Make repo Python 3.9 compatible

.devcontainer/docker-compose.yaml

@@ -6,7 +6,7 @@ services:
       context: ..
       dockerfile: .devcontainer/Dockerfile
       args:
-        IMAGE: python:3.11
+        IMAGE: python:3.12
 
     volumes:
       - ..:/workspace:cached

.github/workflows/app-tests.yaml

@@ -28,8 +28,10 @@ jobs:
       fail-fast: false
       matrix:
         os: ["ubuntu-latest", "macos-latest-xlarge", "macos-13", "windows-latest"]
-        python_version: ["3.10"]
+        python_version: ["3.9", "3.10", "3.11", "3.12"]
         exclude:
+          - os: macos-latest-xlarge
+            python_version: "3.9"
           - os: macos-latest-xlarge
             python_version: "3.10"
     env:

README.md

@@ -69,7 +69,7 @@ A related option is VS Code Dev Containers, which will open the project in your
 
     * [Azure Developer CLI (azd)](https://aka.ms/install-azd)
     * [Node.js 18+](https://nodejs.org/download/)
-    * [Python 3.10+](https://www.python.org/downloads/)
+    * [Python 3.9+](https://www.python.org/downloads/)
     * [PostgreSQL 14+](https://www.postgresql.org/download/)
     * [pgvector](https://github.com/pgvector/pgvector)
     * [Docker Desktop](https://www.docker.com/products/docker-desktop/)

evals/generate_ground_truth.py

@@ -3,6 +3,7 @@
 import os
 from collections.abc import Generator
 from pathlib import Path
+from typing import Union
 
 from azure.identity import AzureDeveloperCliCredential, get_bearer_token_provider
 from dotenv_azd import load_azd_env
@@ -77,9 +78,9 @@ def answer_formatter(answer, source) -> str:
     return f"{answer} [{source['id']}]"
 
 
-def get_openai_client() -> tuple[AzureOpenAI | OpenAI, str]:
+def get_openai_client() -> tuple[Union[AzureOpenAI, OpenAI], str]:
     """Return an OpenAI client based on the environment variables"""
-    openai_client: AzureOpenAI | OpenAI
+    openai_client: Union[AzureOpenAI, OpenAI]
     OPENAI_CHAT_HOST = os.getenv("OPENAI_CHAT_HOST")
     if OPENAI_CHAT_HOST == "azure":
         if api_key := os.getenv("AZURE_OPENAI_KEY"):

pyproject.toml

@@ -1,13 +1,13 @@
 [tool.ruff]
 line-length = 120
-target-version = "py312"
+target-version = "py39"
 lint.select = ["E", "F", "I", "UP"]
 lint.ignore = ["D203"]
 lint.isort.known-first-party = ["fastapi_app"]
 
 [tool.mypy]
 check_untyped_defs = true
-python_version = 3.12
+python_version = 3.9
 exclude = [".venv/*"]
 
 [tool.pytest.ini_options]

src/backend/fastapi_app/__init__.py

@@ -2,7 +2,7 @@
 import os
 from collections.abc import AsyncIterator
 from contextlib import asynccontextmanager
-from typing import TypedDict
+from typing import TypedDict, Union
 
 import fastapi
 from azure.monitor.opentelemetry import configure_azure_monitor
@@ -27,8 +27,8 @@
 class State(TypedDict):
     sessionmaker: async_sessionmaker[AsyncSession]
     context: FastAPIAppContext
-    chat_client: AsyncOpenAI | AsyncAzureOpenAI
-    embed_client: AsyncOpenAI | AsyncAzureOpenAI
+    chat_client: Union[AsyncOpenAI, AsyncAzureOpenAI]
+    embed_client: Union[AsyncOpenAI, AsyncAzureOpenAI]
 
 
 @asynccontextmanager

src/backend/fastapi_app/api_models.py

@@ -1,5 +1,5 @@
 from enum import Enum
-from typing import Any
+from typing import Any, Optional
 
 from openai.types.chat import ChatCompletionMessageParam
 from pydantic import BaseModel
@@ -27,8 +27,8 @@ class ChatRequestOverrides(BaseModel):
     temperature: float = 0.3
     retrieval_mode: RetrievalMode = RetrievalMode.HYBRID
     use_advanced_flow: bool = True
-    prompt_template: str | None = None
-    seed: int | None = None
+    prompt_template: Optional[str] = None
+    seed: Optional[int] = None
 
 
 class ChatRequestContext(BaseModel):
@@ -38,7 +38,7 @@ class ChatRequestContext(BaseModel):
 class ChatRequest(BaseModel):
     messages: list[ChatCompletionMessageParam]
     context: ChatRequestContext
-    sessionState: Any | None = None
+    sessionState: Optional[Any] = None
 
 
 class ThoughtStep(BaseModel):
@@ -50,7 +50,7 @@ class ThoughtStep(BaseModel):
 class RAGContext(BaseModel):
     data_points: dict[int, dict[str, Any]]
     thoughts: list[ThoughtStep]
-    followup_questions: list[str] | None = None
+    followup_questions: Optional[list[str]] = None
 
 
 class ErrorResponse(BaseModel):
@@ -60,13 +60,13 @@ class ErrorResponse(BaseModel):
 class RetrievalResponse(BaseModel):
     message: Message
     context: RAGContext
-    sessionState: Any | None = None
+    sessionState: Optional[Any] = None
 
 
 class RetrievalResponseDelta(BaseModel):
-    delta: Message | None = None
-    context: RAGContext | None = None
-    sessionState: Any | None = None
+    delta: Optional[Message] = None
+    context: Optional[RAGContext] = None
+    sessionState: Optional[Any] = None
 
 
 class ItemPublic(BaseModel):

src/backend/fastapi_app/dependencies.py

@@ -1,7 +1,7 @@
 import logging
 import os
 from collections.abc import AsyncGenerator
-from typing import Annotated
+from typing import Annotated, Optional, Union
 
 import azure.identity
 from fastapi import Depends, Request
@@ -17,7 +17,7 @@ class OpenAIClient(BaseModel):
     OpenAI client
     """
 
-    client: AsyncOpenAI | AsyncAzureOpenAI
+    client: Union[AsyncOpenAI, AsyncAzureOpenAI]
     model_config = {"arbitrary_types_allowed": True}
 
 
@@ -28,9 +28,9 @@ class FastAPIAppContext(BaseModel):
 
     openai_chat_model: str
     openai_embed_model: str
-    openai_embed_dimensions: int | None
-    openai_chat_deployment: str | None
-    openai_embed_deployment: str | None
+    openai_embed_dimensions: Optional[int]
+    openai_chat_deployment: Optional[str]
+    openai_embed_deployment: Optional[str]
     embedding_column: str
 
 
@@ -77,9 +77,9 @@ async def common_parameters():
 
 
 async def get_azure_credential() -> (
-    azure.identity.AzureDeveloperCliCredential | azure.identity.ManagedIdentityCredential
+    Union[azure.identity.AzureDeveloperCliCredential, azure.identity.ManagedIdentityCredential]
 ):
-    azure_credential: azure.identity.AzureDeveloperCliCredential | azure.identity.ManagedIdentityCredential
+    azure_credential: Union[azure.identity.AzureDeveloperCliCredential, azure.identity.ManagedIdentityCredential]
     try:
         if client_id := os.getenv("APP_IDENTITY_ID"):
             # Authenticate using a user-assigned managed identity on Azure

src/backend/fastapi_app/embeddings.py

@@ -1,16 +1,14 @@
-from typing import (
-    TypedDict,
-)
+from typing import Optional, TypedDict, Union
 
 from openai import AsyncAzureOpenAI, AsyncOpenAI
 
 
 async def compute_text_embedding(
     q: str,
-    openai_client: AsyncOpenAI | AsyncAzureOpenAI,
+    openai_client: Union[AsyncOpenAI, AsyncAzureOpenAI],
     embed_model: str,
-    embed_deployment: str | None = None,
-    embedding_dimensions: int | None = None,
+    embed_deployment: Optional[str] = None,
+    embedding_dimensions: Optional[int] = None,
 ) -> list[float]:
     SUPPORTED_DIMENSIONS_MODEL = {
         "text-embedding-ada-002": False,

src/backend/fastapi_app/openai_clients.py

@@ -1,5 +1,6 @@
 import logging
 import os
+from typing import Union
 
 import azure.identity
 import openai
@@ -8,9 +9,9 @@
 
 
 async def create_openai_chat_client(
-    azure_credential: azure.identity.AzureDeveloperCliCredential | azure.identity.ManagedIdentityCredential,
-) -> openai.AsyncAzureOpenAI | openai.AsyncOpenAI:
-    openai_chat_client: openai.AsyncAzureOpenAI | openai.AsyncOpenAI
+    azure_credential: Union[azure.identity.AzureDeveloperCliCredential, azure.identity.ManagedIdentityCredential],
+) -> Union[openai.AsyncAzureOpenAI, openai.AsyncOpenAI]:
+    openai_chat_client: Union[openai.AsyncAzureOpenAI, openai.AsyncOpenAI]
     OPENAI_CHAT_HOST = os.getenv("OPENAI_CHAT_HOST")
     if OPENAI_CHAT_HOST == "azure":
         api_version = os.environ["AZURE_OPENAI_VERSION"] or "2024-03-01-preview"
@@ -57,9 +58,9 @@ async def create_openai_chat_client(
 
 
 async def create_openai_embed_client(
-    azure_credential: azure.identity.AzureDeveloperCliCredential | azure.identity.ManagedIdentityCredential,
-) -> openai.AsyncAzureOpenAI | openai.AsyncOpenAI:
-    openai_embed_client: openai.AsyncAzureOpenAI | openai.AsyncOpenAI
+    azure_credential: Union[azure.identity.AzureDeveloperCliCredential, azure.identity.ManagedIdentityCredential],
+) -> Union[openai.AsyncAzureOpenAI, openai.AsyncOpenAI]:
+    openai_embed_client: Union[openai.AsyncAzureOpenAI, openai.AsyncOpenAI]
     OPENAI_EMBED_HOST = os.getenv("OPENAI_EMBED_HOST")
     if OPENAI_EMBED_HOST == "azure":
         api_version = os.environ["AZURE_OPENAI_VERSION"] or "2024-03-01-preview"

src/backend/fastapi_app/postgres_searcher.py

@@ -1,3 +1,5 @@
+from typing import Optional, Union
+
 import numpy as np
 from openai import AsyncAzureOpenAI, AsyncOpenAI
 from sqlalchemy import Float, Integer, column, select, text
@@ -11,10 +13,10 @@ class PostgresSearcher:
     def __init__(
         self,
         db_session: AsyncSession,
-        openai_embed_client: AsyncOpenAI | AsyncAzureOpenAI,
-        embed_deployment: str | None,  # Not needed for non-Azure OpenAI or for retrieval_mode="text"
+        openai_embed_client: Union[AsyncOpenAI, AsyncAzureOpenAI],
+        embed_deployment: Optional[str],  # Not needed for non-Azure OpenAI or for retrieval_mode="text"
         embed_model: str,
-        embed_dimensions: int | None,
+        embed_dimensions: Optional[int],
         embedding_column: str,
     ):
         self.db_session = db_session
@@ -38,7 +40,7 @@ def build_filter_clause(self, filters) -> tuple[str, str]:
         return "", ""
 
     async def search(
-        self, query_text: str | None, query_vector: list[float] | list, top: int = 5, filters: list[dict] | None = None
+        self, query_text: Optional[str], query_vector: list[float], top: int = 5, filters: Optional[list[dict]] = None
     ):
         filter_clause_where, filter_clause_and = self.build_filter_clause(filters)
         table_name = Item.__tablename__
@@ -100,11 +102,11 @@ async def search(
 
     async def search_and_embed(
         self,
-        query_text: str | None = None,
+        query_text: Optional[str] = None,
         top: int = 5,
         enable_vector_search: bool = False,
         enable_text_search: bool = False,
-        filters: list[dict] | None = None,
+        filters: Optional[list[dict]] = None,
     ) -> list[Item]:
         """
         Search rows by query text. Optionally converts the query text to a vector if enable_vector_search is True.

src/backend/fastapi_app/rag_advanced.py

@@ -1,5 +1,5 @@
 from collections.abc import AsyncGenerator
-from typing import Any, Final
+from typing import Any, Final, Optional, Union
 
 from openai import AsyncAzureOpenAI, AsyncOpenAI, AsyncStream
 from openai.types.chat import ChatCompletion, ChatCompletionChunk, ChatCompletionMessageParam
@@ -24,9 +24,9 @@ def __init__(
         self,
         *,
         searcher: PostgresSearcher,
-        openai_chat_client: AsyncOpenAI | AsyncAzureOpenAI,
+        openai_chat_client: Union[AsyncOpenAI, AsyncAzureOpenAI],
         chat_model: str,
-        chat_deployment: str | None,  # Not needed for non-Azure OpenAI
+        chat_deployment: Optional[str],  # Not needed for non-Azure OpenAI
     ):
         self.searcher = searcher
         self.openai_chat_client = openai_chat_client
@@ -39,8 +39,8 @@ async def generate_search_query(
         original_user_query: str,
         past_messages: list[ChatCompletionMessageParam],
         query_response_token_limit: int,
-        seed: int | None = None,
-    ) -> tuple[list[ChatCompletionMessageParam], Any | str | None, list]:
+        seed: Optional[int] = None,
+    ) -> tuple[list[ChatCompletionMessageParam], Union[Any, str, None], list]:
         """Generate an optimized keyword search query based on the chat history and the last question"""
 
         tools = build_search_function()

src/backend/fastapi_app/rag_simple.py

@@ -1,4 +1,5 @@
 from collections.abc import AsyncGenerator
+from typing import Optional, Union
 
 from openai import AsyncAzureOpenAI, AsyncOpenAI, AsyncStream
 from openai.types.chat import ChatCompletion, ChatCompletionChunk, ChatCompletionMessageParam
@@ -22,9 +23,9 @@ def __init__(
         self,
         *,
         searcher: PostgresSearcher,
-        openai_chat_client: AsyncOpenAI | AsyncAzureOpenAI,
+        openai_chat_client: Union[AsyncOpenAI, AsyncAzureOpenAI],
         chat_model: str,
-        chat_deployment: str | None,  # Not needed for non-Azure OpenAI
+        chat_deployment: Optional[str],  # Not needed for non-Azure OpenAI
     ):
         self.searcher = searcher
         self.openai_chat_client = openai_chat_client

src/backend/fastapi_app/routes/api_routes.py

@@ -1,6 +1,7 @@
 import json
 import logging
 from collections.abc import AsyncGenerator
+from typing import Union
 
 import fastapi
 from fastapi import HTTPException
@@ -93,7 +94,7 @@ async def search_handler(
     return [ItemPublic.model_validate(item.to_dict()) for item in results]
 
 
-@router.post("/chat", response_model=RetrievalResponse | ErrorResponse)
+@router.post("/chat", response_model=Union[RetrievalResponse, ErrorResponse])
 async def chat_handler(
     context: CommonDeps,
     database_session: DBSession,
@@ -110,7 +111,7 @@ async def chat_handler(
             embed_dimensions=context.openai_embed_dimensions,
             embedding_column=context.embedding_column,
         )
-        rag_flow: SimpleRAGChat | AdvancedRAGChat
+        rag_flow: Union[SimpleRAGChat, AdvancedRAGChat]
         if chat_request.context.overrides.use_advanced_flow:
             rag_flow = AdvancedRAGChat(
                 searcher=searcher,
@@ -154,7 +155,7 @@ async def chat_stream_handler(
         embedding_column=context.embedding_column,
     )
 
-    rag_flow: SimpleRAGChat | AdvancedRAGChat
+    rag_flow: Union[SimpleRAGChat, AdvancedRAGChat]
     if chat_request.context.overrides.use_advanced_flow:
         rag_flow = AdvancedRAGChat(
             searcher=searcher,

tests/mocks.py

@@ -1,4 +1,4 @@
-from typing import Any
+from typing import Any, Optional
 
 from azure.core.credentials import AccessToken, TokenCredential
 
@@ -7,8 +7,8 @@ class MockAzureCredential(TokenCredential):
     def get_token(
         self,
         *scopes: str,
-        claims: str | None = None,
-        tenant_id: str | None = None,
+        claims: Optional[str] = None,
+        tenant_id: Optional[str] = None,
         enable_cae: bool = False,
         **kwargs: Any,
     ) -> AccessToken:
@@ -22,8 +22,8 @@ def __init__(self):
     def get_token(
         self,
         *scopes: str,
-        claims: str | None = None,
-        tenant_id: str | None = None,
+        claims: Optional[str] = None,
+        tenant_id: Optional[str] = None,
         enable_cae: bool = False,
         **kwargs: Any,
     ) -> AccessToken: