Implements Phase 5: OpenAI Chat Completions compatibility layer Features: - Wraps Responses API for single source of truth - Automatically enables reasoning generation - Converts reasoning items to <think> tags for Open WebUI - Maintains OpenAI-compatible chat completion format - Supports both streaming and non-streaming modes - Pipeline prefix preservation for model names - System message handling Architecture: - Service layer calls Responses API internally - Streams word-by-word for smooth UX - Reasoning displayed in thought bubbles (Open WebUI) - Main response shown separately from thinking Error Handling: - Enhanced exception types (RateLimitError, ContextLengthError) - OpenAI-compatible error format - Graceful error propagation from Responses API Testing: - 6 unit tests for chat router functionality - 6 unit tests for streaming wrapper behavior - Total: 12 tests with comprehensive coverage 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
174 lines
5.5 KiB
Python
174 lines
5.5 KiB
Python
"""
|
|
Tests for chat completions router.
|
|
"""
|
|
import json
|
|
|
|
import pytest
|
|
from fastapi.testclient import TestClient
|
|
from httpx import AsyncClient
|
|
|
|
from src.chat import constants
|
|
|
|
|
|
@pytest.mark.unit
|
|
def test_chat_completion_non_streaming(
|
|
client: TestClient,
|
|
mock_chat_request: dict,
|
|
) -> None:
|
|
"""Test non-streaming chat completion."""
|
|
response = client.post("/v1/chat/completions", json=mock_chat_request)
|
|
|
|
assert response.status_code == 200
|
|
data = response.json()
|
|
|
|
# Verify response structure
|
|
assert "id" in data
|
|
assert data["object"] == constants.CHAT_COMPLETION_OBJECT
|
|
assert "created" in data
|
|
assert data["model"] == mock_chat_request["model"]
|
|
assert len(data["choices"]) == 1
|
|
|
|
# Verify choice structure
|
|
choice = data["choices"][0]
|
|
assert choice["index"] == 0
|
|
assert choice["message"]["role"] == constants.ROLE_ASSISTANT
|
|
assert choice["message"]["content"] # Should have content
|
|
assert choice["finish_reason"] == constants.FINISH_REASON_STOP
|
|
|
|
# Verify usage
|
|
assert "usage" in data
|
|
assert data["usage"]["prompt_tokens"] > 0
|
|
assert data["usage"]["completion_tokens"] > 0
|
|
assert data["usage"]["total_tokens"] > 0
|
|
|
|
|
|
@pytest.mark.unit
|
|
def test_chat_completion_validation_error(client: TestClient) -> None:
|
|
"""Test chat completion with invalid request."""
|
|
# Missing required field 'messages'
|
|
invalid_request = {"model": "tatlock"}
|
|
|
|
response = client.post("/v1/chat/completions", json=invalid_request)
|
|
|
|
assert response.status_code == 422
|
|
data = response.json()
|
|
assert "error" in data
|
|
assert data["error"]["type"] == "invalid_request_error"
|
|
|
|
|
|
@pytest.mark.unit
|
|
async def test_chat_completion_streaming(
|
|
async_client: AsyncClient,
|
|
mock_streaming_chat_request: dict,
|
|
) -> None:
|
|
"""Test streaming chat completion with 20s timeout per turn."""
|
|
import asyncio
|
|
|
|
async def read_stream_with_timeout():
|
|
"""Read SSE stream with timeout protection."""
|
|
chunks = []
|
|
|
|
async with async_client.stream(
|
|
"POST",
|
|
"/v1/chat/completions",
|
|
json=mock_streaming_chat_request,
|
|
timeout=20.0, # 20 second timeout per turn
|
|
) as response:
|
|
assert response.status_code == 200
|
|
assert response.headers["content-type"] == "text/event-stream; charset=utf-8"
|
|
|
|
async for line in response.aiter_lines():
|
|
# Skip empty lines
|
|
if not line.strip():
|
|
continue
|
|
|
|
# Parse SSE format: "data: {json}"
|
|
if line.startswith("data: "):
|
|
data_str = line[6:].strip() # Remove "data: " prefix
|
|
|
|
# Check for [DONE] marker
|
|
if data_str == "[DONE]":
|
|
break
|
|
|
|
# Parse JSON chunk
|
|
try:
|
|
chunk = json.loads(data_str)
|
|
chunks.append(chunk)
|
|
|
|
# Verify chunk structure
|
|
assert chunk["object"] == constants.CHAT_COMPLETION_CHUNK_OBJECT
|
|
assert chunk["model"] == mock_streaming_chat_request["model"]
|
|
assert len(chunk["choices"]) == 1
|
|
|
|
except json.JSONDecodeError as e:
|
|
pytest.fail(f"Invalid JSON in stream: {data_str[:100]}... Error: {e}")
|
|
|
|
return chunks
|
|
|
|
# Execute with overall timeout
|
|
try:
|
|
chunks = await asyncio.wait_for(read_stream_with_timeout(), timeout=20.0)
|
|
except asyncio.TimeoutError:
|
|
pytest.fail("Streaming test timed out after 20 seconds")
|
|
|
|
# Verify we got chunks
|
|
assert len(chunks) > 0, "No chunks received from stream"
|
|
|
|
# First chunk should have role
|
|
first_chunk = chunks[0]
|
|
assert first_chunk["choices"][0]["delta"].get("role") == constants.ROLE_ASSISTANT
|
|
|
|
# Last chunk should have finish_reason
|
|
last_chunk = chunks[-1]
|
|
assert last_chunk["choices"][0]["finish_reason"] == constants.FINISH_REASON_STOP
|
|
|
|
|
|
@pytest.mark.unit
|
|
def test_chat_completion_temperature_validation(
|
|
client: TestClient,
|
|
mock_chat_request: dict,
|
|
) -> None:
|
|
"""Test temperature parameter validation."""
|
|
# Invalid temperature (too high)
|
|
invalid_request = {**mock_chat_request, "temperature": 3.0}
|
|
response = client.post("/v1/chat/completions", json=invalid_request)
|
|
assert response.status_code == 422
|
|
|
|
# Valid temperature
|
|
valid_request = {**mock_chat_request, "temperature": 0.5}
|
|
response = client.post("/v1/chat/completions", json=valid_request)
|
|
assert response.status_code == 200
|
|
|
|
|
|
@pytest.mark.unit
|
|
def test_chat_completion_message_roles(
|
|
client: TestClient,
|
|
mock_chat_request: dict,
|
|
) -> None:
|
|
"""Test different message roles."""
|
|
request_with_system = {
|
|
**mock_chat_request,
|
|
"messages": [
|
|
{"role": "system", "content": "You are a helpful assistant."},
|
|
{"role": "user", "content": "Hello!"},
|
|
],
|
|
}
|
|
|
|
response = client.post("/v1/chat/completions", json=request_with_system)
|
|
assert response.status_code == 200
|
|
|
|
|
|
@pytest.mark.unit
|
|
def test_chat_completion_invalid_role(
|
|
client: TestClient,
|
|
mock_chat_request: dict,
|
|
) -> None:
|
|
"""Test invalid message role."""
|
|
invalid_request = {
|
|
**mock_chat_request,
|
|
"messages": [{"role": "invalid_role", "content": "test"}],
|
|
}
|
|
|
|
response = client.post("/v1/chat/completions", json=invalid_request)
|
|
assert response.status_code == 422
|