An introduction to search-index.js
summary
Search-index claims a working full‑text search engine in JavaScript that runs on a LevelDB backend and swaps to IndexedDB in browser (or even localStorage) via the Level family and Browserify and NPM. It stores TF/IDF as key/value ranges, supports facets, filters, geohashes for spatial queries, and replicates an index file over HTTP. Examples include client replication and a web‑worker indexer that avoids UI blocking. It targets small‑to‑moderate datasets (hundreds of MB) and scales by offloading thousands of searches to browsers.
AI-written summary · about
chapters (20)
- 0:00Introduction
- 0:24About Mats Olsen
- 1:24About searchindex
- 2:34Features
- 3:28LevelDB
- 4:25LevelUp
- 4:44Backend
- 5:08Browserify
- 5:19Brass Certified
- 5:32Server side
- 5:51Low bandwidth costs
- 7:35Offline performance
- 8:18How it works
- 13:56Simple example
- 16:52Running search index inside the browser
- 21:00Search index and replication
- 24:18Search index in a web worker
- 27:05Future of search index
- 29:16Get involved
- 32:18Question
related talks
27mLyra: Disrupting full-text search industry with JavaScript
28mLyra: Disrupting the full-text search industry with JS
53mWriting a full-text search engine in TypeScript
19mDisrupting Full Text Search Industry With JavaScript
34mBuilding a search engine with Clojure
55mFull-Text Search Demystified
51mFront End Data
24mSearch From the Ground Up