auto-mando is a Typst package for automatic Mandarin annotation of Chinese characters (漢字). It leverages the high-performance Rust-based WASM plugin from rust-mando to segment text and apply ruby annotations in either pīnyīn or zhùyīn (bopomofo / 注音符號).

Features

• Automatic Segmentation: Accurately splits Chinese sentences into words to ensure correct annotation placement.
• Pīnyīn Ruby: Places pīnyīn above (or below) each character via the rubby package.
• Zhùyīn Ruby: Places bopomofo to the right of each character by default (traditional Taiwanese textbook style), with top/bottom/left as alternatives.
• Tone Styles: Supports tone marks (nǐ, default) and tone numbers (ni3) for pīnyīn; tone marks are always used for zhùyīn.
• Content-Aware: Processes both plain strings and Typst content blocks, preserving line breaks and paragraph breaks.

Usage

Basic Pīnyīn Example

#import "@preview/auto-mando:0.2.1": pinyin-ruby

#set text(font: ("Libertinus Serif", "Noto Serif CJK TC"), size: 18pt)
#set par(leading: 1.5em)  // prevent annotation overlap between lines

#pinyin-ruby[北京歡迎你！]

Basic Zhuyin Example

#import "@preview/auto-mando:0.2.1": zhuyin-ruby

#set text(font: ("Libertinus Serif", "Noto Serif CJK TC"), size: 18pt)
#set par(leading: 1.5em)

#zhuyin-ruby[北京歡迎你！]

Bopomofo is placed to the right of each character by default, matching the layout used in Taiwanese elementary school textbooks.

Customizing Word Separation

#import "@preview/auto-mando:0.2.1": pinyin-ruby
#set text(24pt, font: ("Libertinus Serif", "AR PL KaitiM Big5"))
#set par(leading: 1.5em)

#pinyin-ruby(word-sep: 0.7em)[
  北京歡迎你們！

  你們今天過得怎麼樣

  world, 你可以幫幫我嗎？

  能不能告訴我現在你在做甚麼？
]

Pīnyīn Tone Styles

#pinyin-ruby(style: "numbers")[北京]  // → bei3jing1

Customising Ruby Styles

Both functions accept a ruby-styles dictionary that controls the appearance of the annotation. You only need to supply the keys you want to change — the rest are filled in from the defaults.

Pīnyīn defaults (pos must be top or bottom):

// Smaller annotation below the characters
#pinyin-ruby(
  ruby-styles: (size: 0.5em, pos: bottom),
)[北京歡迎你]

Zhùyīn defaults (pos may be top, bottom, left, or right):

// Bopomofo above the characters instead of to the right
#zhuyin-ruby(
  ruby-styles: (size: 0.5em, pos: top),
)[北京歡迎你]

Note on line spacing: At larger font sizes, ruby annotations may overlap the line above. Add #set par(leading: 1.5em) (or a larger value) before your annotated text.

API Reference

pinyin-ruby(it, style: "marks", ruby-styles: (...), word-sep: 0.25em)

Annotates Chinese text with pīnyīn ruby.

• it — string or content block.
• style — "marks" (default) or "numbers".
• ruby-styles — partial or full style dictionary (see table above).
• word-sep — horizontal gap between consecutive Chinese words (default 0.25em). Pass 0em to disable.

zhuyin-ruby(it, ruby-styles: (...), word-sep: 0.25em)

Annotates Chinese text with zhùyīn (bopomofo) ruby.

• it — string or content block.
• ruby-styles — partial or full style dictionary (see table above). Set pos: right (default) for traditional right-side bopomofo, or pos: top / bottom for above/below placement.
• word-sep — horizontal gap between consecutive Chinese words (default 0.25em). Pass 0em to disable.

pinyin-to-zhuyin(syllable)

Converts a single pinyin syllable with tone number (e.g. "pin1") to a bopomofo string (e.g. "ㄆㄧㄣˉ"). Useful for building custom renderers.

syllables-to-zhuyin(syllables)

Converts an array of pinyin-number syllables to an array of bopomofo strings, one per syllable.

flat(txt, style: "marks")

Returns a space-separated pīnyīn string (non-Chinese tokens omitted). Useful for metadata or search indexing.

segment(txt, style: "marks")

Low-level function returning an array of {word, pinyin} dictionaries. pinyin is none for non-Chinese tokens (punctuation, spaces, Latin).

render-segments-pinyin(segs, ruby-styles: (...), word-sep: 0.25em)

render-segments-zhuyin(segs, ruby-styles: (...), word-sep: 0.25em)

Low-level renderers that operate on the array returned by segment(). Use these if you need to pre-process segments before rendering.

Technical Details

• WASM Backend: Powered by rust_mando.wasm for fast, memory-efficient processing.
• Dependencies: rubby:0.10.2 for top/bottom ruby placement.
• Compiler Requirements: Typst 0.14.0 or higher.
• License: MIT.

Author

Vincent Tam (GitHub).