test(azure): drive an async azure_ad_token_provider through the SDK against a scripted Azure peer

tests/integration/sdk/test_azure_async_token_provider_sdk.py

@@ -0,0 +1,52 @@
+import asyncio
+import json
+from typing import Final
+
+import pytest
+from integration._support.wire import Reply, Request, wire_server
+
+import litellm
+
+_TOKEN: Final = "async-provider-token"
+_COMPLETION: Final = json.dumps(
+    {
+        "id": "chatcmpl-async-provider",
+        "object": "chat.completion",
+        "created": 1700000000,
+        "model": "gpt-4o",
+        "choices": [{"index": 0, "message": {"role": "assistant", "content": "ok"}, "finish_reason": "stop"}],
+        "usage": {"prompt_tokens": 1, "completion_tokens": 1, "total_tokens": 2},
+    }
+).encode()
+
+
+def _azure(_: Request) -> Reply:
+    return Reply(body=_COMPLETION)
+
+
+async def _async_token_provider() -> str:
+    await asyncio.sleep(0)
+    return _TOKEN
+
+
+@pytest.mark.parametrize("api_version", ["2024-10-21", "v1"])
+def test_an_async_azure_ad_token_provider_authorizes_the_upstream_call(
+    api_version: str, monkeypatch: pytest.MonkeyPatch
+) -> None:
+    for name in ("AZURE_API_KEY", "AZURE_OPENAI_API_KEY", "AZURE_AD_TOKEN"):
+        monkeypatch.delenv(name, raising=False)
+    with wire_server(_azure) as azure:
+        response: Final = asyncio.run(
+            litellm.acompletion(
+                model="azure/gpt-4o",
+                api_base=azure.url,
+                api_version=api_version,
+                azure_ad_token_provider=_async_token_provider,
+                messages=[{"role": "user", "content": "hi"}],
+                max_retries=0,
+            )
+        )
+        authorizations: Final = tuple(request.headers.get("authorization") for request in azure.drain())
+
+    assert response.choices[0].message.content == "ok"
+    assert authorizations == (f"Bearer {_TOKEN}",)