Logos Seeker

A static, serverless scripture search tool for the Recovery Version Bible, in English and Chinese (中文). One omni search bar accepts either a verse reference or any word/phrase and instantly returns matching verses. It runs entirely in the browser — host it on GitHub Pages with no backend.

Features

• Reference lookup — John 1:1, John 1 (whole chapter), John 1:1-5 (range), 1 John 2:3, plus Chinese forms 约翰福音 1:1, 约 1:1, 创 1.
• Fuzzy / blurred references — type without a colon and get every plausible reading: heb 111 → Heb 1:11 and Heb 11:1; heb11 → Heb 11 (whole chapter) and Heb 1:1. Works for Chinese too (约 111).
• Word / phrase search — Christ, 基督, 耶和华. The query language is auto-detected: Chinese queries search the Chinese text, English the English. Multiple space-separated words are matched as AND, in any order/position — e.g. join spirit finds "…joined to the Lord is one spirit." (1 Cor 6:17). English queries need at least 2 characters; Chinese has no minimum. Results show 20 at a time with a More results button.
• EN / 中文 toggle — switch the display language; results re-render instantly.
• Light / dark theme — a ☀/🌙 switch in the top-right corner; it overrides the OS preference and is remembered across sessions.
• Compact results — each match is a single line: a copy button, the reference, and the verse text.
• Copy — every row has a copy button that puts Reference text on the clipboard.
• Select verses — click a verse's text to select it (click again to deselect). With a selection active, any copy button copies all selected verses (in canonical order, one per line); a floating Deselect all button appears once two or more are selected.
• Click any reference to switch to the full-chapter view for context.
• Chapter navigation — in the full-chapter view, ←/→ buttons (or the ←/→ arrow keys when the search box isn't focused) jump to the previous/next chapter, crossing book boundaries (and disabled at the ends of the canon).
• Interlinear mode — in the full-chapter view, an Interlinear toggle interleaves each verse's English and Chinese lines. Copying an interlinear verse copies both languages. (Chapter view only.)
• Clear quickly — an ✕ button in the search box, or press Esc, clears it.
• Shareable URLs — every search updates the URL hash (#q=John+1:1).
• Offline after first load — the verse data loads once, then search is local.

How it works

data/verses.json ─┐
data/books.json  ─┼─► index.html + src/*.js  (browser app)
                  ┘

• data/verses.json holds the aligned bilingual text — English is the master spine (31,102 verses, the canonical count) with the Chinese for each verse alongside it. data/books.json holds per-book metadata and the reference alias table.
• src/parseQuery.js classifies a query as a reference or a word search and parses references against a per-book alias table.
• src/search.js does O(1) reference lookup and word search via an English token inverted index and a Chinese bigram index (built lazily in the browser).
• src/app.js is the vanilla-JS UI (no framework, native ES modules).

Run locally

npm run serve          # serve the static site at http://localhost:5050

Any static file server works (e.g. python -m http.server). Then open the page in a browser. No build step is required — the data is committed as JSON.

Tests

node scripts/test-parse.mjs    # reference-parsing assertions
node scripts/test-search.mjs   # search-engine assertions (uses local data)

Use as a library (npm)

The search engine and query parser are also published as an npm package, with the verse data bundled under logos-seeker/data/.

npm install logos-seeker

import { BibleSearch, COL } from "logos-seeker";

const bible = new BibleSearch();

// Browser: serve the JSON statically, then point load() at its base path
// (defaults to "" → fetches "data/verses.json" + "data/books.json").
await bible.load("/some/base/");

// Node / bundler: read or import the JSON and inject it directly.
// import verses from "logos-seeker/data/verses.json" with { type: "json" };
// import books  from "logos-seeker/data/books.json"  with { type: "json" };
// bible.setData(verses, books);

const ref = bible.parse("John 3:16");               // { type: "ref", ... }
const rows = bible.lookupReference(ref);            // [rowIdx, ...]
for (const i of rows) {
  console.log(bible.refLabel(bible.verses[i], "en"), bible.verses[i][COL.EN]);
}

const { rows: hits } = bible.wordSearch("基督", "cn"); // word/phrase search

Exported API: BibleSearch, COL, parseQuery, parseReference, buildAliasIndex, hasCJK. The data files are reachable at logos-seeker/data/books.json and logos-seeker/data/verses.json.

Deploy to GitHub Pages

1. Push this repo to GitHub.
2. In Settings → Pages, set the source to GitHub Actions.
3. The included workflow (.github/workflows/deploy.yml) publishes the static site on every push to main.

Because it's fully static, you can also serve it directly from the repo root or a gh-pages branch.

Data source

Text is the Holy Bible Recovery Version (Living Stream Ministry). All rights to the translation belong to their respective publisher; this tool is for personal study and search.