Spaces:
Configuration error
Configuration error
File size: 14,865 Bytes
447ebeb |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 |
import copy
import uuid
from unittest.mock import AsyncMock, MagicMock
import pytest
from fastapi import Request, status
from fastapi.responses import StreamingResponse
import litellm
from litellm.integrations.opentelemetry import UserAPIKeyAuth
from litellm.proxy.common_request_processing import (
ProxyBaseLLMRequestProcessing,
ProxyConfig,
_parse_event_data_for_error,
create_streaming_response,
)
from litellm.proxy.utils import ProxyLogging
class TestProxyBaseLLMRequestProcessing:
@pytest.mark.asyncio
async def test_common_processing_pre_call_logic_pre_call_hook_receives_litellm_call_id(
self, monkeypatch
):
processing_obj = ProxyBaseLLMRequestProcessing(data={})
mock_request = MagicMock(spec=Request)
mock_request.headers = {}
async def mock_add_litellm_data_to_request(*args, **kwargs):
return {}
async def mock_common_processing_pre_call_logic(
user_api_key_dict, data, call_type
):
data_copy = copy.deepcopy(data)
return data_copy
mock_proxy_logging_obj = MagicMock(spec=ProxyLogging)
mock_proxy_logging_obj.pre_call_hook = AsyncMock(
side_effect=mock_common_processing_pre_call_logic
)
monkeypatch.setattr(
litellm.proxy.common_request_processing,
"add_litellm_data_to_request",
mock_add_litellm_data_to_request,
)
mock_general_settings = {}
mock_user_api_key_dict = MagicMock(spec=UserAPIKeyAuth)
mock_proxy_config = MagicMock(spec=ProxyConfig)
route_type = "acompletion"
# Call the actual method.
(
returned_data,
logging_obj,
) = await processing_obj.common_processing_pre_call_logic(
request=mock_request,
general_settings=mock_general_settings,
user_api_key_dict=mock_user_api_key_dict,
proxy_logging_obj=mock_proxy_logging_obj,
proxy_config=mock_proxy_config,
route_type=route_type,
)
mock_proxy_logging_obj.pre_call_hook.assert_called_once()
_, call_kwargs = mock_proxy_logging_obj.pre_call_hook.call_args
data_passed = call_kwargs.get("data", {})
assert "litellm_call_id" in data_passed
try:
uuid.UUID(data_passed["litellm_call_id"])
except ValueError:
pytest.fail("litellm_call_id is not a valid UUID")
assert data_passed["litellm_call_id"] == returned_data["litellm_call_id"]
@pytest.mark.asyncio
class TestCommonRequestProcessingHelpers:
async def consume_stream(self, streaming_response: StreamingResponse) -> list:
content = []
async for chunk_bytes in streaming_response.body_iterator:
content.append(chunk_bytes)
return content
@pytest.mark.parametrize(
"event_line, expected_code",
[
(
'data: {"error": {"code": 400, "message": "bad request"}}',
400,
), # Valid integer code
(
'data: {"error": {"code": "401", "message": "unauthorized"}}',
401,
), # Valid string-integer code
(
'data: {"error": {"code": "invalid_code", "message": "error"}}',
None,
), # Invalid string code
(
'data: {"error": {"code": 99, "message": "too low"}}',
None,
), # Integer code too low
(
'data: {"error": {"code": 600, "message": "too high"}}',
None,
), # Integer code too high
(
'data: {"id": "123", "content": "hello"}',
None,
), # Non-error SSE event
("data: [DONE]", None), # SSE [DONE] event
("data: ", None), # SSE empty data event
(
'data: {"error": {"code": 400',
None,
), # Malformed JSON
("id: 123", None), # Non-SSE event line
(
'data: {"error": {"message": "some error"}}',
None,
), # Error event without 'code' field
(
'data: {"error": {"code": null, "message": "code is null"}}',
None,
), # Error with null code
],
)
async def test_parse_event_data_for_error(self, event_line, expected_code):
assert await _parse_event_data_for_error(event_line) == expected_code
async def test_create_streaming_response_first_chunk_is_error(self):
async def mock_generator():
yield 'data: {"error": {"code": 403, "message": "forbidden"}}\n\n'
yield 'data: {"content": "more data"}\n\n'
yield "data: [DONE]\n\n"
response = await create_streaming_response(
mock_generator(), "text/event-stream", {}
)
assert response.status_code == status.HTTP_403_FORBIDDEN
content = await self.consume_stream(response)
assert content == [
'data: {"error": {"code": 403, "message": "forbidden"}}\n\n',
'data: {"content": "more data"}\n\n',
"data: [DONE]\n\n",
]
async def test_create_streaming_response_first_chunk_not_error(self):
async def mock_generator():
yield 'data: {"content": "first part"}\n\n'
yield 'data: {"content": "second part"}\n\n'
yield "data: [DONE]\n\n"
response = await create_streaming_response(
mock_generator(), "text/event-stream", {}
)
assert response.status_code == status.HTTP_200_OK
content = await self.consume_stream(response)
assert content == [
'data: {"content": "first part"}\n\n',
'data: {"content": "second part"}\n\n',
"data: [DONE]\n\n",
]
async def test_create_streaming_response_empty_generator(self):
async def mock_generator():
if False: # Never yields
yield
# Implicitly raises StopAsyncIteration
response = await create_streaming_response(
mock_generator(), "text/event-stream", {}
)
assert response.status_code == status.HTTP_200_OK
content = await self.consume_stream(response)
assert content == []
async def test_create_streaming_response_generator_raises_stop_async_iteration_immediately(
self,
):
mock_gen = AsyncMock()
mock_gen.__anext__.side_effect = StopAsyncIteration
response = await create_streaming_response(mock_gen, "text/event-stream", {})
assert response.status_code == status.HTTP_200_OK
content = await self.consume_stream(response)
assert content == []
async def test_create_streaming_response_generator_raises_unexpected_exception(
self,
):
mock_gen = AsyncMock()
mock_gen.__anext__.side_effect = ValueError("Test error from generator")
response = await create_streaming_response(mock_gen, "text/event-stream", {})
assert response.status_code == status.HTTP_500_INTERNAL_SERVER_ERROR
content = await self.consume_stream(response)
expected_error_data = {
"error": {
"message": "Error processing stream start",
"code": status.HTTP_500_INTERNAL_SERVER_ERROR,
}
}
assert len(content) == 2
# Use json.dumps to match the formatting in create_streaming_response's exception handler
import json
assert content[0] == f"data: {json.dumps(expected_error_data)}\n\n"
assert content[1] == "data: [DONE]\n\n"
async def test_create_streaming_response_first_chunk_error_string_code(self):
async def mock_generator():
yield 'data: {"error": {"code": "429", "message": "too many requests"}}\n\n'
yield "data: [DONE]\n\n"
response = await create_streaming_response(
mock_generator(), "text/event-stream", {}
)
assert response.status_code == status.HTTP_429_TOO_MANY_REQUESTS
content = await self.consume_stream(response)
assert content == [
'data: {"error": {"code": "429", "message": "too many requests"}}\n\n',
"data: [DONE]\n\n",
]
async def test_create_streaming_response_custom_headers(self):
async def mock_generator():
yield 'data: {"content": "data"}\n\n'
yield "data: [DONE]\n\n"
custom_headers = {"X-Custom-Header": "TestValue"}
response = await create_streaming_response(
mock_generator(), "text/event-stream", custom_headers
)
assert response.headers["x-custom-header"] == "TestValue"
async def test_create_streaming_response_non_default_status_code(self):
async def mock_generator():
yield 'data: {"content": "data"}\n\n'
yield "data: [DONE]\n\n"
response = await create_streaming_response(
mock_generator(),
"text/event-stream",
{},
default_status_code=status.HTTP_201_CREATED,
)
assert response.status_code == status.HTTP_201_CREATED
content = await self.consume_stream(response)
assert content == [
'data: {"content": "data"}\n\n',
"data: [DONE]\n\n",
]
async def test_create_streaming_response_first_chunk_is_done(self):
async def mock_generator():
yield "data: [DONE]\n\n"
response = await create_streaming_response(
mock_generator(), "text/event-stream", {}
)
assert response.status_code == status.HTTP_200_OK # Default status
content = await self.consume_stream(response)
assert content == ["data: [DONE]\n\n"]
async def test_create_streaming_response_first_chunk_is_empty_data(self):
async def mock_generator():
yield "data: \n\n"
yield 'data: {"content": "actual data"}\n\n'
yield "data: [DONE]\n\n"
response = await create_streaming_response(
mock_generator(), "text/event-stream", {}
)
assert response.status_code == status.HTTP_200_OK # Default status
content = await self.consume_stream(response)
assert content == [
"data: \n\n",
'data: {"content": "actual data"}\n\n',
"data: [DONE]\n\n",
]
async def test_create_streaming_response_all_chunks_have_dd_trace(self):
"""Test that all stream chunks are wrapped with dd trace at the streaming generator level"""
import json
from unittest.mock import patch
# Create a mock tracer
mock_tracer = MagicMock()
mock_span = MagicMock()
mock_tracer.trace.return_value.__enter__.return_value = mock_span
mock_tracer.trace.return_value.__exit__.return_value = None
# Mock generator with multiple chunks
async def mock_generator():
yield 'data: {"content": "chunk 1"}\n\n'
yield 'data: {"content": "chunk 2"}\n\n'
yield 'data: {"content": "chunk 3"}\n\n'
yield "data: [DONE]\n\n"
# Patch the tracer in the common_request_processing module
with patch("litellm.proxy.common_request_processing.tracer", mock_tracer):
response = await create_streaming_response(
mock_generator(), "text/event-stream", {}
)
assert response.status_code == 200
# Consume the stream to trigger the tracer calls
content = await self.consume_stream(response)
# Verify all chunks are present
assert len(content) == 4
assert content[0] == 'data: {"content": "chunk 1"}\n\n'
assert content[1] == 'data: {"content": "chunk 2"}\n\n'
assert content[2] == 'data: {"content": "chunk 3"}\n\n'
assert content[3] == "data: [DONE]\n\n"
# Verify that tracer.trace was called for each chunk (4 chunks total)
assert mock_tracer.trace.call_count == 4
# Verify that each call was made with the correct operation name
expected_calls = [
(("streaming.chunk.yield",), {}),
(("streaming.chunk.yield",), {}),
(("streaming.chunk.yield",), {}),
(("streaming.chunk.yield",), {}),
]
actual_calls = mock_tracer.trace.call_args_list
assert len(actual_calls) == 4
for i, call in enumerate(actual_calls):
args, kwargs = call
assert (
args[0] == "streaming.chunk.yield"
), f"Call {i} should have operation name 'streaming.chunk.yield', got {args[0]}"
async def test_create_streaming_response_dd_trace_with_error_chunk(self):
"""Test that dd trace is applied even when the first chunk contains an error"""
from unittest.mock import patch
# Create a mock tracer
mock_tracer = MagicMock()
mock_span = MagicMock()
mock_tracer.trace.return_value.__enter__.return_value = mock_span
mock_tracer.trace.return_value.__exit__.return_value = None
# Mock generator with error in first chunk
async def mock_generator():
yield 'data: {"error": {"code": 400, "message": "bad request"}}\n\n'
yield 'data: {"content": "chunk after error"}\n\n'
yield "data: [DONE]\n\n"
# Patch the tracer in the common_request_processing module
with patch("litellm.proxy.common_request_processing.tracer", mock_tracer):
response = await create_streaming_response(
mock_generator(), "text/event-stream", {}
)
# Even with error, status should be set to error code but tracing should still work
assert response.status_code == 400
# Consume the stream to trigger the tracer calls
content = await self.consume_stream(response)
# Verify all chunks are present
assert len(content) == 3
# Verify that tracer.trace was called for each chunk
assert mock_tracer.trace.call_count == 3
# Verify that each call was made with the correct operation name
actual_calls = mock_tracer.trace.call_args_list
assert len(actual_calls) == 3
for i, call in enumerate(actual_calls):
args, kwargs = call
assert (
args[0] == "streaming.chunk.yield"
), f"Call {i} should have operation name 'streaming.chunk.yield', got {args[0]}"
|