Semantic Code Clone Detection for Enterprise Applications

This project detect semantic code clones in enterprise applications.

Please see the demonstration video

Prerequisites

• Maven 3.6+
• Java 11+

Reproducing results

Running from the source:

Change directory in the console into the cloned project.

Run on Linux:

./mvnw compile quarkus:dev -Dquarkus.args='/pat/to/app/under/test,/path/to/output' -Ddebug=true

Run on Windows:

mvnw.cmd compile quarkus:dev -Dquarkus.args='/pat/to/app/under/test,/path/to/output' -Ddebug=true

Run on MacOS:

sh ./mvnw compile quarkus:dev -Dquarkus.args='/pat/to/app/under/test,/path/to/output' -Ddebug=true

There are 2 arguments specified in -dDquarkus.args separated by a comma.

First is a path to the application that we want to analyze. We utilized project Train Ticket that can be accessed here: https://github.com/FudanSELab/train-ticket

Second argument is a path to a directory where the output of the application is going to be stored.

Note: do not set second path inside this project, it will not compile

Results

Results are stored under /results directory.

It contains:

• module-pair-clones.txt: data on how the microservices are similar to the others
• per module clones.txt: data on how many CFGs are contained within the microservice and what percentage of those CFGs are clones