[foozzie] Refactoring - move source hashing to main script

tools/foozzie/v8_foozzie.py

 #!/usr/bin/env python
 # Copyright 2016 the V8 project authors. All rights reserved.
 # Use of this source code is governed by a BSD-style license that can be
 # found in the LICENSE file.
 
 """
 V8 correctness fuzzer launcher script.
 """
 
 import argparse
+import hashlib
 import itertools
 import json
 import os
 import re
 import sys
 import traceback
 
 import v8_commands
 import v8_suppressions
 
@@ skipping 56 matching lines @@
 ### End of configuration %(first_config_label)s
 #
 ### Start of configuration %(second_config_label)s:
 %(second_config_output)s
 ### End of configuration %(second_config_label)s
 """
 
 FUZZ_TEST_RE = re.compile(r'.*fuzz(-\d+\.js)')
 SOURCE_RE = re.compile(r'print\("v8-foozzie source: (.*)"\);')
 
+# The number of hex digits used from the hash of the original source file path.
+# Keep the number small to avoid duplicate explosion.
+ORIGINAL_SOURCE_HASH_LENGTH = 3
+
+# Placeholder string if no original source file could be determined.
+ORIGINAL_SOURCE_DEFAULT = 'none'
+
 def parse_args():
   parser = argparse.ArgumentParser()
   parser.add_argument(
     '--random-seed', type=int, required=True,
     help='random seed passed to both runs')
   parser.add_argument(
       '--first-arch', help='first architecture', default='x64')
   parser.add_argument(
       '--second-arch', help='second architecture', default='x64')
   parser.add_argument(
@@ skipping 131 matching lines @@
     return RETURN_FAIL
 
   second_config_output = run_d8(options.second_d8, second_config_flags)
 
   # Bailout based on second run's output.
   if pass_bailout(second_config_output, 2):
     return RETURN_PASS
   if fail_bailout(second_config_output, suppress.ignore_by_output2):
     return RETURN_FAIL
 
-  difference, source, source_key = suppress.diff(
+  difference, source = suppress.diff(
       first_config_output.stdout, second_config_output.stdout)
+
+  if source:
+    source_key = hashlib.sha1(source).hexdigest()[:ORIGINAL_SOURCE_HASH_LENGTH]
+  else:
+    source = ORIGINAL_SOURCE_DEFAULT
+    source_key = ORIGINAL_SOURCE_DEFAULT
+
   if difference:
     # The first three entries will be parsed by clusterfuzz. Format changes
     # will require changes on the clusterfuzz side.
     first_config_label = '%s,%s' % (options.first_arch, options.first_config)
     second_config_label = '%s,%s' % (options.second_arch, options.second_config)
     print FAILURE_TEMPLATE % dict(
         configs='%s:%s' % (first_config_label, second_config_label),
         source_key=source_key,
         suppression='', # We can't tie bugs to differences.
         first_config_label=first_config_label,
@@ skipping 25 matching lines @@
         configs='', source_key='', suppression='wrong_usage')
     result = RETURN_FAIL
   except Exception as e:
     print FAILURE_HEADER_TEMPLATE % dict(
         configs='', source_key='', suppression='internal_error')
     print '# Internal error: %s' % e
     traceback.print_exc(file=sys.stdout)
     result = RETURN_FAIL
 
   sys.exit(result)