openaivec 0.99.2__tar.gz → 1.0.0__tar.gz

{openaivec-0.99.2 → openaivec-1.0.0}/.gitignore

@@ -13,6 +13,9 @@ CLAUDE.md
 ### Jupyter
 *.ipynb
 
+### deps
+*.dot
+
 
 ### Python template
 # Byte-compiled / optimized / DLL files

{openaivec-0.99.2 → openaivec-1.0.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: openaivec
-Version: 0.99.2
+Version: 1.0.0
 Summary: Generative mutation for tabular calculation
 Project-URL: Homepage, https://microsoft.github.io/openaivec/
 Project-URL: Repository, https://github.com/microsoft/openaivec
@@ -49,7 +49,7 @@ Description-Content-Type: text/markdown
 - Drop-in `.ai` and `.aio` DataFrame accessors keep pandas analysts in their favorite tools.
 - Smart batching (`BatchingMapProxy`) deduplicates prompts, enforces ordered outputs, and shortens runtimes without manual tuning.
 - Built-in caches, retry logic, and reasoning model safeguards cut noisy boilerplate from production pipelines.
-- Ready-made Spark UDF builders and Microsoft Fabric guides take AI workloads from notebooks into enterprise-scale ETL.
+- Ready-made Spark UDF helpers and Microsoft Fabric guides take AI workloads from notebooks into enterprise-scale ETL.
 - Pre-configured task library and `FewShotPromptBuilder` ship curated prompts and structured outputs validated by Pydantic.
 - Supports OpenAI and Azure OpenAI clients interchangeably, including async workloads and embeddings.
 
@@ -142,7 +142,7 @@ automatically in notebook environments when `show_progress=True`.
 - Vectorized request batching with automatic deduplication, retries, and cache hooks for any OpenAI-compatible client.
 - pandas `.ai` and `.aio` accessors for synchronous and asynchronous DataFrame pipelines, including `ai.extract` helpers.
 - Task library with Pydantic-backed schemas for consistent structured outputs across pandas and Spark jobs.
-- Spark UDF builders (`responses_udf`, `embeddings_udf`, `parse_udf`, `task_udf`, etc.) for large-scale ETL and BI.
+- Spark UDF helpers (`responses_udf`, `embeddings_udf`, `parse_udf`, `task_udf`, etc.) for large-scale ETL and BI.
 - Embeddings, token counting, and similarity utilities for search and retrieval use cases.
 - Prompt tooling (`FewShotPromptBuilder`, `improve`) to craft and iterate production-ready instructions.
 

{openaivec-0.99.2 → openaivec-1.0.0}/README.md

@@ -23,7 +23,7 @@
 - Drop-in `.ai` and `.aio` DataFrame accessors keep pandas analysts in their favorite tools.
 - Smart batching (`BatchingMapProxy`) deduplicates prompts, enforces ordered outputs, and shortens runtimes without manual tuning.
 - Built-in caches, retry logic, and reasoning model safeguards cut noisy boilerplate from production pipelines.
-- Ready-made Spark UDF builders and Microsoft Fabric guides take AI workloads from notebooks into enterprise-scale ETL.
+- Ready-made Spark UDF helpers and Microsoft Fabric guides take AI workloads from notebooks into enterprise-scale ETL.
 - Pre-configured task library and `FewShotPromptBuilder` ship curated prompts and structured outputs validated by Pydantic.
 - Supports OpenAI and Azure OpenAI clients interchangeably, including async workloads and embeddings.
 
@@ -116,7 +116,7 @@ automatically in notebook environments when `show_progress=True`.
 - Vectorized request batching with automatic deduplication, retries, and cache hooks for any OpenAI-compatible client.
 - pandas `.ai` and `.aio` accessors for synchronous and asynchronous DataFrame pipelines, including `ai.extract` helpers.
 - Task library with Pydantic-backed schemas for consistent structured outputs across pandas and Spark jobs.
-- Spark UDF builders (`responses_udf`, `embeddings_udf`, `parse_udf`, `task_udf`, etc.) for large-scale ETL and BI.
+- Spark UDF helpers (`responses_udf`, `embeddings_udf`, `parse_udf`, `task_udf`, etc.) for large-scale ETL and BI.
 - Embeddings, token counting, and similarity utilities for search and retrieval use cases.
 - Prompt tooling (`FewShotPromptBuilder`, `improve`) to craft and iterate production-ready instructions.
 

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/__init__.py

@@ -2,7 +2,7 @@ from ._embeddings import AsyncBatchEmbeddings, BatchEmbeddings
 from ._model import PreparedTask
 from ._prompt import FewShotPrompt, FewShotPromptBuilder
 from ._responses import AsyncBatchResponses, BatchResponses
-from ._schema import InferredSchema, SchemaInferenceInput, SchemaInferer
+from ._schema import SchemaInferenceInput, SchemaInferenceOutput, SchemaInferer
 
 __all__ = [
     "AsyncBatchEmbeddings",
@@ -11,7 +11,7 @@ __all__ = [
     "BatchResponses",
     "FewShotPrompt",
     "FewShotPromptBuilder",
-    "InferredSchema",
+    "SchemaInferenceOutput",
     "PreparedTask",
     "SchemaInferenceInput",
     "SchemaInferer",

openaivec-1.0.0/src/openaivec/_cache/__init__.py

@@ -0,0 +1,12 @@
+"""Caching utilities used across OpenAIVec."""
+
+from .optimize import BatchSizeSuggester, PerformanceMetric
+from .proxy import AsyncBatchingMapProxy, BatchingMapProxy, ProxyBase
+
+__all__ = [
+    "AsyncBatchingMapProxy",
+    "BatchSizeSuggester",
+    "BatchingMapProxy",
+    "PerformanceMetric",
+    "ProxyBase",
+]

openaivec-0.99.2/src/openaivec/_proxy.py → openaivec-1.0.0/src/openaivec/_cache/proxy.py

@@ -4,7 +4,7 @@ from collections.abc import Awaitable, Callable, Hashable
 from dataclasses import dataclass, field
 from typing import Any, Generic, TypeVar
 
-from openaivec._optimize import BatchSizeSuggester
+from openaivec._cache import BatchSizeSuggester
 
 __all__ = []
 

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/_embeddings.py

@@ -5,8 +5,8 @@ import numpy as np
 from numpy.typing import NDArray
 from openai import AsyncOpenAI, InternalServerError, OpenAI, RateLimitError
 
+from openaivec._cache import AsyncBatchingMapProxy, BatchingMapProxy
 from openaivec._log import observe
-from openaivec._proxy import AsyncBatchingMapProxy, BatchingMapProxy
 from openaivec._util import backoff, backoff_async
 
 __all__ = [

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/_prompt.py

@@ -6,7 +6,7 @@ construction of a prompt in a structured way, including setting the
 purpose, adding cautions, and providing examples.
 
 ```python
-from openaivec.prompt import FewShotPromptBuilder
+from openaivec import FewShotPromptBuilder
 
 prompt_str: str = (
     FewShotPromptBuilder()

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/_responses.py

@@ -7,9 +7,9 @@ from openai import AsyncOpenAI, BadRequestError, InternalServerError, OpenAI, Ra
 from openai.types.responses import ParsedResponse
 from pydantic import BaseModel
 
+from openaivec._cache import AsyncBatchingMapProxy, BatchingMapProxy
 from openaivec._log import observe
 from openaivec._model import PreparedTask, ResponseFormat
-from openaivec._proxy import AsyncBatchingMapProxy, BatchingMapProxy
 from openaivec._util import backoff, backoff_async
 
 __all__ = [

openaivec-1.0.0/src/openaivec/_schema/__init__.py

@@ -0,0 +1,9 @@
+"""Schema inference package.
+
+Internal helpers now live in :mod:`openaivec._schema.infer`; this module simply
+re-exports the main entry points so ``from openaivec._schema import ...`` still
+behaves the same."""
+
+from .infer import SchemaInferenceInput, SchemaInferenceOutput, SchemaInferer
+
+__all__ = ["SchemaInferenceOutput", "SchemaInferenceInput", "SchemaInferer"]

openaivec-0.99.2/src/openaivec/_schema.py → openaivec-1.0.0/src/openaivec/_schema/infer.py

@@ -61,14 +61,14 @@ from openai import OpenAI
 from openai.types.responses import ParsedResponse
 from pydantic import BaseModel, Field
 
-from openaivec._dynamic import ObjectSpec, _build_model
 from openaivec._model import PreparedTask
+from openaivec._schema.spec import ObjectSpec, _build_model
 
 # Internal module: explicitly not part of public API
 __all__: list[str] = []
 
 
-class InferredSchema(BaseModel):
+class SchemaInferenceOutput(BaseModel):
     """Result of a schema inference round.
 
     Contains the normalized *instructions*, objective *examples_summary*, the root
@@ -123,7 +123,7 @@ class InferredSchema(BaseModel):
     )
 
     @classmethod
-    def load(cls, path: str) -> "InferredSchema":
+    def load(cls, path: str) -> "SchemaInferenceOutput":
         """Load an inferred schema from a JSON file.
 
         Args:
@@ -265,7 +265,7 @@ class SchemaInferer:
     client: OpenAI
     model_name: str
 
-    def infer_schema(self, data: SchemaInferenceInput, *args, max_retries: int = 8, **kwargs) -> InferredSchema:
+    def infer_schema(self, data: SchemaInferenceInput, *args, max_retries: int = 8, **kwargs) -> SchemaInferenceOutput:
         """Infer a validated schema from representative examples.
 
           Workflow:
@@ -315,11 +315,11 @@ class SchemaInferer:
                 )
                 instructions = _INFER_INSTRUCTIONS + "\n\n" + "\n".join(feedback_lines)
 
-            response: ParsedResponse[InferredSchema] = self.client.responses.parse(
+            response: ParsedResponse[SchemaInferenceOutput] = self.client.responses.parse(
                 model=self.model_name,
                 instructions=instructions,
                 input=data.model_dump_json(),
-                text_format=InferredSchema,
+                text_format=SchemaInferenceOutput,
                 *args,
                 **kwargs,
             )

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/pandas_ext.py

@@ -54,12 +54,12 @@ import tiktoken
 from openai import AsyncOpenAI, OpenAI
 from pydantic import BaseModel
 
+from openaivec._cache import AsyncBatchingMapProxy, BatchingMapProxy
 from openaivec._embeddings import AsyncBatchEmbeddings, BatchEmbeddings
 from openaivec._model import EmbeddingsModelName, PreparedTask, ResponseFormat, ResponsesModelName
 from openaivec._provider import CONTAINER, _check_azure_v1_api_url
-from openaivec._proxy import AsyncBatchingMapProxy, BatchingMapProxy
 from openaivec._responses import AsyncBatchResponses, BatchResponses
-from openaivec._schema import InferredSchema, SchemaInferenceInput, SchemaInferer
+from openaivec._schema import SchemaInferenceInput, SchemaInferenceOutput, SchemaInferer
 from openaivec.task.table import FillNaResponse, fillna
 
 __all__ = [
@@ -308,7 +308,7 @@ class OpenAIVecSeriesAccessor:
 
         Example:
             ```python
-            from openaivec._proxy import BatchingMapProxy
+            from openaivec._cache import BatchingMapProxy
             import numpy as np
 
             # Create a shared cache with custom batch size
@@ -387,7 +387,7 @@ class OpenAIVecSeriesAccessor:
 
         Example:
             ```python
-            from openaivec._proxy import BatchingMapProxy
+            from openaivec._cache import BatchingMapProxy
             shared_cache = BatchingMapProxy(batch_size=64)
             reviews.ai.task_with_cache(sentiment_task, cache=shared_cache)
             ```
@@ -503,7 +503,7 @@ class OpenAIVecSeriesAccessor:
                 schema model, aligned with the original Series index.
         """
 
-        schema: InferredSchema | None = None
+        schema: SchemaInferenceOutput | None = None
         if response_format is None:
             schema = self.infer_schema(instructions=instructions, max_examples=max_examples, **api_kwargs)
 
@@ -588,7 +588,7 @@ class OpenAIVecSeriesAccessor:
             **api_kwargs,
         )
 
-    def infer_schema(self, instructions: str, max_examples: int = 100, **api_kwargs) -> InferredSchema:
+    def infer_schema(self, instructions: str, max_examples: int = 100, **api_kwargs) -> SchemaInferenceOutput:
         """Infer a structured data schema from Series content using AI.
 
         This method analyzes a sample of Series values to automatically generate
@@ -730,7 +730,7 @@ class OpenAIVecDataFrameAccessor:
 
         Example:
             ```python
-            from openaivec._proxy import BatchingMapProxy
+            from openaivec._cache import BatchingMapProxy
 
             # Create a shared cache with custom batch size
             shared_cache = BatchingMapProxy(batch_size=64)
@@ -990,7 +990,7 @@ class OpenAIVecDataFrameAccessor:
             **api_kwargs,
         )
 
-    def infer_schema(self, instructions: str, max_examples: int = 100, **api_kwargs) -> InferredSchema:
+    def infer_schema(self, instructions: str, max_examples: int = 100, **api_kwargs) -> SchemaInferenceOutput:
         """Infer a structured data schema from DataFrame rows using AI.
 
         This method analyzes a sample of DataFrame rows to automatically infer
@@ -1317,7 +1317,7 @@ class AsyncOpenAIVecSeriesAccessor:
 
         Example:
             ```python
-            from openaivec._proxy import AsyncBatchingMapProxy
+            from openaivec._cache import AsyncBatchingMapProxy
             import numpy as np
 
             # Create a shared cache with custom batch size and concurrency
@@ -1424,7 +1424,7 @@ class AsyncOpenAIVecSeriesAccessor:
         Example:
             ```python
             from openaivec._model import PreparedTask
-            from openaivec._proxy import AsyncBatchingMapProxy
+            from openaivec._cache import AsyncBatchingMapProxy
 
             # Create a shared cache with custom batch size and concurrency
             shared_cache = AsyncBatchingMapProxy(batch_size=64, max_concurrency=4)
@@ -1556,7 +1556,7 @@ class AsyncOpenAIVecSeriesAccessor:
         Note:
             This is an asynchronous method and must be awaited.
         """
-        schema: InferredSchema | None = None
+        schema: SchemaInferenceOutput | None = None
         if response_format is None:
             # Use synchronous schema inference
             schema = self._obj.ai.infer_schema(instructions=instructions, max_examples=max_examples)
@@ -1650,7 +1650,7 @@ class AsyncOpenAIVecDataFrameAccessor:
 
         Example:
             ```python
-            from openaivec._proxy import AsyncBatchingMapProxy
+            from openaivec._cache import AsyncBatchingMapProxy
 
             # Create a shared cache with custom batch size and concurrency
             shared_cache = AsyncBatchingMapProxy(batch_size=64, max_concurrency=4)

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/spark.py

@@ -142,10 +142,10 @@ from pyspark.sql.udf import UserDefinedFunction
 from typing_extensions import Literal
 
 from openaivec import pandas_ext
+from openaivec._cache import AsyncBatchingMapProxy
 from openaivec._model import EmbeddingsModelName, PreparedTask, ResponseFormat, ResponsesModelName
 from openaivec._provider import CONTAINER
-from openaivec._proxy import AsyncBatchingMapProxy
-from openaivec._schema import InferredSchema, SchemaInferenceInput, SchemaInferer
+from openaivec._schema import SchemaInferenceInput, SchemaInferenceOutput, SchemaInferer
 from openaivec._serialize import deserialize_base_model, serialize_base_model
 from openaivec._util import TextChunker
 
@@ -518,7 +518,7 @@ def infer_schema(
     example_table_name: str,
     example_field_name: str,
     max_examples: int = 100,
-) -> InferredSchema:
+) -> SchemaInferenceOutput:
     """Infer the schema for a response format based on example data.
 
     This function retrieves examples from a Spark table and infers the schema
@@ -606,7 +606,7 @@ def parse_udf(
     if not response_format and not (example_field_name and example_table_name):
         raise ValueError("Either response_format or example_table_name and example_field_name must be provided.")
 
-    schema: InferredSchema | None = None
+    schema: SchemaInferenceOutput | None = None
 
     if not response_format:
         schema = infer_schema(

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/task/__init__.py

@@ -32,7 +32,7 @@ Specialized tasks for customer service operations:
 ### Quick Start with Default Tasks
 ```python
 from openai import OpenAI
-from openaivec._responses import BatchResponses
+from openaivec import BatchResponses
 from openaivec.task import nlp, customer_support
 
 client = OpenAI()
@@ -90,15 +90,17 @@ results_df = df.ai.extract("sentiment")
 
 ### Spark Integration
 ```python
-from openaivec.spark import ResponsesUDFBuilder
+from openaivec.spark import task_udf
 
 # Register UDF for large-scale processing
 spark.udf.register(
     "analyze_sentiment",
-    ResponsesUDFBuilder.of_openai(
-        api_key=api_key,
-        model_name="gpt-4.1-mini"
-    ).build_from_task(task=nlp.SENTIMENT_ANALYSIS)
+    task_udf(
+        task=nlp.SENTIMENT_ANALYSIS,
+        model_name="gpt-4.1-mini",
+        batch_size=64,
+        max_concurrency=8,
+    ),
 )
 
 # Use in Spark SQL

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/task/customer_support/customer_sentiment.py

@@ -9,7 +9,7 @@ Example:
 
     ```python
     from openai import OpenAI
-    from openaivec._responses import BatchResponses
+    from openaivec import BatchResponses
     from openaivec.task import customer_support
 
     client = OpenAI()

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/task/customer_support/inquiry_classification.py

@@ -8,7 +8,7 @@ Example:
 
     ```python
     from openai import OpenAI
-    from openaivec._responses import BatchResponses
+    from openaivec import BatchResponses
     from openaivec.task import customer_support
 
     client = OpenAI()

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/task/customer_support/inquiry_summary.py

@@ -9,7 +9,7 @@ Example:
 
     ```python
     from openai import OpenAI
-    from openaivec._responses import BatchResponses
+    from openaivec import BatchResponses
     from openaivec.task import customer_support
 
     client = OpenAI()

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/task/customer_support/intent_analysis.py

@@ -8,7 +8,7 @@ Example:
 
     ```python
     from openai import OpenAI
-    from openaivec._responses import BatchResponses
+    from openaivec import BatchResponses
     from openaivec.task import customer_support
 
     client = OpenAI()

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/task/customer_support/response_suggestion.py

@@ -9,7 +9,7 @@ Example:
 
     ```python
     from openai import OpenAI
-    from openaivec._responses import BatchResponses
+    from openaivec import BatchResponses
     from openaivec.task import customer_support
 
     client = OpenAI()

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/task/customer_support/urgency_analysis.py

@@ -8,7 +8,7 @@ Example:
 
     ```python
     from openai import OpenAI
-    from openaivec._responses import BatchResponses
+    from openaivec import BatchResponses
     from openaivec.task import customer_support
 
     client = OpenAI()

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/task/nlp/dependency_parsing.py

@@ -8,7 +8,7 @@ Example:
 
     ```python
     from openai import OpenAI
-    from openaivec._responses import BatchResponses
+    from openaivec import BatchResponses
     from openaivec.task import nlp
 
     client = OpenAI()

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/task/nlp/keyword_extraction.py

@@ -8,7 +8,7 @@ Example:
 
     ```python
     from openai import OpenAI
-    from openaivec._responses import BatchResponses
+    from openaivec import BatchResponses
     from openaivec.task import nlp
 
     client = OpenAI()

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/task/nlp/morphological_analysis.py

@@ -9,7 +9,7 @@ Example:
 
     ```python
     from openai import OpenAI
-    from openaivec._responses import BatchResponses
+    from openaivec import BatchResponses
     from openaivec.task import nlp
 
     client = OpenAI()

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/task/nlp/named_entity_recognition.py

@@ -8,7 +8,7 @@ Example:
 
     ```python
     from openai import OpenAI
-    from openaivec._responses import BatchResponses
+    from openaivec import BatchResponses
     from openaivec.task import nlp
 
     client = OpenAI()

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/task/nlp/sentiment_analysis.py

@@ -8,7 +8,7 @@ Example:
 
     ```python
     from openai import OpenAI
-    from openaivec._responses import BatchResponses
+    from openaivec import BatchResponses
     from openaivec.task import nlp
 
     client = OpenAI()

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/task/nlp/translation.py

@@ -13,7 +13,7 @@ Example:
 
     ```python
     from openai import OpenAI
-    from openaivec._responses import BatchResponses
+    from openaivec import BatchResponses
     from openaivec.task import nlp
 
     client = OpenAI()

{openaivec-0.99.2 → openaivec-1.0.0}/src/openaivec/task/table/fillna.py

@@ -33,7 +33,7 @@ Example:
 
     ```python
     from openai import OpenAI
-    from openaivec._responses import BatchResponses
+    from openaivec import BatchResponses
     from openaivec.task.table import fillna
 
     client = OpenAI()

{openaivec-0.99.2/tests → openaivec-1.0.0/tests/_cache}/test_optimize.py

@@ -4,7 +4,7 @@ from threading import Thread
 
 import pytest
 
-from openaivec._optimize import BatchSizeSuggester, PerformanceMetric
+from openaivec._cache import BatchSizeSuggester, PerformanceMetric
 
 
 class TestPerformanceMetric:

{openaivec-0.99.2/tests → openaivec-1.0.0/tests/_cache}/test_proxy.py

@@ -5,7 +5,7 @@ import time
 
 import pytest
 
-from openaivec._proxy import AsyncBatchingMapProxy, BatchingMapProxy
+from openaivec._cache import AsyncBatchingMapProxy, BatchingMapProxy
 
 
 def test_batching_map_proxy_batches_calls_by_batch_size():
@@ -115,7 +115,7 @@ def test_batching_map_proxy_rechecks_cache_within_batch_iteration():
 
 
 def test_batching_map_proxy_map_func_length_mismatch_raises_and_releases():
-    from openaivec._proxy import BatchingMapProxy
+    from openaivec._cache import BatchingMapProxy
 
     p = BatchingMapProxy[int, int](batch_size=3)
 
@@ -134,14 +134,14 @@ def test_batching_map_proxy_map_func_length_mismatch_raises_and_releases():
 
 # -------------------- Internal methods tests --------------------
 def test_internal_unique_in_order():
-    from openaivec._proxy import BatchingMapProxy
+    from openaivec._cache import BatchingMapProxy
 
     p = BatchingMapProxy[int, int]()
     assert p._unique_in_order([1, 1, 2, 3, 2, 4]) == [1, 2, 3, 4]
 
 
 def test_internal_normalized_batch_size():
-    from openaivec._proxy import BatchingMapProxy
+    from openaivec._cache import BatchingMapProxy
 
     p = BatchingMapProxy[int, int]()
     assert p._normalized_batch_size(5) == 5  # default None => total
@@ -152,7 +152,7 @@ def test_internal_normalized_batch_size():
 
 
 def test_internal_all_cached_and_values():
-    from openaivec._proxy import BatchingMapProxy
+    from openaivec._cache import BatchingMapProxy
 
     p = BatchingMapProxy[int, int]()
     # fill cache via public API
@@ -167,7 +167,7 @@ def test_internal_all_cached_and_values():
 def test_internal_acquire_ownership():
     import threading
 
-    from openaivec._proxy import BatchingMapProxy
+    from openaivec._cache import BatchingMapProxy
 
     p = BatchingMapProxy[int, int]()
     # Cache 1; mark 2 inflight; 3 is missing
@@ -185,7 +185,7 @@ def test_internal_acquire_ownership():
 def test_internal_finalize_success_and_failure():
     import threading
 
-    from openaivec._proxy import BatchingMapProxy
+    from openaivec._cache import BatchingMapProxy
 
     p = BatchingMapProxy[int, int]()
     inflight = getattr(p, "_inflight")
@@ -214,7 +214,7 @@ def test_internal_finalize_success_and_failure():
 
 
 def test_internal_process_owned_batches_and_skip_cached():
-    from openaivec._proxy import BatchingMapProxy
+    from openaivec._cache import BatchingMapProxy
 
     calls: list[list[int]] = []
 
@@ -242,7 +242,7 @@ def test_internal_wait_for_with_inflight_event():
     import threading
     import time
 
-    from openaivec._proxy import BatchingMapProxy
+    from openaivec._cache import BatchingMapProxy
 
     p = BatchingMapProxy[int, int]()
 
@@ -284,7 +284,7 @@ async def _afunc_echo(xs: list[int]) -> list[int]:
 
 
 def test_async_localproxy_basic(event_loop=None):
-    from openaivec._proxy import AsyncBatchingMapProxy
+    from openaivec._cache import AsyncBatchingMapProxy
 
     calls: list[list[int]] = []
 
@@ -304,7 +304,7 @@ def test_async_localproxy_basic(event_loop=None):
 
 
 def test_async_localproxy_dedup_and_cache(event_loop=None):
-    from openaivec._proxy import AsyncBatchingMapProxy
+    from openaivec._cache import AsyncBatchingMapProxy
 
     calls: list[list[int]] = []
 
@@ -326,7 +326,7 @@ def test_async_localproxy_dedup_and_cache(event_loop=None):
 
 
 def test_async_localproxy_concurrent_requests(event_loop=None):
-    from openaivec._proxy import AsyncBatchingMapProxy
+    from openaivec._cache import AsyncBatchingMapProxy
 
     calls: list[list[int]] = []
 
@@ -355,7 +355,7 @@ def test_async_localproxy_concurrent_requests(event_loop=None):
 
 
 def test_async_localproxy_max_concurrency_limit(event_loop=None):
-    from openaivec._proxy import AsyncBatchingMapProxy
+    from openaivec._cache import AsyncBatchingMapProxy
 
     current = 0
     peak = 0
@@ -383,7 +383,7 @@ def test_async_localproxy_max_concurrency_limit(event_loop=None):
 
 
 def test_async_localproxy_map_func_length_mismatch_raises_and_releases(event_loop=None):
-    from openaivec._proxy import AsyncBatchingMapProxy
+    from openaivec._cache import AsyncBatchingMapProxy
 
     async def bad(xs: list[int]) -> list[int]:
         return xs[:-1]
@@ -426,7 +426,7 @@ def test_sync_clear_releases_memory_and_recomputes():
 
 def test_batch_size_maximization_with_cache_hits():
     """Test that batch_size is maximized even when some items are cached."""
-    from openaivec._proxy import BatchingMapProxy
+    from openaivec._cache import BatchingMapProxy
 
     calls: list[list[int]] = []
 
@@ -458,7 +458,7 @@ def test_batch_size_maximization_with_cache_hits():
 
 def test_batch_size_maximization_complex_scenario():
     """Test batch_size maximization with more complex cache hit patterns."""
-    from openaivec._proxy import BatchingMapProxy
+    from openaivec._cache import BatchingMapProxy
 
     calls: list[list[int]] = []
 
@@ -513,7 +513,7 @@ async def test_async_clear_releases_memory_and_recomputes():
 @pytest.mark.asyncio
 async def test_async_batch_size_maximization_with_cache_hits():
     """Test that batch_size is maximized even when some items are cached (async version)."""
-    from openaivec._proxy import AsyncBatchingMapProxy
+    from openaivec._cache import AsyncBatchingMapProxy
 
     calls: list[list[int]] = []
 
@@ -547,7 +547,7 @@ async def test_async_batch_size_maximization_with_cache_hits():
 @pytest.mark.asyncio
 async def test_async_batch_size_maximization_complex_scenario():
     """Test batch_size maximization with more complex cache hit patterns (async version)."""
-    from openaivec._proxy import AsyncBatchingMapProxy
+    from openaivec._cache import AsyncBatchingMapProxy
 
     calls: list[list[int]] = []
 
@@ -584,7 +584,7 @@ async def test_async_batch_size_maximization_complex_scenario():
 
 def test_notebook_environment_detection():
     """Test notebook environment detection functionality."""
-    from openaivec._proxy import ProxyBase
+    from openaivec._cache import ProxyBase
 
     proxy = ProxyBase()
     # The method should return a boolean and not raise an exception
@@ -594,7 +594,7 @@ def test_notebook_environment_detection():
 
 def test_progress_bar_methods():
     """Test progress bar creation and management methods."""
-    from openaivec._proxy import ProxyBase
+    from openaivec._cache import ProxyBase
 
     proxy = ProxyBase()
     proxy.show_progress = True
@@ -677,7 +677,7 @@ async def test_async_batching_proxy_with_progress_enabled():
 
 def test_progress_bar_with_forced_notebook_environment():
     """Test progress bar functionality with forced notebook environment."""
-    from openaivec._proxy import ProxyBase
+    from openaivec._cache import ProxyBase
 
     # Monkey patch the notebook detection to return True
     original_method = ProxyBase._is_notebook_environment
@@ -707,7 +707,7 @@ def test_progress_bar_with_forced_notebook_environment():
 @pytest.mark.asyncio
 async def test_async_progress_bar_with_forced_notebook_environment():
     """Test async progress bar functionality with forced notebook environment."""
-    from openaivec._proxy import ProxyBase
+    from openaivec._cache import ProxyBase
 
     # Monkey patch the notebook detection to return True
     original_method = ProxyBase._is_notebook_environment

{openaivec-0.99.2/tests → openaivec-1.0.0/tests/_cache}/test_proxy_suggester.py

@@ -4,7 +4,7 @@ import asyncio
 
 import pytest
 
-from openaivec._proxy import AsyncBatchingMapProxy, BatchingMapProxy
+from openaivec._cache import AsyncBatchingMapProxy, BatchingMapProxy
 
 
 def test_sync_proxy_uses_suggester_when_batch_size_none():

openaivec-0.99.2/tests/test_schema.py → openaivec-1.0.0/tests/_schema/test_infer.py

@@ -4,8 +4,8 @@ from typing import get_args, get_origin
 import pytest
 from pydantic import BaseModel
 
-from openaivec._dynamic import EnumSpec, FieldSpec, ObjectSpec  # internal types for constructing test schemas
-from openaivec._schema import InferredSchema, SchemaInferenceInput, SchemaInferer  # type: ignore
+from openaivec._schema import SchemaInferenceInput, SchemaInferenceOutput, SchemaInferer  # type: ignore
+from openaivec._schema.spec import EnumSpec, FieldSpec, ObjectSpec  # internal types for constructing test schemas
 
 
 @pytest.fixture(scope="session")
@@ -136,7 +136,7 @@ class TestInferredSchemaBuildModel:
 
     def test_build_model_primitive_types(self):
         """Test that all primitive types are correctly mapped to Python types."""
-        schema = InferredSchema(
+        schema = SchemaInferenceOutput(
             instructions="Test primitive types",
             examples_summary="Various primitive type examples",
             examples_instructions_alignment="Primitive examples justify coverage of all base types",
@@ -167,7 +167,7 @@ class TestInferredSchemaBuildModel:
 
     def test_build_model_enum_field(self):
         """Test that enum fields generate proper Enum classes."""
-        schema = InferredSchema(
+        schema = SchemaInferenceOutput(
             instructions="Test enum types",
             examples_summary="Enum examples",
             examples_instructions_alignment="Stable status labels appear repeatedly, supporting enum creation",
@@ -207,7 +207,7 @@ class TestInferredSchemaBuildModel:
             FieldSpec(name="second_field", type="boolean", description="Second field"),
         ]
 
-        schema = InferredSchema(
+        schema = SchemaInferenceOutput(
             instructions="Test field ordering",
             examples_summary="Field ordering examples",
             examples_instructions_alignment="Ordering matters for deterministic downstream column alignment",
@@ -223,7 +223,7 @@ class TestInferredSchemaBuildModel:
 
     def test_build_model_field_descriptions(self):
         """Test that field descriptions are correctly included in the model."""
-        schema = InferredSchema(
+        schema = SchemaInferenceOutput(
             instructions="Test field descriptions",
             examples_summary="Description examples",
             examples_instructions_alignment="Descriptions guide extraction disambiguation",
@@ -246,7 +246,7 @@ class TestInferredSchemaBuildModel:
 
     def test_build_model_empty_fields(self):
         """Test behavior with empty fields list."""
-        schema = InferredSchema(
+        schema = SchemaInferenceOutput(
             instructions="Test empty fields",
             examples_summary="Empty examples",
             examples_instructions_alignment="Edge case of no extractable signals",
@@ -266,7 +266,7 @@ class TestInferredSchemaBuildModel:
 
     def test_build_model_mixed_enum_and_regular_fields(self):
         """Test a complex scenario with both enum and regular fields of all types."""
-        schema = InferredSchema(
+        schema = SchemaInferenceOutput(
             instructions="Test mixed field types",
             examples_summary="Mixed type examples",
             examples_instructions_alignment="Examples demonstrate diverse field types including enums",
@@ -313,7 +313,7 @@ class TestInferredSchemaBuildModel:
 
     def test_build_model_multiple_calls_independence(self):
         """Test that multiple calls to build_model return independent model classes."""
-        schema = InferredSchema(
+        schema = SchemaInferenceOutput(
             instructions="Test independence",
             examples_summary="Independence examples",
             examples_instructions_alignment="Independence ensures rebuilding yields fresh class objects",
@@ -338,7 +338,7 @@ class TestInferredSchemaBuildModel:
 
     def test_build_model_array_types(self):
         """Test that *_array types map to list element annotations and proper JSON Schema arrays."""
-        schema = InferredSchema(
+        schema = SchemaInferenceOutput(
             instructions="Test array types",
             examples_summary="Array type examples",
             examples_instructions_alignment="Examples justify homogeneous primitive arrays",

openaivec-0.99.2/tests/test_dynamic.py → openaivec-1.0.0/tests/_schema/test_spec.py

@@ -5,7 +5,7 @@ from typing import get_args, get_origin
 
 import pytest
 
-from openaivec._dynamic import _MAX_ENUM_VALUES, EnumSpec, FieldSpec, ObjectSpec, _build_model
+from openaivec._schema.spec import _MAX_ENUM_VALUES, EnumSpec, FieldSpec, ObjectSpec, _build_model
 
 # ----------------------------- Success Cases -----------------------------
 

{openaivec-0.99.2 → openaivec-1.0.0}/tests/conftest.py

@@ -269,7 +269,7 @@ def performance_timer():
 @pytest.fixture
 def batch_cache():
     """BatchingMapProxy cache for testing."""
-    from openaivec._proxy import BatchingMapProxy
+    from openaivec._cache import BatchingMapProxy
 
     return BatchingMapProxy(batch_size=32)
 
@@ -277,7 +277,7 @@ def batch_cache():
 @pytest.fixture
 def async_batch_cache():
     """AsyncBatchingMapProxy cache for testing."""
-    from openaivec._proxy import AsyncBatchingMapProxy
+    from openaivec._cache import AsyncBatchingMapProxy
 
     return AsyncBatchingMapProxy(batch_size=32, max_concurrency=4)
 

{openaivec-0.99.2 → openaivec-1.0.0}/tests/test_pandas_ext.py

@@ -532,7 +532,7 @@ class TestPandasExt:
 
     def test_shared_cache_responses_sync(self):
         """Test shared cache functionality for responses."""
-        from openaivec._proxy import BatchingMapProxy
+        from openaivec._cache import BatchingMapProxy
 
         shared_cache = BatchingMapProxy(batch_size=32)
         series1 = pd.Series(["cat", "dog", "elephant"])
@@ -557,7 +557,7 @@ class TestPandasExt:
 
     def test_shared_cache_embeddings_sync(self):
         """Test shared cache functionality for embeddings."""
-        from openaivec._proxy import BatchingMapProxy
+        from openaivec._cache import BatchingMapProxy
 
         shared_cache = BatchingMapProxy(batch_size=32)
         series1 = pd.Series(["apple", "banana", "cherry"])
@@ -582,7 +582,7 @@ class TestPandasExt:
 
     def test_shared_cache_async(self):
         """Test shared cache functionality for async methods."""
-        from openaivec._proxy import AsyncBatchingMapProxy
+        from openaivec._cache import AsyncBatchingMapProxy
 
         async def run_test():
             shared_cache = AsyncBatchingMapProxy(batch_size=32, max_concurrency=4)
@@ -723,7 +723,7 @@ class TestPandasExt:
 
     def test_parse_with_cache_methods(self):
         """Test parse_with_cache methods for both Series and DataFrame."""
-        from openaivec._proxy import BatchingMapProxy
+        from openaivec._cache import BatchingMapProxy
 
         # Test Series parse_with_cache
         series = pd.Series(["Good product", "Bad experience"])