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

Wiktionary page text for every Han-script character (Unicode Script_Extensions=Han), plus every component recursively referenced from those pages' ids= template fields, 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-ids-page-text

Usage

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

How the data is generated

scripts/parse.ts streams the Wiktionary pages-articles-multistream dump and:

1. Scans the index for titles whose every code point matches \p{Script_Extensions=Han} — the seed set.
2. Decompresses only the bz2 sub-streams that contain seed titles and extracts those pages.
3. Parses every fetched page's {{...|ids=…}} template fields (IDS = Ideographic Description Sequence, the Unicode notation that uses Ideographic Description Characters like ⿰ / ⿱ / ⿻ to describe how a Han character decomposes into components), gathers each single component code point referenced inside (skipping the IDCs themselves at U+2FF0–U+2FFF and U+31EF, plus ASCII, CJK punctuation, and fullwidth forms), and recursively fetches any newly-referenced titles that exist in the dump's index.
4. Repeats step 3 until no new references are discovered, then 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.

Non-Han glyphs pulled in via ids=

Versus en.wiktionary.org-wiki-scx-han-page-text, this package additionally ships every non-Han glyph that appears inside a Han page's ids= field. Wiktionary IDS authors frequently borrow Kana, Bopomofo, geometric shapes, etc. as visual stand-ins for components that have no encoded code point of their own. The Referenced by column lists Han titles whose ids= field contains the glyph.

| Char | Codepoint | Unicode block | Referenced by | |---|---|---|---| | ʓ | U+0293 | IPA Extensions | 𠍋 | | ॸ | U+0978 | Devanagari | 𦹗 | | ∾ | U+223E | Mathematical Operators | 𢀓 | | ◇ | U+25C7 | Geometric Shapes | 𠄷 | | ◊ | U+25CA | Geometric Shapes | 𤕈 | | ➿ | U+27BF | Dingbats | 𡦹 | | は | U+306F | Hiragana | 𭁇 | | コ | U+30B3 | Katakana | 𠁲, 𠂶, 𠂼 … (51 total) | | サ | U+30B5 | Katakana | 𬻀 | | ス | U+30B9 | Katakana | 𢀖, 𬽡, 𮌇 | | タ | U+30BF | Katakana | 罗 | | チ | U+30C1 | Katakana | 刋 | | ツ | U+30C4 | Katakana | 𫝂 | | テ | U+30C6 | Katakana | 𫝆 | | ノ | U+30CE | Katakana | 𣘺 | | モ | U+30E2 | Katakana | 𪜈 | | ユ | U+30E6 | Katakana | 𠀌, 𠇛, 𠇡 … (15 total) | | リ | U+30EA | Katakana | 𠇬, 𱍬, 𱍭 … (9 total) | | ㄣ | U+3123 | Bopomofo | 𪜀, 专 | | ㅿ | U+317F | Hangul Compatibility Jamo | 㐃 | | ㆳ | U+31B3 | Bopomofo Extended | 𠇇 | | 𛀲 | U+1B032 | Kana Supplement | 𠆭 | | 𛂦 | U+1B0A6 | Kana Supplement | 𭤪 |

License

MIT