Add graphrag-factory package

Author: dworthen

packages/graphrag-factory/README.md

@@ -0,0 +1,49 @@
+# GraphRAG Factory
+
+```python
+from abc import ABC, abstractmethod
+
+from graphrag_factory import Factory
+
+class SampleABC(ABC):
+
+    @abstractmethod
+    def get_value(self) -> str:
+        msg = "Subclasses must implement the get_value method."
+        raise NotImplementedError(msg)
+
+
+class ConcreteClass(SampleABC):
+    def __init__(self, value: str):
+        self._value = value
+
+    def get_value(self) -> str:
+        return self._value
+
+class SampleFactory(Factory[SampleABC]):
+"""A Factory for SampleABC classes."""
+
+factory = SampleFactory()
+
+# Registering transient services
+# A new one is created for every request
+factory.register("some_strategy", ConcreteTestClass)
+
+trans1 = factory.create("some_strategy", {"value": "test1"})
+trans2 = factory.create("some_strategy", {"value": "test2"})
+
+assert trans1 is not trans2
+assert trans1.get_value() == "test1"
+assert trans2.get_value() == "test2"
+
+# Registering singleton services
+# After first creation, the same one is returned every time
+factory.register("some_other_strategy", ConcreteTestClass, scope="singleton")
+
+single1 = factory.create("some_other_strategy", {"value": "singleton"})
+single2 = factory.create("some_other_strategy", {"value": "ignored"})
+
+assert single1 is single2
+assert single1.get_value() == "singleton"
+assert single2.get_value() == "singleton"
+```

packages/graphrag-factory/graphrag_factory/__init__.py

@@ -0,0 +1,8 @@
+# Copyright (c) 2024 Microsoft Corporation.
+# Licensed under the MIT License
+
+"""The GraphRAG Factory package."""
+
+from graphrag_factory.factory import Factory
+
+__all__ = ["Factory"]

…ges/graphrag/graphrag/factory/factory.py …hrag-factory/graphrag_factory/factory.pypackages/graphrag/graphrag/factory/factory.py renamed to packages/graphrag-factory/graphrag_factory/factory.py packages/graphrag/graphrag/factory/factory.py renamed to packages/graphrag-factory/graphrag_factory/factory.py

@@ -5,10 +5,12 @@
 
 from abc import ABC
 from collections.abc import Callable
-from typing import Any, ClassVar, Generic, TypeVar
+from typing import Any, ClassVar, Generic, Literal, TypeVar
 
 T = TypeVar("T", covariant=True)
 
+ServiceScope = Literal["singleton", "transient"]
+
 
 class Factory(ABC, Generic[T]):
     """Abstract base class for factories."""
@@ -23,27 +25,36 @@ def __new__(cls, *args: Any, **kwargs: Any) -> "Factory":
 
     def __init__(self):
         if not hasattr(self, "_initialized"):
-            self._services: dict[str, Callable[..., T]] = {}
+            self._service_initializers: dict[
+                str, tuple[ServiceScope, Callable[..., T]]
+            ] = {}
+            self._initialized_services: dict[str, T] = {}
             self._initialized = True
 
     def __contains__(self, strategy: str) -> bool:
         """Check if a strategy is registered."""
-        return strategy in self._services
+        return strategy in self._service_initializers
 
     def keys(self) -> list[str]:
         """Get a list of registered strategy names."""
-        return list(self._services.keys())
-
-    def register(self, strategy: str, initializer: Callable[..., T]) -> None:
+        return list(self._service_initializers.keys())
+
+    def register(
+        self,
+        strategy: str,
+        initializer: Callable[..., T],
+        scope: ServiceScope = "transient",
+    ) -> None:
         """
         Register a new service.
 
         Args
         ----
             strategy: The name of the strategy.
             initializer: A callable that creates an instance of T.
+            scope: The service scope, either 'singleton' or 'transient'.
         """
-        self._services[strategy] = initializer
+        self._service_initializers[strategy] = (scope, initializer)
 
     def create(self, strategy: str, init_args: dict[str, Any] | None = None) -> T:
         """
@@ -62,7 +73,16 @@ def create(self, strategy: str, init_args: dict[str, Any] | None = None) -> T:
         ------
             ValueError: If the strategy is not registered.
         """
-        if strategy not in self._services:
+        if strategy not in self._service_initializers:
             msg = f"Strategy '{strategy}' is not registered."
             raise ValueError(msg)
-        return self._services[strategy](**(init_args or {}))
+
+        scope, service_initializer = self._service_initializers[strategy]
+        if scope == "singleton":
+            if strategy not in self._initialized_services:
+                self._initialized_services[strategy] = service_initializer(
+                    **(init_args or {})
+                )
+            return self._initialized_services[strategy]
+
+        return service_initializer(**(init_args or {}))

packages/graphrag-factory/graphrag_factory/py.typed

@@ -0,0 +1,41 @@
+[project]
+name = "graphrag-factory"
+version = "2.7.0"
+description = "GraphRAG: A graph-based retrieval-augmented generation (RAG) system."
+authors = [
+    {name = "Alonso Guevara Fernández", email = "[email protected]"},
+    {name = "Andrés Morales Esquivel", email = "[email protected]"},
+    {name = "Chris Trevino", email = "[email protected]"},
+    {name = "David Tittsworth", email = "[email protected]"},
+    {name = "Dayenne de Souza", email = "[email protected]"},
+    {name = "Derek Worthen", email = "[email protected]"},
+    {name = "Gaudy Blanco Meneses", email = "[email protected]"},
+    {name = "Ha Trinh", email = "[email protected]"},
+    {name = "Jonathan Larson", email = "[email protected]"},
+    {name = "Josh Bradley", email = "[email protected]"},
+    {name = "Kate Lytvynets", email = "[email protected]"},
+    {name = "Kenny Zhang", email = "[email protected]"},
+    {name = "Mónica Carvajal"},
+    {name = "Nathan Evans", email = "[email protected]"},
+    {name = "Rodrigo Racanicci", email = "[email protected]"},
+    {name = "Sarah Smith", email = "[email protected]"},
+]
+license = "MIT"
+readme = "README.md"
+license-files = ["LICENSE"]
+requires-python = ">=3.10,<3.13"
+classifiers = [
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+]
+dependencies = []
+
+[project.urls]
+Source = "https://github.com/microsoft/graphrag"
+
+[build-system]
+requires = ["hatchling>=1.27.0,<2.0.0"]
+build-backend = "hatchling.build"
+

packages/graphrag-factory/pyproject.toml

@@ -5,12 +5,13 @@
 
 from __future__ import annotations
 
+from graphrag_factory import Factory
+
 from graphrag.cache.json_pipeline_cache import JsonPipelineCache
 from graphrag.cache.memory_pipeline_cache import InMemoryCache
 from graphrag.cache.noop_pipeline_cache import NoopPipelineCache
 from graphrag.cache.pipeline_cache import PipelineCache
 from graphrag.config.enums import CacheType
-from graphrag.factory.factory import Factory
 from graphrag.storage.blob_pipeline_storage import BlobPipelineStorage
 from graphrag.storage.cosmosdb_pipeline_storage import CosmosDBPipelineStorage
 from graphrag.storage.file_pipeline_storage import FilePipelineStorage

packages/graphrag/graphrag/cache/factory.py

@@ -5,8 +5,9 @@
 
 import logging
 
+from graphrag_factory import Factory
+
 from graphrag.config.enums import InputFileType
-from graphrag.factory.factory import Factory
 from graphrag.index.input.csv import CSVFileReader
 from graphrag.index.input.input_reader import InputReader
 from graphrag.index.input.json import JSONFileReader

packages/graphrag/graphrag/factory/__init__.py

@@ -3,8 +3,9 @@
 
 """A package containing a factory for supported llm types."""
 
+from graphrag_factory import Factory
+
 from graphrag.config.enums import ModelType
-from graphrag.factory.factory import Factory
 from graphrag.language_model.protocol.base import ChatModel, EmbeddingModel
 from graphrag.language_model.providers.litellm.chat_model import LitellmChatModel
 from graphrag.language_model.providers.litellm.embedding_model import (

packages/graphrag/graphrag/index/input/factory.py

@@ -3,7 +3,8 @@
 
 """LiteLLM Rate Limiter Factory."""
 
-from graphrag.factory.factory import Factory
+from graphrag_factory import Factory
+
 from graphrag.language_model.providers.litellm.services.rate_limiter.rate_limiter import (
     RateLimiter,
 )