ChangZhuo Chen (陳昌倬)

I just reduce the second loop to `MAX_PHRASE_LEN` so that O(n^2) become O(n). It is the easiest fix. ``` c for ( begin = 0; begin < pgdata->nPhoneSeq; begin++ )...

As suggestion by [Yuan Chao](https://www.facebook.com/yuanchao), we can do brute force phrasing only when user presses tab at the end of preedit buffer. In other scenarios, we just use DP based...

See #123

The scenario of this API is? I think the easiest way to do so is to create another `ChewingContext` and replace the old one.

@PCMan Just hack an API `chewing_reload` to reload everything in https://github.com/czchen/libchewing/tree/feature/chewing_reload, is that what you need?

`MIN_CHI_SYMBOL_LEN` means the minimal valid input for function `chewing_set_maxChiSymbolLen`. I think it shall be `1` instead of `0`.

This can be implemented by modify CountReleaseNum like [this](https://gist.github.com/4232958). After applying this patch, set maxChiSymbolLen to and you will only have one word in preedit buffer. However, I cannot find...

Need help to reprocedure this one.

The following APIs do not have test cases: - `chewing_cand_CheckDone` - `chewing_cand_ChoicePerPage` The keyboard type other than `KB_DEFAULT` do not have any test cases.

[Andrew Lee](https://www.facebook.com/ajqlee) 在 [MiniDebConf2013](https://wiki.debian.org/DebianTaiwan/MiniDebConf2013) 有分享他目前在做的 [TALOS project](http://talostw.github.io/). 目標是讓原住民語可以在 debian 上使用。或許可以參考他目前已完成的資料。 印象中，原住民的文字是當時教會為了翻譯聖經，用西歐字母組出來了。由於傳教士來自不同的國家，每個原住民語都會有不同的版本。在輸入方面，鍵盤上會有 [Compose key](http://en.wikipedia.org/wiki/Compose_key) 來輔助輸入。這部份不確定是要用特殊鍵盤還是可以用輸入法搞定。 「想學原住民的漢語使用者」這件事，Andrew 目前在教原住民使用 debian，或許可以請他幫忙？不過由於山區網路不穩定，這部份需要 offline 可以使用會比較符合需求。