JavaRepoMiner is a python tool which can detect:
- Java method signature after and before a commit. In this repo, JavaRepoMiner mines a given Java GitHub code repository and analyses all the commits in that repository to find the commits that have added a parameter to an existing method. For example, assume that we have a method test(int x) in a Java file. If a commit changes this method to be test(int x, int y) then this commit added a parameter to this function and this project will detect that.
- Java class before and after commit,
- Java Library before and after commit,
- Java Loops before and after commit.
JavaRepoMiner makes a report in CSV file with:
- Columns of "Commit SHA, Java File, Before Commit Status, After Commit Status”.
- Frequency of Method, Class, Library, and Loop.
- Python 3.X
- Git
- To install git in Linux:
sudo apt-get install git- To install PyDriller[1]:
pip install pydriller-
RepoMiner.py This module wil take the Java code repository as input and write the output as described in a CSV file upder the Outputs folder.
-
RepoMiner method takes the git repo name as input and returns the results to write in CSV.
-
isMethod method checks the JAVA statements and ensures that statement is a method or not.
After running the RepoMiner.py module, it will prompt for input. The input is the JAVA code repository. For example, we have a repo named JavaTestingRepo in parent directory, then input will be like ../JavaTestingRepo. Then it write the result in this project in a folder name output as Result.csv.
To run the code:
$ git clone https://github.com/habibrahmanbd/JavaRepoMiner.git
$ python RepoMiner.py
Enter the git name with path: <RepoNameWithDirectory>
Output printed successfully...
Two popular JAVA git repo, java-design-patterns [2] and RxJava [3] is used as dataset. JavaTestingRepo [4] is a demo git repo for running this code while developing.