Project 2: Scrape a site
The very first thing I built using coding was a web forum. I didn’t know the first thing about coding - it was a mess of unformatted code, mangled file names, dozens of files without any order or structure. It was a nightmare. But it worked. And once you get that high, once you see it work once, you never really get over it.
If you’ve not written code before, I hope that this post makes you feel the same high I felt that day.
Back then, I didn’t have ChatGPT. Today, you do. Which means you can skip a lot of the needless pain and still experience the thrill - the joy - of bringing something new into existence.
In this post, we’ll create a script that will print the text of any website into a text file.
The only different from the 1st project is that here, you have to install a cool python package, called trafilatura. It’s best to do that first:
On Windows:
Press the Windows key, type cmd, and hit Enter. This will open the command prompt.
Type this exactly: pip install trafilatura
Then press enter
On Mac:
Press Command + Space, type terminal, and hit Enter. This will open Terminal.
Type this exactly: pip install trafilatura
Wait for the “Successfully installed trafilatura” message.
Now we’re ready!
Download the script from here.
Run it like we did in the first project, enter any url (I tried https://en.wikipedia.org/wiki/OpenAI), and watch it print that website into a file called output.txt. Cool huh?
Here’s the chat that helped create this script. Now it’s your turn to make it yours.
Tinker with it. Change the name of the file it creates. Add a timestamp so every run feels like a snapshot in time. Break it, fix it, and see what happens.
Or take it even further: make an input.txt file, drop in a bunch of URLs (one per line), and teach your script to fetch them all. Suddenly you’re not just running a program. You’re building a tool.
This is how we learn: by poking at the edges of what already works until something new emerges.
See you next time!
Hope to read about such more projects.
"Once you see it work once, you never really get over it"—that high of building something from nothing is exactly why I'm on Substack.
I'm 19, not a coder, but I recognize that feeling. The first time I published a blog post and 2 real people subscribed? That was my version of your web forum moment.
Your approach (download script → run it → tinker with it → break it → fix it → make it yours) is how real learning happens. Not tutorials. Not passive consumption. Building, breaking, rebuilding.
The suggestion to add a timestamp or create an input.txt for batch URL scraping? That's teaching people to think beyond the template.
"This is how we learn: by poking at the edges of what already works until something new emerges."
That applies to everything, not just code 💙