Fix Logging and Error Handling for get_benchmark_final_target_code (#910)

## Description
This PR addresses the issue of missing source code by adding enhanced
logging and improving error handling in the
`get_benchmark_final_target_code()` method. Previously, silent failures
could occur, making it difficult to diagnose the root cause of missing
source code.

## Changes
- **Enhanced Logging**: Logs are added to track when the `sample_id` is
invalid or the source code is missing.
- **Error Handling**: Ensures that malformed or empty `sample_id` values
are handled gracefully and logged appropriately.
- **Return Null for Missing Source Code**: In case of an error, the
function explicitly returns `None` rather than an empty string for
clarity in JSON outputs.
#908
It may also refer to #625
## Benefits
- Improved **debuggability** by making it easier to identify the reasons
behind missing source code.
- Increased **robustness** by preventing silent failures and ensuring
more informative logs and error handling.

Author: Ekam219

report/common.py

@@ -256,6 +256,32 @@ def getsize(self) -> int:
 class Results:
   """Results provides functions to explore the experiment results in a
   particular directory."""
+  
+  def get_benchmark_final_target_code(self, sample_id: str) -> Optional[str]:
+    """Retrieve source code for a sample_id (format: 'benchmark/sample').
+
+    Args:
+        sample_id: A string in the format 'benchmark/sample'.
+
+    Returns:
+        The source code as a string, or None if the source code is missing or the sample_id is invalid.
+    """
+    try:
+        # Split the sample_id into benchmark and sample
+        benchmark, sample = sample_id.split('/')
+    except ValueError:
+        # Log an error if the sample_id format is invalid
+        logging.error(f"Invalid sample_id format: '{sample_id}'. Expected 'benchmark/sample'.")
+        return None
+
+    # Retrieve the source code using the existing method
+    code = self.get_final_target_code(benchmark, sample)
+    if not code:
+        # Log a warning if the source code is missing
+        logging.warning(f"Missing source code for {sample_id}")
+        return None
+
+    return code
 
   def __init__(self, results_dir='results', benchmark_set='all'):
     self._results_dir = results_dir
@@ -284,7 +310,7 @@ def get_final_target_code(self, benchmark: str, sample: str) -> str:
     targets_dir = os.path.join(self._results_dir, benchmark, 'fixed_targets')
     # TODO(donggeliu): Make this consistent with agent output.
     if not os.path.exists(targets_dir):
-      return ''
+      return None
 
     for name in sorted(FileSystem(targets_dir).listdir()):
       path = os.path.join(targets_dir, name)
@@ -293,7 +319,7 @@ def get_final_target_code(self, benchmark: str, sample: str) -> str:
           code = f.read()
           code = json.dumps(code)
         return code
-    return ''
+    return None
 
   def get_logs(self, benchmark: str, sample: str) -> list[LogPart]:
     status_dir = os.path.join(self._results_dir, benchmark, 'status')

report/templates/crash.json

@@ -17,7 +17,7 @@
         "target_binary": "{{ sample.target_binary }}",
         "reproducer": "{{ sample.reproducer }}",
         "run_log": "{{ sample.run_log }}",
-        "source_code": {{ get_benchmark_final_target_code(sample.id) | replace('\\n', '\\\\n')}},
+        "source_code": {{ get_benchmark_final_target_code(sample.id) | tojson }},
         "model": "{{ model }}"
     }{% if not loop.last %},{% endif %}
 {% endfor %}

report/web.py

@@ -218,12 +218,13 @@ def _write_benchmark_index(self, benchmark: Benchmark, samples: List[Sample],
   def _write_benchmark_crash(self, benchmark: Benchmark, samples: List[Sample]):
     """Generate the benchmark crash.json and write to filesystem."""
     try:
-      rendered = self._jinja.render('crash.json',
-                                    benchmark=benchmark.signature,
-                                    samples=samples,
-                                    get_benchmark_final_target_code=partial(
-                                        self._results.get_final_target_code,
-                                        benchmark.id))
+      rendered = self._jinja.render(
+          'crash.json',
+          benchmark=benchmark.signature,
+          samples=samples,
+          # Changed line below to use new error-handling method
+          get_benchmark_final_target_code=self._results.get_benchmark_final_target_code
+      )
       self._write(f'benchmark/{benchmark.id}/crash.json', rendered)
     except Exception as e:
       logging.error('Failed to write benchmark/%s/crash.json:\n%s',
@@ -328,4 +329,4 @@ def main():
 
 if __name__ == '__main__':
   logging.getLogger().setLevel(os.environ.get('LOGLEVEL', 'WARN').upper())
-  main()
+  main()