Package Exports

@fasteroid/partial-pattern-tree
@fasteroid/partial-pattern-tree/index.js

This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (@fasteroid/partial-pattern-tree) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

Readme

@fasteroid/partial-pattern-tree

“Not quite partial regex matching, but pretty darn close!”

Premise

Say you want to assign "names" to a series of values and then search for them.

The "best" result—since users are creative and unpredictable—is to look for partial matches. This is really easy if your "names" are just unchanging string literals (just see if they contain the search query), but becomes an absolute head-scratcher when you decide you want to name your list entry Order ID\d+ where \d+ is any sequence of digits.

And no, you can't just naively regex this either, because doing that forfeits partial matching. "Order" (or just "der" if your user is a lunatic) doesn't match Order ID\d+, so the option won't show up 😭

So are you just cooked, or is there a way? Fear not, as this package (kind of) solves your problem!

Usage

import { PartialPatternTree } from "@fasteroid/partial-pattern-tree"

type Pet = {
  importance: string,
  name: string
}

const tree = new PartialPatternTree<Pet>(
    [
        ["parake", /^e+/m, "t"], // this "name" matches parakeet, parakeeet, parakeeeee... you get the idea.
        {
            name: "Newton"
            importance: "actually ate my homework once (not kidding)"
        }
    ],
    [
        ["parrot"], // this name just looks for "parrot"
        {
            name: "Polly"
            importance: "has a crippling cracker addiction"
        }
    ]
)

tree.search("keeeeeeeeeee"); // returns an array containing Newton

But what if..?

Q: I want to order the search results based on relevance and I want it to not be complete garbage!
- A: First of all that's not a question—but yes it does do that. The metric I chose is "how many characters into the name must we skip to find the search query?"
Q: This is literally just a generalized suffix tree with extra steps!
- Also not a question. But good observation!
Q: If this is a suffix tree why aren't you constructing it with Ukkonen's algorithm?
- A: I'm an idiot 🤡
  ...but if you know how, I'd love if you made a pull request!
Q: Why is this README so GOOFY?
- A: To make you smile! Duh!!!

License

MIT License