File size: 9,483 Bytes
246d201 |
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 |
from unittest.mock import MagicMock, patch
import pytest
import requests
from litellm.exceptions import RateLimitError
from openhands.core.config import LLMConfig
from openhands.events.action.message import MessageAction
from openhands.llm.llm import LLM
from openhands.resolver.github_issue import GithubIssue
from openhands.resolver.issue_definitions import IssueHandler, PRHandler
@pytest.fixture(autouse=True)
def mock_logger(monkeypatch):
# suppress logging of completion data to file
mock_logger = MagicMock()
monkeypatch.setattr('openhands.llm.debug_mixin.llm_prompt_logger', mock_logger)
monkeypatch.setattr('openhands.llm.debug_mixin.llm_response_logger', mock_logger)
return mock_logger
@pytest.fixture
def default_config():
return LLMConfig(
model='gpt-4o',
api_key='test_key',
num_retries=2,
retry_min_wait=1,
retry_max_wait=2,
)
def test_handle_nonexistent_issue_reference():
llm_config = LLMConfig(model='test', api_key='test')
handler = PRHandler('test-owner', 'test-repo', 'test-token', llm_config)
# Mock the requests.get to simulate a 404 error
mock_response = MagicMock()
mock_response.raise_for_status.side_effect = requests.exceptions.HTTPError(
'404 Client Error: Not Found'
)
with patch('requests.get', return_value=mock_response):
# Call the method with a non-existent issue reference
result = handler._PRHandler__get_context_from_external_issues_references(
closing_issues=[],
closing_issue_numbers=[],
issue_body='This references #999999', # Non-existent issue
review_comments=[],
review_threads=[],
thread_comments=None,
)
# The method should return an empty list since the referenced issue couldn't be fetched
assert result == []
def test_handle_rate_limit_error():
llm_config = LLMConfig(model='test', api_key='test')
handler = PRHandler('test-owner', 'test-repo', 'test-token', llm_config)
# Mock the requests.get to simulate a rate limit error
mock_response = MagicMock()
mock_response.raise_for_status.side_effect = requests.exceptions.HTTPError(
'403 Client Error: Rate Limit Exceeded'
)
with patch('requests.get', return_value=mock_response):
# Call the method with an issue reference
result = handler._PRHandler__get_context_from_external_issues_references(
closing_issues=[],
closing_issue_numbers=[],
issue_body='This references #123',
review_comments=[],
review_threads=[],
thread_comments=None,
)
# The method should return an empty list since the request was rate limited
assert result == []
def test_handle_network_error():
llm_config = LLMConfig(model='test', api_key='test')
handler = PRHandler('test-owner', 'test-repo', 'test-token', llm_config)
# Mock the requests.get to simulate a network error
with patch(
'requests.get', side_effect=requests.exceptions.ConnectionError('Network Error')
):
# Call the method with an issue reference
result = handler._PRHandler__get_context_from_external_issues_references(
closing_issues=[],
closing_issue_numbers=[],
issue_body='This references #123',
review_comments=[],
review_threads=[],
thread_comments=None,
)
# The method should return an empty list since the network request failed
assert result == []
def test_successful_issue_reference():
llm_config = LLMConfig(model='test', api_key='test')
handler = PRHandler('test-owner', 'test-repo', 'test-token', llm_config)
# Mock a successful response
mock_response = MagicMock()
mock_response.raise_for_status.return_value = None
mock_response.json.return_value = {'body': 'This is the referenced issue body'}
with patch('requests.get', return_value=mock_response):
# Call the method with an issue reference
result = handler._PRHandler__get_context_from_external_issues_references(
closing_issues=[],
closing_issue_numbers=[],
issue_body='This references #123',
review_comments=[],
review_threads=[],
thread_comments=None,
)
# The method should return a list with the referenced issue body
assert result == ['This is the referenced issue body']
class MockLLMResponse:
"""Mock LLM Response class to mimic the actual LLM response structure."""
class Choice:
class Message:
def __init__(self, content):
self.content = content
def __init__(self, content):
self.message = self.Message(content)
def __init__(self, content):
self.choices = [self.Choice(content)]
class DotDict(dict):
"""
A dictionary that supports dot notation access.
"""
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
for key, value in self.items():
if isinstance(value, dict):
self[key] = DotDict(value)
elif isinstance(value, list):
self[key] = [
DotDict(item) if isinstance(item, dict) else item for item in value
]
def __getattr__(self, key):
if key in self:
return self[key]
else:
raise AttributeError(
f"'{self.__class__.__name__}' object has no attribute '{key}'"
)
def __setattr__(self, key, value):
self[key] = value
def __delattr__(self, key):
if key in self:
del self[key]
else:
raise AttributeError(
f"'{self.__class__.__name__}' object has no attribute '{key}'"
)
@patch('openhands.llm.llm.litellm_completion')
def test_guess_success_rate_limit_wait_time(mock_litellm_completion, default_config):
"""Test that the retry mechanism in guess_success respects wait time between retries."""
with patch('time.sleep') as mock_sleep:
# Simulate a rate limit error followed by a successful response
mock_litellm_completion.side_effect = [
RateLimitError(
'Rate limit exceeded', llm_provider='test_provider', model='test_model'
),
DotDict(
{
'choices': [
{
'message': {
'content': '--- success\ntrue\n--- explanation\nRetry successful'
}
}
]
}
),
]
llm = LLM(config=default_config)
handler = IssueHandler('test-owner', 'test-repo', 'test-token', default_config)
handler.llm = llm
# Mock issue and history
issue = GithubIssue(
owner='test-owner',
repo='test-repo',
number=1,
title='Test Issue',
body='This is a test issue.',
thread_comments=['Please improve error handling'],
)
history = [MessageAction(content='Fixed error handling.')]
# Call guess_success
success, _, explanation = handler.guess_success(issue, history)
# Assertions
assert success is True
assert explanation == 'Retry successful'
assert mock_litellm_completion.call_count == 2 # Two attempts made
mock_sleep.assert_called_once() # Sleep called once between retries
# Validate wait time
wait_time = mock_sleep.call_args[0][0]
assert (
default_config.retry_min_wait <= wait_time <= default_config.retry_max_wait
), f'Expected wait time between {default_config.retry_min_wait} and {default_config.retry_max_wait} seconds, but got {wait_time}'
@patch('openhands.llm.llm.litellm_completion')
def test_guess_success_exhausts_retries(mock_completion, default_config):
"""Test the retry mechanism in guess_success exhausts retries and raises an error."""
# Simulate persistent rate limit errors by always raising RateLimitError
mock_completion.side_effect = RateLimitError(
'Rate limit exceeded', llm_provider='test_provider', model='test_model'
)
# Initialize LLM and handler
llm = LLM(config=default_config)
handler = PRHandler('test-owner', 'test-repo', 'test-token', default_config)
handler.llm = llm
# Mock issue and history
issue = GithubIssue(
owner='test-owner',
repo='test-repo',
number=1,
title='Test Issue',
body='This is a test issue.',
thread_comments=['Please improve error handling'],
)
history = [MessageAction(content='Fixed error handling.')]
# Call guess_success and expect it to raise an error after retries
with pytest.raises(RateLimitError):
handler.guess_success(issue, history)
# Assertions
assert (
mock_completion.call_count == default_config.num_retries
) # Initial call + retries
|