Licences & credits

Sources

rime-cantoneseCC BY 4.0

Jyutping readings, including the colloquial tone-change forms that most sources omit.

The rime-cantonese contributors

CC-CEDICTCC BY-SA 3.0

English definitions for Chinese words.

MDBG and contributors

CC-CantoCC BY-SA 3.0

Cantonese-specific definitions and readings.

Pleco Software and the CC-Canto contributors

to-jyutpingBSD 2-Clause

Choosing the right reading for a character from the words around it.

CanCLID

SUBTLEX-CHCC BY

Character frequencies drawn from film subtitles, which reflect spoken rather than written Chinese. Used when the app is built, to choose which character best represents a syllable; the dataset itself is not shipped.

Cai, Q. & Brysbaert, M. (2010). PLoS ONE 5(6): e10729

Tesseract OCR and tesseract.jsApache License 2.0

Recognising Chinese text in a photograph, on your device.

Google and the Tesseract contributors

LucideISC

Interface icons.

The Lucide contributors

Anton and Noto Serif HKSIL Open Font License 1.1

Display type, and the Han face with Hong Kong glyph forms.

Vernon Adams and Google

Share-alike

CC-CEDICT and CC-Canto are licensed CC BY-SA 3.0. The dictionary data we derive from them is offered under the same licence, and we will provide that derived data on request to anyone who asks at support@snap2story.com. The application code itself is a separate work and is not covered by that licence.

Recordings

The syllable recordings shipped with Canto Reader are generated with a licensed text-to-speech voice and are part of the application. They are not offered for reuse.

A thank you

Cantonese is badly served by language tools, and what does exist is mostly the work of volunteers who built it anyway. rime-cantonese in particular is why the tones in this app are right — including the several hundred syllables that exist only through tone change and that most systems silently get wrong. We are grateful.