First Interaction With Open Source

#opensource #java

Introduction

I've never thought that my hand will ever touch an open source, as it sounded like something unreachable to me. However, starting this semester I chose the course called "Open Source Development" by David Humphrey, not knowing what is waiting me, my expectation was like "Another boring course to fulfill my program", but once we started I realized that it is going to be something different from other courses. On the first lecture we were tasked with building a simple a CLI tool that allows us to use any compatible AI API endpoint in order to transform files in some helpful way. Probably, it changed my life. Literally, I have never been that passionated about programming, had thoughts that I don't belong to this sphere, but we all know that road isn't always straight, sometimes it's narrow. I witnessed so many students in our class which were so skilled, the only thought that revived my love to coding was: "Am I worse than them? Nahhh... It pushed me so hard, so I am sitting third week without leaving my laptop. I chose completely new language, started my research, and testings. Eventually, I came up with something simple, but interesting that helped me to understand that I can, literally, build anything I want.

PolyglotCode

PolyglotCode is a command-line tool that aids developers to translate their files with the code in ANY other programming language. This tool was written in Java using Maven dependencies manager. It simply takes file, language user wants to translate sends it as a query to the API endpoint and outputs translated material to the terminal or writes it to the file (user's choice.) Going forward, I want to mention that I used Cohere's AI v1 endpoint.

I uploaded video with detailed demo of the tool.

Getting Started

Step 1. Generate and provide Cohere API key into the defaultValue of -a && --api-key flag inside of the main class:

@Option(
        names = {"-a", "--api-key"},
        defaultValue = "YOUR_API-KEY",
        description = "Modifying api key manually"
)
private String api;

Specify call and provide the api-key into the command line, using flags(Available only after Step 3) :

./polyglot <file(s)...> <language> -a <YOUR-API-KEY>

Step 2. Make bash script executable, so user would be able to use ./polyglot command to run code

chmod +x polyglot

Step 3. Run Maven Package to compile the source code

mvn package

Features

For the reason, that is release v0.1, I was struggling crafting this tool(struggling, I mean it took a lot of time I ENJOYED IT :D) it doesn't contain all the features that I planned to implement. However, it still contains some of the required features, I just simply was following the requirements. So here they are:

Help

Isn't peak of programming, (-h, --help) flags. This flag helps user to learn the usage and all flags. Usage:

./polyglot -h

./polyglot --help

Version

Another peak of programming is (-v, --version) flag. Shows current version of the tool. Usage:

./polyglot -v

./polyglot --version

Api-Key

As it was mentioned earlier, user is able to specify Cohere-Api key without going into the code, simply using flags (-a, --api-key) Usage:

./polyglot <file(s)...> <language> -a <YOUR-API>

./polyglot <file(s)...> <language> --api-key <YOUR-API>

BaseUrl

It is not recommended to change the baseUrl of this application, as it may produce an unexpected output; however this option is still available to the user. Usage:

./polyglot <file(s)...> <language> -u <baseUrl>

./polyglot <file(s)...> <language> --base-url <baseURL>

Output

Creates the file, specified by user and writes the result there. It's highly recommended to use this option with one file at the time. Usage:

./polyglot <file(s)...> <language> -o file.txt

./polyglot <file(s)...> <language> --output file.txt

How it looks in the new file:

Challenges I faced

After I shared all my features with you, it's time to talk about the challenges that I faced. One of the challenges was learning new language, it wasn't that hard as I learnt before C/C++. Second one was finding appropriate AI endpoint because many of them wouldn't be that developed to answer even simple questions, and my goal was to find it for free :) Third challenge came with learning the usage of picoCLI, tool that helps you to setup CLI tool, at first I created a bash script to help me with it, I was trying to avoid use of picoCLI, then It became soooo inconvenient, and I decided to learn about picoCLI better.

Overall, this project helped me to understand that everything is possible, if you work hard!