GitHub Repo Java Method Analyzer

This program analyzes the increases of the parameters of the method declarations of Java source files in the commits of the master branch of any GitHub repository.

Objectives

Say, you have edited your Java file src/MyFile.java to change a method signature from void doTask(int x) into void doTask(int x, String y) in a commit (SHA: 0798eaa) of your repository your-username/your-repo hosted on GitHub. Now if you run GitHub Repo Java Method Analyzer tool to analyze your repository, the tool will detect the method signature change, and it will report the change into a CSV file named your-username_your-repo_methodAnalysis.csv. The CSV file will contain the following information:

Commit SHA	Java File	Old function signature	New function signature
0798eaa	src/MyFile.java	void:doTask:int	void:doTask:int,String

Build Project

Requirements:

• JDK 1.8 or later
• Maven, or any Java IDE which supports Maven

Build steps:

1. If you want to use Maven directly (without any IDE), run the following command in the project directory where pom.xml file exists:

   mvn package

   Or, you can use any supported Java IDE (like IntelliJ, Eclipse) to build the project.

2. If the build process ends successfully, it will generate github-repo-java-method-analyzer-1.0-SNAPSHOT.jar file in /target directory.

Run Project

To run the project, use the command in /target directory:

java -jar github-repo-java-method-analyzer-1.0-SNAPSHOT.jar

The program will show the following usage arguments:

repo_full_name [-u username password] [-p processed_commit_limit] [-t thread_count]

Here:

• repo_full_name: Use both owner's name and repo name like google/guava.

• [-u username password]: It is optional but recommended. GitHub API has a certain rate limit. For unauthenticated requests (without using a username and its corresponding password/token), the rate limit allows for up to 60 requests per hour. For authenticated requests, the rate limit allows at least 5,000 requests per hour. The program uses 1 API call per processed commit.

• [-p processed_commit_limit]: It is optional. You can enforce the program to limit how many commits to be processed. The program will start from the latest commit and process older commits one by one up to the limit. The default behavior of the program is to process all the commits of the master branch of a repository.

• [-t thread_count]: It is optional. You can enforce the program to use a certain number of threads to process each commit. A larger number of threads will be helpful for the commits which contain a lot of eligible files (modified Java files) per commit. But, by design, a larger number of threads will not be helpful for the commits which contain only a single eligible file per commit. The default behavior of the program is to use only a single thread to process each commit.

For example:

java -jar github-repo-java-method-analyzer-1.0-SNAPSHOT.jar google/guava -u yourname yourtoken -p 1000 -t 100

Output

Console output:

In the console, the program will provide a similar output of the following screenshot:

File output:

The program will generate two files for each repository:

1. CSV file: It is the main output file. It has been already stated that the CSV file will contain the following information:

   Commit SHA, Java File, Old function signature, New function signature

2. State file: This file is for the program's internal purpose. The file will contain the SHA value of the last processed commit. If you terminate the program, it will start from the commit stored in the file, instead of the latest commit.

Sample Results

I have run the program to analyze the following repositories:

• Google Guava
• Apache Dubbo (incubating) Project
• Spring Boot
• JavaParser

I have attached all the results of these repositories in _SampleResults directory.

Acknowledgement

I would like to thank Professor Dr. Sarah Nadi, University of Alberta, for assigning me to do the project. I have learned a lot of things while working on the project.