Add Responses API with streaming, history, and advanced features
Implements Phases 2, 3, and 6: Complete Responses API implementation Core API (Phase 2): - OpenAI Responses API format with structured output items - Streaming and non-streaming support via SSE-Starlette - Reasoning items (thinking summaries) - Function call items (tool execution) - Message items (assistant responses) - Router, schemas, service, and streaming coordinator Conversation History (Phase 3): - Hybrid client/server approach - Auto-generated deterministic conversation IDs - Configurable max turns with automatic trimming - Context window management with token counting - Token usage statistics - Placeholder for future vector memory integration Advanced Features (Phase 6): - Parameter validation with Pydantic field validators: - Temperature: 0.0-2.0 range enforcement - Reasoning effort: 6 levels (none to xhigh) - Max output tokens: positive integer enforcement - Stop sequences: up to 4, non-empty strings - Real-time stop sequence detection during streaming - Real-time max tokens enforcement with token counting - Graceful error handling and OpenAI-compatible error format Testing: - 9 unit tests for API endpoints and streaming - 11 unit tests for error handling - 13 unit tests for conversation history and context - 12 unit tests for advanced features and validation - Total: 45 tests with comprehensive coverage 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
This commit is contained in:
Claude
@@ -0,0 +1,281 @@
|
||||
"""
|
||||
Tests for Responses API router.
|
||||
"""
|
||||
|
||||
import json
|
||||
import pytest
|
||||
from fastapi.testclient import TestClient
|
||||
from httpx import AsyncClient
|
||||
|
||||
|
||||
@pytest.mark.unit
|
||||
def test_create_response_non_streaming(client: TestClient) -> None:
|
||||
"""Test non-streaming response creation."""
|
||||
request_data = {
|
||||
"model": "lorem-tester",
|
||||
"input": [
|
||||
{"role": "user", "content": "Hello, how are you?"}
|
||||
],
|
||||
"stream": False
|
||||
}
|
||||
|
||||
response = client.post("/v1/responses", json=request_data)
|
||||
|
||||
assert response.status_code == 200
|
||||
data = response.json()
|
||||
|
||||
# Verify response structure
|
||||
assert "id" in data
|
||||
assert data["object"] == "response"
|
||||
assert "created_at" in data
|
||||
assert data["model"] == "lorem-tester"
|
||||
assert data["status"] == "completed"
|
||||
assert "output" in data
|
||||
assert "usage" in data
|
||||
|
||||
# Verify output array
|
||||
assert isinstance(data["output"], list)
|
||||
assert len(data["output"]) >= 1 # At least message item
|
||||
|
||||
# Verify usage
|
||||
assert data["usage"]["input_tokens"] > 0
|
||||
assert data["usage"]["output_tokens"] > 0
|
||||
assert data["usage"]["total_tokens"] > 0
|
||||
|
||||
|
||||
@pytest.mark.unit
|
||||
def test_create_response_with_reasoning(client: TestClient) -> None:
|
||||
"""Test response with reasoning enabled."""
|
||||
request_data = {
|
||||
"model": "lorem-tester",
|
||||
"input": [
|
||||
{"role": "user", "content": "Explain something"}
|
||||
],
|
||||
"reasoning": {
|
||||
"effort": "medium",
|
||||
"summary": "auto"
|
||||
},
|
||||
"stream": False
|
||||
}
|
||||
|
||||
response = client.post("/v1/responses", json=request_data)
|
||||
|
||||
assert response.status_code == 200
|
||||
data = response.json()
|
||||
|
||||
# Should have reasoning item and message item
|
||||
assert len(data["output"]) >= 2
|
||||
|
||||
# Check for reasoning item
|
||||
reasoning_items = [item for item in data["output"] if item["type"] == "reasoning"]
|
||||
assert len(reasoning_items) >= 1
|
||||
|
||||
reasoning_item = reasoning_items[0]
|
||||
assert "summary" in reasoning_item
|
||||
assert isinstance(reasoning_item["summary"], list)
|
||||
assert len(reasoning_item["summary"]) > 0
|
||||
|
||||
# Check for message item
|
||||
message_items = [item for item in data["output"] if item["type"] == "message"]
|
||||
assert len(message_items) >= 1
|
||||
|
||||
# Verify reasoning tokens counted
|
||||
assert data["usage"]["reasoning_tokens"] > 0
|
||||
|
||||
|
||||
@pytest.mark.unit
|
||||
def test_create_response_with_tools(client: TestClient) -> None:
|
||||
"""Test response with tools available."""
|
||||
request_data = {
|
||||
"model": "lorem-tester",
|
||||
"input": [
|
||||
{"role": "user", "content": "Use a tool to help"}
|
||||
],
|
||||
"tools": [
|
||||
{
|
||||
"name": "search_knowledge",
|
||||
"description": "Search knowledge base",
|
||||
"parameters": {}
|
||||
}
|
||||
],
|
||||
"stream": False
|
||||
}
|
||||
|
||||
response = client.post("/v1/responses", json=request_data)
|
||||
|
||||
assert response.status_code == 200
|
||||
data = response.json()
|
||||
|
||||
# May or may not have function_call items (randomized in lorem-tester)
|
||||
# But should always have message item
|
||||
message_items = [item for item in data["output"] if item["type"] == "message"]
|
||||
assert len(message_items) >= 1
|
||||
|
||||
|
||||
@pytest.mark.unit
|
||||
def test_create_response_invalid_model(client: TestClient) -> None:
|
||||
"""Test response with non-existent model."""
|
||||
request_data = {
|
||||
"model": "nonexistent-model",
|
||||
"input": [
|
||||
{"role": "user", "content": "Hello"}
|
||||
],
|
||||
"stream": False
|
||||
}
|
||||
|
||||
response = client.post("/v1/responses", json=request_data)
|
||||
|
||||
assert response.status_code == 404
|
||||
data = response.json()
|
||||
assert "detail" in data
|
||||
|
||||
|
||||
@pytest.mark.unit
|
||||
def test_create_response_validation_error(client: TestClient) -> None:
|
||||
"""Test response with invalid request data."""
|
||||
# Missing required 'model' field
|
||||
request_data = {
|
||||
"input": [
|
||||
{"role": "user", "content": "Hello"}
|
||||
]
|
||||
}
|
||||
|
||||
response = client.post("/v1/responses", json=request_data)
|
||||
|
||||
assert response.status_code == 422
|
||||
|
||||
|
||||
@pytest.mark.unit
|
||||
@pytest.mark.asyncio
|
||||
async def test_create_response_streaming(async_client: AsyncClient) -> None:
|
||||
"""Test streaming response creation."""
|
||||
request_data = {
|
||||
"model": "lorem-tester",
|
||||
"input": [
|
||||
{"role": "user", "content": "Hello"}
|
||||
],
|
||||
"stream": True
|
||||
}
|
||||
|
||||
async with async_client.stream(
|
||||
"POST",
|
||||
"/v1/responses",
|
||||
json=request_data,
|
||||
timeout=20.0,
|
||||
) as response:
|
||||
assert response.status_code == 200
|
||||
assert response.headers["content-type"] == "text/event-stream; charset=utf-8"
|
||||
|
||||
events = []
|
||||
async for line in response.aiter_lines():
|
||||
if not line.strip():
|
||||
continue
|
||||
|
||||
# Parse SSE format: "event: event_type" and "data: {json}"
|
||||
if line.startswith("event: "):
|
||||
event_type = line[7:].strip()
|
||||
elif line.startswith("data: "):
|
||||
data_str = line[6:].strip()
|
||||
try:
|
||||
data = json.loads(data_str)
|
||||
events.append({"event": event_type, "data": data})
|
||||
except json.JSONDecodeError:
|
||||
pass # Skip malformed data
|
||||
|
||||
# Should have multiple events
|
||||
assert len(events) > 0
|
||||
|
||||
# Should have response.done event as last event
|
||||
assert events[-1]["event"] == "response.done"
|
||||
|
||||
# response.done should have complete response
|
||||
done_data = events[-1]["data"]
|
||||
assert "response" in done_data
|
||||
assert done_data["response"]["status"] == "completed"
|
||||
|
||||
|
||||
@pytest.mark.unit
|
||||
@pytest.mark.asyncio
|
||||
async def test_create_response_streaming_with_reasoning(async_client: AsyncClient) -> None:
|
||||
"""Test streaming response with reasoning."""
|
||||
request_data = {
|
||||
"model": "lorem-tester",
|
||||
"input": [
|
||||
{"role": "user", "content": "Explain"}
|
||||
],
|
||||
"reasoning": {
|
||||
"effort": "medium",
|
||||
"summary": "auto"
|
||||
},
|
||||
"stream": True
|
||||
}
|
||||
|
||||
async with async_client.stream(
|
||||
"POST",
|
||||
"/v1/responses",
|
||||
json=request_data,
|
||||
timeout=20.0,
|
||||
) as response:
|
||||
assert response.status_code == 200
|
||||
|
||||
events = []
|
||||
event_type = None
|
||||
|
||||
async for line in response.aiter_lines():
|
||||
if not line.strip():
|
||||
continue
|
||||
|
||||
if line.startswith("event: "):
|
||||
event_type = line[7:].strip()
|
||||
elif line.startswith("data: "):
|
||||
data_str = line[6:].strip()
|
||||
try:
|
||||
data = json.loads(data_str)
|
||||
if event_type:
|
||||
events.append({"event": event_type, "data": data})
|
||||
except json.JSONDecodeError:
|
||||
pass
|
||||
|
||||
# Should have reasoning events
|
||||
reasoning_events = [
|
||||
e for e in events
|
||||
if e["event"] == "response.reasoning_summary_text.delta"
|
||||
]
|
||||
assert len(reasoning_events) > 0
|
||||
|
||||
# Should have output text events
|
||||
output_events = [
|
||||
e for e in events
|
||||
if e["event"] == "response.output_text.delta"
|
||||
]
|
||||
assert len(output_events) > 0
|
||||
|
||||
|
||||
@pytest.mark.unit
|
||||
def test_create_response_temperature_parameter(client: TestClient) -> None:
|
||||
"""Test temperature parameter handling."""
|
||||
request_data = {
|
||||
"model": "lorem-tester",
|
||||
"input": [{"role": "user", "content": "Test"}],
|
||||
"temperature": 0.5,
|
||||
"stream": False
|
||||
}
|
||||
|
||||
response = client.post("/v1/responses", json=request_data)
|
||||
|
||||
assert response.status_code == 200
|
||||
|
||||
|
||||
@pytest.mark.unit
|
||||
def test_create_response_max_tokens_parameter(client: TestClient) -> None:
|
||||
"""Test max_output_tokens parameter handling."""
|
||||
request_data = {
|
||||
"model": "lorem-tester",
|
||||
"input": [{"role": "user", "content": "Test"}],
|
||||
"max_output_tokens": 100,
|
||||
"stream": False
|
||||
}
|
||||
|
||||
response = client.post("/v1/responses", json=request_data)
|
||||
|
||||
assert response.status_code == 200
|
||||
Reference in New Issue
Block a user