Features/libpinyin

= libpinyin =

Summary
The feature covers the integration for ibus-pinyin of a new libpinyin backend to provide intelligent sentence-based Chinese pinyin input.

Owner

 * Name: Peng Wu
 * Email: pwu@redhat.com

Current status

 * Targeted release: Fedora 17
 * Last updated: 2012-02-16
 * Percentage of completion: 100%

Detailed Description
User requirements:
 * More Intelligent Pinyin Input Method
 * Some users get used to some intelligent pinyin imes, which can input a whole sentence by only choosing zero, one or two candidates, the pinyin ime will figure out the actual sentence the user wants.

Benefit to Fedora

 * This will make the Fedora pinyin input method more convenient and intelligent to use, also speed up the user input when using pinyin.
 * Currently ibus-pinyin needs more user interactions to input a sentence.

Scope

 * Make ibus-pinyin more easier and intelligent to input Chinese sentences.
 * Use libpinyin as a back end of ibus-pinyin. (done)


 * libpinyin large web corpus training. (done)
 * rewrite full pinyin/double pinyin/chewing parsers. (done)


 * In Fedora 17 the remained tasks are:
 * Improves the libpinyin self-learning code; (in progress)

How To Test

 * Check whether "Intelligent Pinyin" is listed in ibus-setup dialog and default ime list.
 * Try to add "Intelligent Pinyin" to ime list, begin to use the new pinyin ime.
 * Get some testing sentences, try to input it under both "Pinyin" and "Intelligent Pinyin", to see whether the "Intelligent Pinyin" can reduce the number of keyboard press.

User Experience

 * With the libpinyin back end, Chinese users can speed up the pinyin input speed by pressing less keys.

Dependencies

 * Need to modify ibus-pinyin code to integrate ibus-pinyin with libpinyin, and upstream accept the patch.

Contingency Plan
libpinyin is already in testing currently in F16 updates-testing but in worst case we could revert to current ibus-pinyin without libpinyin.

Documentation

 * github repo
 * wiki

Release Notes
ibus-pinyin with new libpinyin backend provides a more intelligent pinyin input method and speeds up the user input of Chinese characters.