")
+        if index_options.vector_indexes is not None:
+            for vector_index in index_options.vector_indexes:
+                if vector_index.method is not None:
+                    raise ValueError(
+                        "Vector index method is not configurable for LanceDB yet"
+                    )
         return _State(
             key_field_schema=key_fields_schema[0],
             value_fields_schema=value_fields_schema,

{cocoindex-0.2.13.dist-info → cocoindex-0.2.15.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cocoindex
-Version: 0.2.13
+Version: 0.2.15
 Classifier: Development Status :: 3 - Alpha
 Classifier: License :: OSI Approved :: Apache Software License
 Classifier: Operating System :: OS Independent

{cocoindex-0.2.13.dist-info → cocoindex-0.2.15.dist-info}/RECORD

@@ -1,15 +1,15 @@
-cocoindex-0.2.13.dist-info/METADATA,sha256=08dUNFSA-_C_GyIUPTzNCV5ss-L-GP8r07ucSnlU7N4,13316
-cocoindex-0.2.13.dist-info/WHEEL,sha256=cVaoL47Ex1FxzwnkO_WCjy3a1Wl6mtZbBPTvTiNCHdY,103
-cocoindex-0.2.13.dist-info/entry_points.txt,sha256=_NretjYVzBdNTn7dK-zgwr7YfG2afz1u1uSE-5bZXF8,46
-cocoindex-0.2.13.dist-info/licenses/THIRD_PARTY_NOTICES.html,sha256=kOcN_-Ew74skSjMDUYvpRfdbB2pO24VPcbup3dwagb4,717769
-cocoindex/__init__.py,sha256=AsoNLBgjJ-1AEKtwTnENJM9dEilHTq6rJ4qWUzGKSdc,2428
-cocoindex/_engine.abi3.so,sha256=gKWLgibmJt1uyCgB9SKHVDF9NnjjjKiepnCGwml840U,67174768
+cocoindex-0.2.15.dist-info/METADATA,sha256=3u9PxzHp4NgjbknLyCtuASJeXskfIpODzlIpSyaBcSI,13316
+cocoindex-0.2.15.dist-info/WHEEL,sha256=cVaoL47Ex1FxzwnkO_WCjy3a1Wl6mtZbBPTvTiNCHdY,103
+cocoindex-0.2.15.dist-info/entry_points.txt,sha256=_NretjYVzBdNTn7dK-zgwr7YfG2afz1u1uSE-5bZXF8,46
+cocoindex-0.2.15.dist-info/licenses/THIRD_PARTY_NOTICES.html,sha256=LV2CtczxHQ5mN6Rj9TgnBa06dAxNPXZLXc0hMbco8tU,717769
+cocoindex/__init__.py,sha256=6qZWVkK4WZ01BIAg3CPh_bRRdA6Clk4d4Q6OnZ2jFa4,2630
+cocoindex/_engine.abi3.so,sha256=fCnSkunJcYsjrtesJqRupKG7o9neU6pQ25vZRHVOjUU,67288912
 cocoindex/auth_registry.py,sha256=_DOIY42C79joLCY_XczHwP5uebkmSavweoAHc0L3hQY,1334
 cocoindex/cli.py,sha256=69X30bFTFdM7c0_6lgIHR19CeQ7UEkobEQYihy8IdOQ,21599
 cocoindex/convert.py,sha256=itkUBCriOk8fdauahHRqJ-L8mnHehNZsBe_FouN0K1Q,28695
 cocoindex/flow.py,sha256=Vk72dX_svfpinvsolQ11aw6YDqbzaafrAi7xrQHo1i0,39844
-cocoindex/functions.py,sha256=09erNt3WbzY9l1KER-akBF2O5-6xEahV2ORBECaL6yk,12260
-cocoindex/index.py,sha256=j93B9jEvvLXHtpzKWL88SY6wCGEoPgpsQhEGHlyYGFg,540
+cocoindex/functions.py,sha256=V4ljBnCprvA25XlCVvNLwK5ergXiEcKU76jkOGC-X3A,12882
+cocoindex/index.py,sha256=tz5ilvmOp0BtroGehCQDqWK_pIX9m6ghkhcxsDVU8WE,982
 cocoindex/lib.py,sha256=0XheDF7fiFdqExpdqzU-VKun_Zll6DwZ5JfTm7u42aY,2284
 cocoindex/llm.py,sha256=Pv_cdnRngTLtuLU9AUmS8izIHhcKVnuBNolC33f9BDI,851
 cocoindex/op.py,sha256=c1xzoiWoPb6PYiCQAUsyzMRJwPvWaC3o5emPk38oY-4,26551
@@ -22,7 +22,7 @@ cocoindex/sources.py,sha256=FYz7cWYasLGDaYoIEQ1dF2uprgUETHWsTIrIS7n6pQE,3188
 cocoindex/subprocess_exec.py,sha256=r1xO84uek4VP4I6i87JMwsH5xFm3vKW0ABvgn0jskt4,10088
 cocoindex/targets/__init__.py,sha256=HQG7I4U0xQhHiYctiUvwEBLxT2727oHP3xwrqotjmhk,78
 cocoindex/targets/_engine_builtin_specs.py,sha256=DM7vyO0pkoukA-aBbvm_J4irgXhXIEqWdp-hwVpVRU4,2800
-cocoindex/targets/lancedb.py,sha256=pEHEBCOEe358e14cqAI1SytUTBCgRkYCsv9fJE7biTA,15440
+cocoindex/targets/lancedb.py,sha256=1nzCre5p-fvKkmLOTvfpiLTfnhF3qMLqTvsTwNuGwVU,15749
 cocoindex/tests/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 cocoindex/tests/test_convert.py,sha256=pG1AkEdIKSRE0trMV2dQ6VSQEqQJnPTJzfQCVrZSz8w,50791
 cocoindex/tests/test_load_convert.py,sha256=XHuXhgLcazniEYrhoz_L5RFvgUwu-iy3EB9zgR6p95A,3339
@@ -34,4 +34,4 @@ cocoindex/typing.py,sha256=jZO3meRVL_RsFdhj8Sx6gWF-Z207VhoPtb1ZmqzAnH0,19974
 cocoindex/user_app_loader.py,sha256=bc3Af-gYRxJ9GpObtpjegZY855oQBCv5FGkrkWV2yGY,1873
 cocoindex/utils.py,sha256=hUhX-XV6XGCtJSEIpBOuDv6VvqImwPlgBxztBTw7u0U,598
 cocoindex/validation.py,sha256=PZnJoby4sLbsmPv9fOjOQXuefjfZ7gmtsiTGU8SH-tc,3090
-cocoindex-0.2.13.dist-info/RECORD,,
+cocoindex-0.2.15.dist-info/RECORD,,

{cocoindex-0.2.13.dist-info → cocoindex-0.2.15.dist-info}/licenses/THIRD_PARTY_NOTICES.html

@@ -2428,7 +2428,7 @@ Software.
                 <h3 id="Apache-2.0">Apache License 2.0</h3>
                 <h4>Used by:</h4>
                 <ul class="license-used-by">
-                    <li><a href=" https://crates.io/crates/cocoindex ">cocoindex 0.2.13</a></li>
+                    <li><a href=" https://crates.io/crates/cocoindex ">cocoindex 0.2.15</a></li>
                     <li><a href=" https://github.com/awesomized/crc-fast-rust ">crc-fast 1.3.0</a></li>
                     <li><a href=" https://github.com/qdrant/rust-client ">qdrant-client 1.15.0</a></li>
                 </ul>