xiaoven / ccstokener Goto Github PK

• ccstokener: clone detection tool
• 8projects: four java open source projects and four c open source projects.
• java-label-pairs: the files in pairs.zip are the sampled true pairs and false pairs from four java open source projects. astnn.zip and tbccd.zip are the clone pairs organized into the format used by ASTNN and TBCCD.

CCStokener runs under Python3 environment, and you should first install pandarallel package to run CCStokener. After performing clone detection, the results of clone pairs will be stored in results directory.

pip3 install pandarallel

Our tool, CCStokener, is in directory ccstokener, now it supports detection for c and java language. runner.py is the runner file.

python runner.py -i /path/to/dataset -m common/bcb -l java/c [-t 0.6]

• -i: path to dataset
• -m: detection mode. If you detect on BigCloneBench, the mode will be bcb, otherwise, the mode will be common.
• -l: language of datasets, now we support detection for java and c languages.
• -t: similarity threshold of clone pairs, it is optional. If it is not setting, we will use the default threshold.

Example: You can run CCStokener as follows to perform clone detection for BigCloneBench.

python runner.py -i /home/datasets/BigCloneEval-master/ijadataset/bcb_reduced -m bcb -l java

We support verify whether clone pairs are true or not. The clone pairs should edit in a file firstly, then call verify.py as follows to verify.

python verify.py -i /path/to/dataset -l java/c

• -i: path to the clone pair files directory.
• -l: language of clone pair, support java or c language.