This library adds functions to romanize (convert a different character set into a Latin alphabet phonetic representation) strings from all languages.
It is mainly intended for comparison of Speech Recognition results (which Android returns as Unicode characters) with their Latin romanized form.
Currently, only Chinese characters are supported (in the form of Pinyin).
The Unicode-Pinyin table can be found here.
-
Build this repository into a .aar library
-
In your app's
build.gradlefile, include the following line in thedependenciesgroup:
dependencies { compile(name: 'android-romanization-0.1', ext: 'aar')
3. At any time in your application, initialize the language to load the data
from the Context (included in the library).
```
import life.knowledge4.romanization.PinYinTable;
...
protected void onCreate(Bundle savedInstanceState) {
super.onCreate(savedInstanceState);
PinYinTable.initialize(getContext());
}
-
Use the class
ChineseStringto obtain the string in its romanized form:
ChineseString("人").getRomanized() {"ren2"}
Another possibility is to return the string without the tone indicator:
```
> ChineseString("人").getRomanized(ChineseString.MODE_NO_TONES)
{"ren"}
Please note that, because some characters may have different pronunciations,
and thus may be represented by different Latin strings, the function will
return all possible values. Thus, you shouldn't create a `ChineseString`
instance with too many characters, as they may grow fast.
```
ChineseString("中文").getRomanized() {"zhong1 wen2", "zhong1 wen4", "zhong4 wen2", "zhong4 wen4"}
## TODO (contributions accepted)
* Pinyin diacriticals (ā, á, ǎ, à)
* More intelligent string romanization
* Other alphabets and languages