I have extracted MDBG cantonese input data:
aa: 那 阿 呀 亞 啊 丫 啞 吖 厂 婭 扼 亚 嗄 氬 鴉 錒 掗 錏 厊 咓 瘂
aai: 唉 哎 嗌 欸 挨 隘 誒 餲
aak: 呃 軛 齷 軶
aan: 晏 鷃 鴳
aang: 甇 罌 甖….
You can test drive this input here: http://pcwong.org/gwdw/mdbg/
My Cantonese Input Data:
Total characters in syllables:12479
Take into account that Cantonese is dynamic and the total syllables will keep increasing. The lost character-syllable will constantly be found and understood. Keep the same data will not be appropriate and with accessing data through my site, user will get the most upldated set.
You can test drive this input here: http://pcwong.org/cantonese/mobile/
Other than inputing chinese, my data includes the tone information and which is important to let users get the exact pronounciation of the syllable for native and non-native speaker which is a good learning tool as welll.
Hope that you will enjoy the system.