cocoindex 0.2.14__cp311-abi3-manylinux_2_28_x86_64.whl → 0.2.16__cp311-abi3-manylinux_2_28_x86_64.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
@@ -42,6 +43,8 @@ from .typing import (
     Json,
 )
 
+_engine.init_pyo3_runtime()
+
 __all__ = [
     # Submodules
     "_engine",

cocoindex/flow.py

@@ -17,7 +17,6 @@ from typing import (
     Callable,
     Generic,
     Iterable,
-    NamedTuple,
     Sequence,
     TypeVar,
     cast,
@@ -575,7 +574,8 @@ class FlowLiveUpdaterOptions:
     print_stats: bool = False
 
 
-class FlowUpdaterStatusUpdates(NamedTuple):
+@dataclass
+class FlowUpdaterStatusUpdates:
     """
     Status updates for a flow updater.
     """
@@ -1060,12 +1060,14 @@ def _get_data_slice_annotation_type(
 _transform_flow_name_builder = _NameBuilder()
 
 
-class TransformFlowInfo(NamedTuple):
+@dataclass
+class TransformFlowInfo(Generic[T]):
     engine_flow: _engine.TransientFlow
     result_decoder: Callable[[Any], T]
 
 
-class FlowArgInfo(NamedTuple):
+@dataclass
+class FlowArgInfo:
     name: str
     type_hint: Any
     encoder: Callable[[Any], Any]
@@ -1081,7 +1083,7 @@ class TransformFlow(Generic[T]):
     _args_info: list[FlowArgInfo]
 
     _lazy_lock: asyncio.Lock
-    _lazy_flow_info: TransformFlowInfo | None = None
+    _lazy_flow_info: TransformFlowInfo[T] | None = None
 
     def __init__(
         self,
@@ -1123,12 +1125,12 @@ class TransformFlow(Generic[T]):
         return self._flow_fn(*args, **kwargs)
 
     @property
-    def _flow_info(self) -> TransformFlowInfo:
+    def _flow_info(self) -> TransformFlowInfo[T]:
         if self._lazy_flow_info is not None:
             return self._lazy_flow_info
         return execution_context.run(self._flow_info_async())
 
-    async def _flow_info_async(self) -> TransformFlowInfo:
+    async def _flow_info_async(self) -> TransformFlowInfo[T]:
         if self._lazy_flow_info is not None:
             return self._lazy_flow_info
         async with self._lazy_lock:
@@ -1136,7 +1138,7 @@ class TransformFlow(Generic[T]):
                 self._lazy_flow_info = await self._build_flow_info_async()
             return self._lazy_flow_info
 
-    async def _build_flow_info_async(self) -> TransformFlowInfo:
+    async def _build_flow_info_async(self) -> TransformFlowInfo[T]:
         flow_builder_state = _FlowBuilderState(self._flow_name)
         kwargs: dict[str, DataSlice[T]] = {}
         for arg_info in self._args_info:

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-0.2.14.dist-info → cocoindex-0.2.16.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cocoindex
-Version: 0.2.14
+Version: 0.2.16
 Classifier: Development Status :: 3 - Alpha
 Classifier: License :: OSI Approved :: Apache Software License
 Classifier: Operating System :: OS Independent

{cocoindex-0.2.14.dist-info → cocoindex-0.2.16.dist-info}/RECORD

@@ -1,14 +1,14 @@
-cocoindex-0.2.14.dist-info/METADATA,sha256=rohqn4_9s8I4l6dxYLSTqdphN-fogyzjZNFU65w_ZiE,13316
-cocoindex-0.2.14.dist-info/WHEEL,sha256=IG-K_sumA04dNIpy5J1b3kZo_HELEjvxxDWHD32zTgo,107
-cocoindex-0.2.14.dist-info/entry_points.txt,sha256=_NretjYVzBdNTn7dK-zgwr7YfG2afz1u1uSE-5bZXF8,46
-cocoindex-0.2.14.dist-info/licenses/THIRD_PARTY_NOTICES.html,sha256=3Rq0MMyXp4exsfuop4rrPxu53e48XPYvzTfvV8EBY-Y,717769
-cocoindex/__init__.py,sha256=DP7l1Z05oxOiBl-PmLhqs3Lh2HcGg589DrSv4KCEvio,2563
-cocoindex/_engine.abi3.so,sha256=8_CUvr9MtBRWyow-Zp0n0e2OUUEf3W9DMIEV0c98HPQ,73509920
+cocoindex-0.2.16.dist-info/METADATA,sha256=PlewUqWQWLCwQG8HT1lJed7JG9nyvPH61KjhNB_3U7Q,13316
+cocoindex-0.2.16.dist-info/WHEEL,sha256=IG-K_sumA04dNIpy5J1b3kZo_HELEjvxxDWHD32zTgo,107
+cocoindex-0.2.16.dist-info/entry_points.txt,sha256=_NretjYVzBdNTn7dK-zgwr7YfG2afz1u1uSE-5bZXF8,46
+cocoindex-0.2.16.dist-info/licenses/THIRD_PARTY_NOTICES.html,sha256=fxZXw4-pEtDeJix0DJuH2VrSsp0CTjsExMxjxU4p0ds,717769
+cocoindex/__init__.py,sha256=6qZWVkK4WZ01BIAg3CPh_bRRdA6Clk4d4Q6OnZ2jFa4,2630
+cocoindex/_engine.abi3.so,sha256=xnhUzk7vP6YNFdJkkG-jleX9YqXv5KkKIDrExl4h0mk,73561928
 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/flow.py,sha256=oAFerPoarOS9XWVumYIzT4EHJyV3Pixv2mfqpCLHNOw,39849
+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
@@ -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-0.2.14.dist-info/RECORD,,
+cocoindex-0.2.16.dist-info/RECORD,,

{cocoindex-0.2.14.dist-info → cocoindex-0.2.16.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 0.2.14</a></li>
+                    <li><a href=" https://crates.io/crates/cocoindex ">cocoindex 0.2.16</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>