Fix training data generator (#453)

Training data generator crashes in recent experiments because we added
triage directories (e.g.,
[03-triage/](https://pantheon.corp.google.com/storage/browser/oss-fuzz-gcb-experiment-run-logs/Result-reports/scheduled/2024-07-06-weekly-all/results/output-eigen-_zn5eigen9transposeins_5blockins_6matrixinst3__17complexideelin1elin1eli0elin1elin1eeelin1eli1elb1eeee6resi/fixed_targets/03-triage?pageState=(%22StorageObjectListTable%22:(%22f%22:%22%255B%255D%22))&e=-13802955&mods=logs_tg_prod&project=oss-fuzz))
and breaks this line:

https://github.com/google/oss-fuzz-gen/blob/main/data_prep/parse_training_data.py#L83

This PR fixes it with a more fine-grained regex to identify code fixing
directories (e.g.,
[03-F5/](https://pantheon.corp.google.com/storage/browser/oss-fuzz-gcb-experiment-run-logs/Result-reports/scheduled/2024-07-06-weekly-all/results/output-eigen-_zn5eigen9transposeins_5blockins_6matrixinst3__17complexideelin1elin1eli0elin1elin1eeelin1eli1elb1eeee6resi/fixed_targets/03-F5?pageState=(%22StorageObjectListTable%22:(%22f%22:%22%255B%255D%22))&e=-13802955&mods=logs_tg_prod&project=oss-fuzz)).

data_prep/parse_training_data.py

@@ -27,13 +27,15 @@
 import json
 import logging
 import os
+import re
 import sys
 from concurrent.futures import ThreadPoolExecutor
 from typing import Any, Dict, List
 
 from google.cloud import storage
 
 STORAGE_CLIENT = storage.Client()
+FUZZ_TARGET_FIXING_DIR_PATTERN = r'\d+-F\d+'
 
 
 class Benchmark:
@@ -53,6 +55,14 @@ def prompt(self) -> str:
     with open(prompt_path) as prompt_file:
       return prompt_file.read()
 
+  def _get_code_fixing_dirs(self, fixed_target_dir):
+    """Gets the directories for fixing fuzz targets."""
+    return [
+        item for item in os.listdir(fixed_target_dir)
+        if (os.path.isdir(os.path.join(fixed_target_dir, item)) and
+            re.match(FUZZ_TARGET_FIXING_DIR_PATTERN, item))
+    ]
+
   @property
   def targets(self) -> Dict[str, List[str]]:
     """Returns the generated targets of a benchmark in a directory, mapping
@@ -75,10 +85,7 @@ def targets(self) -> Dict[str, List[str]]:
     if not os.path.isdir(fixed_target_dir):
       logging.warning('Fixed target dir does not exist: %s', fixed_target_dir)
       return {}
-    fix_dirs = [
-        instance for instance in os.listdir(fixed_target_dir)
-        if os.path.isdir(os.path.join(fixed_target_dir, instance))
-    ]
+    fix_dirs = self._get_code_fixing_dirs(fixed_target_dir)
     for fix_dir in sorted(fix_dirs):
       instance, _ = fix_dir.split('-F')
       code_path = [