Add FastAPI HTTP request benchmark

pyperformance/data-files/benchmarks/MANIFEST

@@ -42,6 +42,7 @@ django_template	<local>
 dulwich_log	<local>
 docutils	<local>
 fannkuch	<local>
+fastapi	<local>
 float	<local>
 genshi	<local>
 go	<local>

pyperformance/data-files/benchmarks/bm_fastapi/pyproject.toml

@@ -0,0 +1,15 @@
+[project]
+name = "pyperformance_bm_fastapi"
+requires-python = ">=3.10"
+dependencies = [
+    "pyperf",
+    "fastapi",
+    "uvicorn",
+    "httpx",
+]
+urls = {repository = "https://github.com/python/pyperformance"}
+dynamic = ["version"]
+
+[tool.pyperformance]
+name = "fastapi"
+tags = "apps"

pyperformance/data-files/benchmarks/bm_fastapi/requirements.txt

@@ -0,0 +1,3 @@
+fastapi==0.121.0
+httpx==0.28.1
+uvicorn==0.38.0

pyperformance/data-files/benchmarks/bm_fastapi/run_benchmark.py

@@ -0,0 +1,87 @@
+"""
+Test the performance of simple HTTP serving with FastAPI.
+
+This benchmark tests FastAPI's request handling, including:
+- Path parameter extraction and validation
+- Pydantic model serialization
+- JSON response encoding
+
+The bench serves a REST API endpoint that returns JSON objects,
+simulating a typical web application scenario.
+
+Author: Savannah Ostrowski
+"""
+
+import asyncio
+import socket
+import time
+
+import httpx
+import pyperf
+import uvicorn
+from fastapi import FastAPI
+from pydantic import BaseModel
+
+HOST = "127.0.0.1"
+
+CONCURRENCY = 150
+
+class Item(BaseModel):
+    id: int
+    name: str
+    price: float
+    tags: list[str] = []
+
+app = FastAPI()
+
+@app.get("/items/{item_id}", response_model=Item)
+async def get_item(item_id: int):
+    return {
+        "id": item_id,
+        "name": "Sample Item",
+        "price": 9.99,
+        "tags": ["sample", "item", "fastapi"]
+    }
+
+
+async def run_server_and_benchmark(loops):
+    with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as s:
+        s.bind((HOST, 0))
+        s.listen(1)
+        port = s.getsockname()[1]
+
+    config = uvicorn.Config(app, host=HOST, port=port, log_level="error")
+    server = uvicorn.Server(config)
+    server_task = asyncio.create_task(server.serve())
+
+    await asyncio.sleep(0.5)
+
+    async with httpx.AsyncClient() as client:
+        t0 = time.perf_counter()
+
+        for i in range(loops):
+            tasks = [
+                client.get(f"http://{HOST}:{port}/items/{i}")
+                for _ in range(CONCURRENCY)
+            ]
+            responses = await asyncio.gather(*tasks)
+            for response in responses:
+                response.raise_for_status()
+                data = response.json()
+                assert data["id"] == i
+                assert "tags" in data
+
+        elapsed = time.perf_counter() - t0
+
+    server.should_exit = True
+    await server_task
+    return elapsed
+
+def bench_fastapi(loops):
+    return asyncio.run(run_server_and_benchmark(loops))
+
+
+if __name__ == "__main__":
+    runner = pyperf.Runner()
+    runner.metadata['description'] = "Test the performance of HTTP requests with FastAPI"
+    runner.bench_time_func("fastapi_http", bench_fastapi)