def generate():
total_count = 0
syntax_valid_count = 0
files = []
for root, d_names, f_names in os.walk(seed_path):
for f in f_names:
files.append(os.path.join(root, f))
for file in files:
if 'deepfuzz' in file:
continue
if not file.endswith('.kt'):
continue
#try:
text = open(file, 'r').read()
# text = pp.replace_macro(text, file)
text = pp.remove_comment(text)
#text = pp.remove_space(text)
is_valid = pp.verify_correctness(text, file, 'deepfuzz_original')
if not is_valid:
continue
total_count += 1
text = synthesis(text, 'g3', 'sample')
is_valid = pp.verify_correctness(text, file, 'deepfuzz_g1_nosample')
if (is_valid):
syntax_valid_count += 1
#except:
#continue
pass_rate = syntax_valid_count / total_count
print('syntax_valid_count: %d' % syntax_valid_count)
print('total_count: %d' % total_count)
print('pass rate: %f' % pass_rate)
print(sentences[i] + "\t" + next_chars[i])
path = './testData1'
files = []
valid_count = 0
for root, d_names, f_names in os.walk(path):
try:
for f in f_names:
files.append(os.path.join(root, f))
except Exception:
continue
for file in files:
try:
print('--------------------------------------------------')
print(file)
# if 'nocomment' in file or 'nospace' in file or 'nomacro' in file or 'raw' in file:
# command = 'rm ' + file
# p = Popen(command, shell=True, stdin=PIPE, stdout=PIPE, stderr=STDOUT, close_fds=True)
text = open(file, 'r').read()
text = pp.remove_comment(text)
# text = pp.replace_macro(text, file)
text = pp.remove_space(text)
is_valid = pp.verify_correctness(text, file, 'nospace')
if is_valid:
valid_count += 1
generate_training_data(text)
except Exception:
continue
print(valid_count)