build-search-index.mjs

#!/usr/bin/env node

/**
 * Builds a MiniSearch full-text search index from all HTML files in the tree.
 * Outputs search-index.json which is loaded by index.html at runtime.
 */

import { readFileSync, writeFileSync } from "node:fs";
import { glob } from "node:fs/promises";
import * as cheerio from "cheerio";
import MiniSearch from "minisearch";

const ROOT = process.cwd();
const EXCLUDED = new Set(["index.html", "sitemap.html"]);

const documents = [];
let id = 0;

for await (const file of glob("**/*.html", { cwd: ROOT })) {
  if (EXCLUDED.has(file)) continue;

  const html = readFileSync(file, "utf8");
  const $ = cheerio.load(html);

  const title =
    $(".page-title").first().text().trim() ||
    $("title").first().text().trim() ||
    $("h1").first().text().trim() ||
    file;

  // Extract visible body text, stripped of HTML
  // Remove script and style elements first
  $("script, style").remove();
  const text = $("body").text().replace(/\s+/g, " ").trim();

  documents.push({ id: id++, title, text, url: file });
}

console.log(`Indexed ${documents.length} documents`);

const miniSearch = new MiniSearch({
  fields: ["title", "text"],
  storeFields: ["title", "url"],
});

miniSearch.addAll(documents);

writeFileSync("search-index.json", JSON.stringify(miniSearch));
console.log("Wrote search-index.json");