Support text normalization

[csukuangfj]

TODOs

• Add documentation about how to use it
• Add examples about how to generate the rule.fst

Example

#!/usr/bin/env python3

import kaldifst

rule = "./rule.fst"
normalizer = kaldifst.TextNormalizer(rule)
text = "3年前中国总人口为1411778724 人"
out = normalizer(text)
print(out)

The output is given below:

三年前中国总人口为十四亿一千一百七十七万八千七百二十四 人

The rule.fst used in the above example is attached below.
rule.fst.zip

---

Note: The above example uses Chinese numbers, but the implementation is very generic.