I have two problems, and I wonder if I can solve them in one go. I am trying to do a natural language date partitioning in Java (well, Scala) and used JChronic, the excellent chronicler RubyGem port.
There are two problems:
JChronic uses java.util.Calendar, not Joda-Time, and I think it's pretty reasonable to say that Joda-Time is or should be a replacement for JDK date libraries. And if Joda-Time does not replace existing date libraries, JSR 310 is confident that someday Oracle will finish the lawsuit and return to Java.
JChronic does not handle general date and time parsing. If I tell him to parse "next Thursday 4pm" or something like that, he will correctly handle it and give me a Calendar object at the right time. But if I just say “2011” or “January 1963” or something like that, he will not be able to handle general date ranges or Partials in Joda —Time to speak.
The second of them is much more worrying than the first. I am trying to extract dates from web pages regarding documents (books, newspaper articles, web pages, etc.), Where publication dates and copyright dates are important.
Currently, I feel like I’ve come to terms with writing my own or perhaps porting an aging JChronic to use Joda-Time and adding support for partial ones. Is there an alternative solution that could satisfy at least (2) and possibly (1)?
source share