def main(program_slice, vulnerability_pattern, debug=False):
program_slice_json = read_from_json(program_slice)
vuln_pattern_json = read_from_json(vulnerability_pattern)
output_filename = program_slice[:-5] + ".output.json"
output = program_analysis(program_slice_json,
vuln_pattern_json,
debug=debug)
write_to_json(output, output_filename)
def test_slice1():
program_slice = "{}/data/slice1.json".format(ROOT_DIR)
pattern = "{}/tests/test_slice1/slice1_vuln_pattern.json".format(ROOT_DIR)
analyzer(program_slice, pattern, debug=True)
output = read_from_json("{}/data/slice1.output.json".format(ROOT_DIR))
assert output == [{
"vulnerability": "SQL injection",
"source": "a",
"sink": "z",
"sanitizer": "",
}]
def test_slice4():
program_slice = "{}/data/slice4.json".format(ROOT_DIR)
pattern = "{}/tests/test_slice4/slice4_vuln_pattern.json".format(ROOT_DIR)
analyzer(program_slice, pattern, debug=False)
output = read_from_json("{}/data/slice4.output.json".format(ROOT_DIR))
assert output == [{
"vulnerability": "SQL injection",
"source": "koneksi",
"sink": "s",
"sanitizer": "",
}]
y_column_name,
step_size=step_size)
best_score, best_number_features, best_feature_sets = get_scores_from_feature_sub_sets(
df, features_in_sub_sets, y_column_name)
return best_score, best_number_features, data[best_feature_sets +
[y_column_name]]
if __name__ == "__main__":
data: pd.DataFrame = pd.read_csv(
"{}/data/df_without_corr.csv".format(ROOT_DIR))
find_best_feature_sub_sets(data,
y_column_name="class",
step_size=1,
verbose=True,
save_to_file=True)
feature_sets = read_from_json("feature_sets")
print(
get_scores_from_feature_sub_sets(
data,
feature_sets=feature_sets,
y_column_name="class",
classifier=knn_model,
save_to_file=True,
))
feature_sub_set_scores = read_from_json("feature_sub_set_scores")
plots_feature_sub_set_scores(feature_sub_set_scores)
