How I Turned an Arabic Word List into an Automated YouTube Channel

Hello! I am a web developer with knowledge of Python, and I am bilingual. I love educating people and learning new things. Last week, I came across an amazing but unmaintained GitHub repository created by Dawi Alotaibi, who allegedly works at IBM. The repository contains a huge dataset of Arabic words, sorted alphabetically. At first, I wasn’t sure whether my idea would succeed, but after some time thinking, I decided to give it a try. I primarily used ChatGPT because I had never worked with MoviePy (a Python library for generating videos). I had a great time experimenting with it. After an hour of working between ChatGPT and the codebase, I managed to write my first simple script that generates a short video with text and audio. The audio is generated using gTTS (Google Text-to-Speech), which gave me a somewhat human-like voice. On the second day, I connected my simple script to the dataset I obtained from the repository mentioned earlier. I added it as a sub-module and used the built-in open function to read the words from the JSON file. After understanding how to work with the JSON file in Python, I had to make a decision: bulk-generate videos, risking damaging my laptop, or upload them one at a time. I decided to generate one video at a time. To automate the process and avoid manually entering the last word, I created an index.txt file to store the last processed index number. The next step was uploading the videos to YouTube. I created a new YouTube channel, which was quite easy. Then, I asked ChatGPT to help me create a simple script for uploading videos to YouTube and guide me on how to use it. While the process of getting the token and integrating it was difficult and frustrating, I hope Google can improve it in the future. Once everything was working, I needed a script to automatically upload one video every day. Although I could have run it locally or bought a server, I’m more familiar with GitHub Actions, so I created a simple script to run every day at midnight UTC. This script generates a video, uploads it to YouTube, and pushes a commit with the updated index. And it worked! Throughout this project, I learned a lot of cool things, like how to use MoviePy and the YouTube API. I’m hopeful that it will succeed. I almost forgot to mention that it's open source and MIT-licensed, so you can check out the code, contribute, or even use it in your personal projects!

Apr 16, 2025 - 04:56

How I Turned an Arabic Word List into an Automated YouTube Channel

Hello!
I am a web developer with knowledge of Python, and I am bilingual.

I love educating people and learning new things.

Last week, I came across an amazing but unmaintained GitHub repository created by Dawi Alotaibi, who allegedly works at IBM. The repository contains a huge dataset of Arabic words, sorted alphabetically.

At first, I wasn’t sure whether my idea would succeed, but after some time thinking, I decided to give it a try. I primarily used ChatGPT because I had never worked with MoviePy (a Python library for generating videos). I had a great time experimenting with it. After an hour of working between ChatGPT and the codebase, I managed to write my first simple script that generates a short video with text and audio. The audio is generated using gTTS (Google Text-to-Speech), which gave me a somewhat human-like voice.

On the second day, I connected my simple script to the dataset I obtained from the repository mentioned earlier. I added it as a sub-module and used the built-in open function to read the words from the JSON file. After understanding how to work with the JSON file in Python, I had to make a decision: bulk-generate videos, risking damaging my laptop, or upload them one at a time. I decided to generate one video at a time. To automate the process and avoid manually entering the last word, I created an index.txt file to store the last processed index number.

The next step was uploading the videos to YouTube. I created a new YouTube channel, which was quite easy. Then, I asked ChatGPT to help me create a simple script for uploading videos to YouTube and guide me on how to use it. While the process of getting the token and integrating it was difficult and frustrating, I hope Google can improve it in the future. Once everything was working, I needed a script to automatically upload one video every day. Although I could have run it locally or bought a server, I’m more familiar with GitHub Actions, so I created a simple script to run every day at midnight UTC. This script generates a video, uploads it to YouTube, and pushes a commit with the updated index. And it worked!

Throughout this project, I learned a lot of cool things, like how to use MoviePy and the YouTube API. I’m hopeful that it will succeed. I almost forgot to mention that it's open source and MIT-licensed, so you can check out the code, contribute, or even use it in your personal projects!