Language Support

Mobile-Pretext handles text in any script that the platform's word segmenter and font engine support. No special configuration is needed — all languages work out of the box.

Tested Scripts

Script	Languages	Special handling
Latin	English, French, German, etc.	Standard word-boundary breaking
Hebrew	Hebrew	RTL bidi levels
Arabic	Arabic, Urdu, Farsi	RTL bidi levels, no-space punctuation rules
CJK	Chinese, Japanese, Korean	Per-character breaking, kinsoku shori
Thai	Thai	Dictionary-based word segmentation
Devanagari	Hindi, Sanskrit	Combining mark handling
Myanmar	Burmese	Medial glue rules
Khmer	Cambodian	Dictionary-based segmentation
Lao	Lao	Dictionary-based segmentation
Korean	Korean	Hangul syllable-level breaking
Emoji	All platforms	Correct ZWJ sequence widths

CJK Line Breaking

Chinese, Japanese, and Korean text uses per-character line breaking — each grapheme is a potential break point. This matches browser behavior.

Kinsoku Shori

Japanese/Chinese line-breaking rules prevent certain characters from appearing at line boundaries:

Line-start prohibited (can't start a line): 。、！：；？）」】＞」」 and iteration marks ゝゞヽヾ々〻

Line-end prohibited (can't end a line): "（「【＜「 and opening quotes "'«‹

Bidirectional Text

Mixed LTR/RTL text (e.g., English mixed with Hebrew or Arabic) gets bidi embedding levels computed via a simplified Unicode Bidi Algorithm. These levels are available on the PreparedTextWithSegments handle for custom rendering.

INFO

The line-breaking engine itself does not reorder text based on bidi levels — it breaks lines in logical order. Bidi levels are metadata for renderers that need to display text in visual order.

Emoji

All emoji are supported, including:

• Basic emoji (😀, 🚀)
• Emoji with variation selectors (❤️)
• Skin tone modifiers (👋🏽)
• ZWJ sequences (👨‍👩‍👧, 🏳️‍🌈)
• Regional indicators / flag sequences (🇯🇵)

Each emoji grapheme cluster is measured as a single unit with correct width.