GitHub LOC Counter

A powerful command-line tool designed to analyze GitHub repositories and local directories, accurately counting Lines of Code (LOC) and visualizing repository structures.

Unlike other online counters (such as ghloc or codetabs):

• True SLOC Count: Instead of just counting text-containing lines (like standard engines), this tool computes actual Source Lines of Code (SLOC), comments, and blank lines.
• Private & Multiple Repositories: It fully supports private repositories (using a GitHub token) and analysis of all public repositories for a given user.

---

1. Features

• Accurate Line Counting (SLOC): Counts exact Source Lines of Code, comments, and blank lines with high precision using the locc4j engine.
• Repository Tree Visualization: Generates and prints a directory tree structure directly to the console.
• Multi-Repository Analysis (User Mode): Analyzes and processes all public repositories of a given GitHub user in a single command.
• File Sorting: Groups and ranks files by LOC or programming languages.
• JSON Export: Export results into JSON files for easy parsing or external use.

Preview

📄 Example JSON Result (describing the structure and LOC of this project)

---

2. Usage

There are two main ways to run this application:

1. Option 1: From Source (Java 17 & Maven)

   mvn clean package
   java -jar target/githubloc-1.0.jar <command>

2. Option 2: Prebuilt Executable (.exe for Windows / binary for Linux) Download the latest archive from the Releases tab on GitHub and run:

   ./github-loc.exe <command>

For detailed, step-by-step setup and GitHub API token configuration instructions, see the Setup Guide.

---

3. Command Guide

Commands follow a simple format: [TARGET] [ACTION] [SORT]

Quick Examples

• Analyze a single GitHub repository:

  facebook/react

• Analyze all public repositories of a specific user:

  sonle1501

• Sort and group files by programming language:

  sonle1501/github-loc -a sort BYLANG

For the complete documentation on modes, action flags (-a/--action), sorting arguments, and interactive prompt mode, see the Command Guide.

---

4. Output Storage

Results and intermediate files are neatly organized in the local storage/ directory:

• storage/repos/: Unzipped repository source folders.
• storage/zip-repos/: Raw downloaded repository .zip archives.
• storage/json-results/: JSON analysis tree outputs and sorted lists.
• storage/user-repos/: JSON results of multi-repository analysis.

---

5. Benchmark

To evaluate the performance of github-loc, I ran a benchmark on the Elasticsearch repository (~6M LOC) using the JMH (Java Microbenchmark Harness).

Results

Below are the average execution times (in milliseconds per operation, ms/op) for each approach:

Benchmark	Mode	Cnt	Score	Error	Units
CreateTreeBenchmark.measureBuildTreeMultithreading	avgt	5	3412.355	± 110.319	ms/op
CreateTreeBenchmark.measureBuildTreeMultithreadingWithBatch	avgt	5	4074.866	± 373.230	ms/op
CreateTreeBenchmark.measureBuildTreeSequential	avgt	5	15530.403	± 297.930	ms/op

Execution results visualized below:

---

6. License

This project utilizes the locc4j counting engine.

The github-loc project is licensed under the MIT License.