Integrating Myself With Wikipedia

min read

2026-07-29 Back to posts

Nickipedia Wikipedia-inspired personal wiki

How I turned a part of this website into a small Wikipedia-inspired ecosystem with rabbit holes, featured pictures, search, and live encyclopedia data.

Tech Coding All Tags

Share X Facebook LinkedIn Reddit Email

Table of Contents

Integrating Myself With Wikipedia

I’ve always loved the idea behind Wikipedia. So I decided to make a page on this website in a similar style.

I call it Nickipedia.

Charlie Day conspiracy board

Making a clone of Wikipedia itself is impossible and also completely unnecessary. The goal was to make my own website feel less like a collection of pages and more like a connected knowledge base.

The Problem With Personal Websites

Most personal websites are boring.

There isn’t anything wrong with that but I wanted something closer to a personal archive, something closer to how I’ve learned information over the years. Something that represented all the random things I write about.

Over the last few months, the site had grown into a weird combination of:

Blog posts
Books I’ve read
Movies and media I like
Music I’ve released
Software projects
Recipes
Random historical rabbit holes

Here was the issue: The library didn’t know about the blog. The blog didn’t know about the projects. The projects didn’t know about anything else.

Wikipedia solved this problem decades ago with links. Everything leads somewhere else.

Building The Data Layer

The first step was figuring out what information already existed.

The homepage already imported almost everything:

import { books } from "../data/library";
import { mediaList } from "../data/media";
import { tracks } from "../data/music";
import { featuredCases, openSourceProjects } from "../data/projects";

Instead of creating a database, I realized the site already had a lot of structured data.

The next step was combining everything into one searchable collection:

const unifiedSearchData = [
  ...postSearchData,
  ...librarySearchData,
  ...mediaSearchData,
  ...musicSearchData,
  ...projectSearchData,
];

Now the search bar searches the entire site. A book, a song, a project, or a blog post all exist in the same universe.

Creating The Wikipedia Rabbit Holes

A big reason Wikipedia is addictive is because every article is a doorway. I wanted to capture that feeling.

Since my posts already contained Wikipedia links, I wrote a script that scans every article and collects them.

const wikipediaLinksMap = new Map();

const wikiHrefPattern =
  /href="(https?:\/\/(?:[a-z]{2,3}\.)?wikipedia\.org\/wiki\/[^"]+)"/gi;

for (const post of allPosts) {
  let html = await post.compiledContent();

  for (const match of html.matchAll(wikiHrefPattern)) {
    const url = match[1];

    if (!wikipediaLinksMap.has(url)) {
      wikipediaLinksMap.set(url, {
        url,
        postTitle: post.frontmatter.title,
      });
    }
  }
}

Live Wikipedia Data

The next step was making the site pull information directly from Wikipedia. Wikipedia provides public APIs that anyone can use.

For example, the featured picture:

fetch(
  `https://en.wikipedia.org/api/rest_v1/feed/featured/${yyyy}/${mm}/${dd}`,
)
.then((res) => res.json())
.then((data) => {
  const image = data.image;
});

Every day, the site gets a new featured picture from Wikimedia Commons.

I also added “On This Day.”

fetch(
  `https://en.wikipedia.org/api/rest_v1/feed/onthisday/selected/${month}/${day}`,
)
.then((res) => res.json())
.then((data) => {
  console.log(data.selected);
});

Making Search Feel Like Wikipedia

The search system uses Fuse.js, which provides fuzzy matching.

Basically, it lets users find things even if they don’t type the exact title.

The search data looks like this:

{
  type: "Article",
  title: "The Labor Force Never Fully Came Back",
  description: "Analyzing BLS labor data",
  url: "/posts/labor-force"
}

Then Fuse searches across titles and descriptions:

const fuse = new Fuse(items, {
  keys: ["title", "description"],
  threshold: 0.3,
});

The result is a small internal search engine for everything I’ve created using an existing search algorithm.

Why Do This?

The modern web is increasingly becoming a collection of feeds controlled by algorithms. Everything is optimized for keeping you scrolling. Nickipedia lets you explore my personal knowledge base, enhanced by the best knowledge database we have: Wikipedia.

I wanted something slower. Something you can sink into.

Homer Simpson disappearing into bushes