Spaces:
Sleeping
Sleeping
gary-boon
Claude
commited on
Commit
·
9dbec03
1
Parent(s):
c0d95bf
Fix syntax error in swe_bench_service.py
Browse files- Remove commented-out mock data function with nested triple quotes
- Triple quotes inside comments were causing Python syntax errors
- Clean removal ensures no mock data and fixes deployment
🤖 Generated with Claude Code
Co-Authored-By: Claude <noreply@anthropic.com>
- backend/swe_bench_service.py +2 -67
backend/swe_bench_service.py
CHANGED
|
@@ -95,73 +95,8 @@ class SWEBenchService:
|
|
| 95 |
self.metrics_cache: Dict[str, Any] = {}
|
| 96 |
|
| 97 |
# Removed _load_mock_tasks - real data only for research
|
| 98 |
-
|
| 99 |
-
|
| 100 |
-
# Load mock tasks when dataset isn't available
|
| 101 |
-
repos = [
|
| 102 |
-
"astropy/astropy", "django/django", "matplotlib/matplotlib",
|
| 103 |
-
"pandas-dev/pandas", "pytest-dev/pytest", "scikit-learn/scikit-learn"
|
| 104 |
-
]
|
| 105 |
-
|
| 106 |
-
statements = [
|
| 107 |
-
"""Modeling's `separability_matrix` does not compute separability correctly for nested CompoundModels
|
| 108 |
-
|
| 109 |
-
Consider the following model:
|
| 110 |
-
|
| 111 |
-
```python
|
| 112 |
-
from astropy.modeling import models as m
|
| 113 |
-
from astropy.modeling.separable import separable_matrix
|
| 114 |
-
|
| 115 |
-
cm = m.Linear1D(10) & m.Linear1D(5)
|
| 116 |
-
```
|
| 117 |
-
|
| 118 |
-
It's separability matrix as you might expect is a diagonal:
|
| 119 |
-
|
| 120 |
-
```python
|
| 121 |
-
>>> separability_matrix(cm)
|
| 122 |
-
array([[ True, False],
|
| 123 |
-
[False, True]])
|
| 124 |
-
```""",
|
| 125 |
-
"""Please support header rows in RestructuredText output
|
| 126 |
-
|
| 127 |
-
### Description
|
| 128 |
-
It would be great if the RestructuredText output could have header rows for tables, similar to what MySQL does for pipe formatting.
|
| 129 |
-
|
| 130 |
-
### Expected behavior
|
| 131 |
-
According to the documentation for MyST parsers, the docutils RST table expects the first row to be treated as a header row.
|
| 132 |
-
|
| 133 |
-
### Actual behavior
|
| 134 |
-
The RST output treats the first row as a regular data row and doesn't mark it as a header.""",
|
| 135 |
-
"""Issue when parsing empty lists/arrays in configuration
|
| 136 |
-
|
| 137 |
-
When attempting to parse empty lists or arrays from configuration files, the parser incorrectly raises a ValueError instead of returning an empty list.
|
| 138 |
-
|
| 139 |
-
```python
|
| 140 |
-
>>> config.parse_list("[]")
|
| 141 |
-
ValueError: invalid literal for int() with base 10: '[]'
|
| 142 |
-
```
|
| 143 |
-
|
| 144 |
-
Expected behavior: Should return an empty list []"""
|
| 145 |
-
]
|
| 146 |
-
|
| 147 |
-
for i in range(300): # Create 300 mock tasks for better testing
|
| 148 |
-
repo = repos[i % len(repos)]
|
| 149 |
-
repo_name = repo.split('/')[1]
|
| 150 |
-
issue_number = 11000 + i
|
| 151 |
-
|
| 152 |
-
task = SWEBenchTask(
|
| 153 |
-
instance_id=f"{repo_name}__{repo_name}-{issue_number}",
|
| 154 |
-
repo=repo,
|
| 155 |
-
problem_statement=statements[i % len(statements)],
|
| 156 |
-
base_commit=f"commit_{i:04d}",
|
| 157 |
-
patch="# Mock patch\n+ line added\n- line removed",
|
| 158 |
-
FAIL_TO_PASS=["test_1", "test_2"] if i % 2 == 0 else ["test_a"],
|
| 159 |
-
PASS_TO_PASS=["test_pass_1", "test_pass_2"]
|
| 160 |
-
)
|
| 161 |
-
self.tasks[task.instance_id] = task
|
| 162 |
-
|
| 163 |
-
logger.info(f"Loaded {len(self.tasks)} mock SWE-bench tasks")
|
| 164 |
-
"""
|
| 165 |
|
| 166 |
async def load_dataset(self, dataset_name: str = "princeton-nlp/SWE-bench_Lite"):
|
| 167 |
"""Load SWE-bench dataset from Hugging Face"""
|
|
|
|
| 95 |
self.metrics_cache: Dict[str, Any] = {}
|
| 96 |
|
| 97 |
# Removed _load_mock_tasks - real data only for research
|
| 98 |
+
# Mock data generation has been completely removed to ensure
|
| 99 |
+
# only real SWE-bench tasks are used for PhD research integrity
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 100 |
|
| 101 |
async def load_dataset(self, dataset_name: str = "princeton-nlp/SWE-bench_Lite"):
|
| 102 |
"""Load SWE-bench dataset from Hugging Face"""
|