en.wiktionary.org-wiki-scx-han-page-text

Wiktionary page text for every Han-script character (Unicode Script_Extensions=Han), extracted from the English Wiktionary XML dump and shipped as per-character dynamic imports.

Each character maps to an async function that lazy-loads only the JSON bucket it lives in, so consumers pay for the characters they actually look up — not the entire dataset.

Install

npm install en.wiktionary.org-wiki-scx-han-page-text

Usage

import wiktionaryPages from "en.wiktionary.org-wiki-scx-han-page-text";

const value = await wiktionaryPages["漢"]();
// value is either the raw wikitext (string) or a redirect target
// represented as a single-element tuple: [string].

Return value shape

• string — the raw wikitext of that character's Wiktionary page.
• [string] — the page is a redirect; the array contains the normalized target title.

import type { WiktionaryPageValue } from "en.wiktionary.org-wiki-scx-han-page-text";
// WiktionaryPageValue = string | [string]

How the data is generated

scripts/parse.ts streams the Wiktionary pages-articles-multistream dump, scans the index for titles whose every code point matches \p{Script_Extensions=Han}, decompresses only the bz2 sub-streams that contain those titles, and writes:

• unicode/<hi-byte>xx.json — bucketed page data, keyed by the high byte of each title's first code point.
• index.js / index.d.ts — the default-exported wiktionaryPages map of character → dynamic import.

To regenerate from a fresh dump:

node --experimental-strip-types scripts/parse.ts \
  path/to/enwiktionary-YYYYMMDD-pages-articles-multistream.xml.bz2 \
  path/to/enwiktionary-YYYYMMDD-pages-articles-multistream-index.txt.bz2

Requires bzip2 on PATH.

License

MIT