* down and parse cedict

* remove useless

* using third party python pinyin

* jieba as default wordseg

* remove useless

* remove pinyin dict

* using jieba.lcut

* remove doc of cedict egs

* add fan2jian test

* add description for say_digit