测试无限输入域（如解析器或基于规则的系统）的最佳实践？

Testing a text-to-speech engine is a rather daunting task. The engine itself parses input and applies rules for pronunciation based on phonetic analysis of individual words. In addition exception lists for the pronounciation rules exist to improve the end result. Projects such as Testing a text-to-speech engine is a rather dau