Introduction - If you have any usage issues, please Google them yourself
For Java-word analysis, the mainstream approach is to use middleware to visit, but you need to use JNI Apache and Microsoft to develop the POI, but POI though powerful, but it is also a lot of trouble with the 3.5 version and the previous version 3.0 to modify a lot of , but not yet stabilized. For general applications, such as mobile email like QQ, only need to read the text, we can directly use tm-extractors, of course, it is also a call to the underlying Apache POI