0

Hi all i read some answers from other questions like that and didnt obtain a good answer...

So i would like to know some good api's to read and write xml strings. To read i'm thinking to use SAX i already started to write a xmlhandler with a state machine to process it but the handler starts to become big... on method

 public void startElement(String uri, String name, String n, Attributes at)

i have a infinite else if tree, Is this normal and right ?

To write xml strings i have no idea yet...

thanks for your time

Winter
  • 1,896
  • 4
  • 32
  • 41
  • 3
    What do you want to do with the XML? If you're trying to turn XML into Java objects (and/or vice versa) you can just use [JAXB](http://en.wikipedia.org/wiki/JAXB) and not worry about parsing XML. – Matt Ball May 24 '11 at 14:03
  • For writing XML see here: http://stackoverflow.com/questions/4898590/generating-xml-using-sax-and-java/4898623#4898623 – morja May 24 '11 at 14:04
  • Can you show your infinite else if tree? – Andreas Dolk May 24 '11 at 14:04
  • 3
    @Andreas_D: It probably requires infinite space to show ;-) – Olaf May 24 '11 at 14:24
  • Possible duplicate of [Java: How to read and write xml files?](http://stackoverflow.com/questions/7373567/java-how-to-read-and-write-xml-files) – Raedwald Jan 25 '16 at 17:01

1 Answers1

3

The answer depends on what you want to do with your XML. If you want to create Java objects, use JAXB, as Matt Ball suggested.

If you want to do some things (i.e. execute some code) depending on the names of XML elements (or content of their attributes, or whatever), then most likely you will end up with endless if-else. Why? Because you are branching your code depending on what you have in the XML.

You can go around it by:

  • putting each branch inside a class (or a method) and use class loader (or reflection) to construct an instance of it (or call this method);
  • in case of having a finite set of elements you must respond to, having a HashMap<String, Method> and populating it before parsing, then simply invoking method the hash gives you; this implies the use of reflection.

Both solutions negatively affect the speed, but greatly (in my opinion) increase the readability, reusability and understanding of your code.

You might consider using DOM instead of SAX. Drawback: loads the whole XML to the memory; advantage: you can navigate through the tree of elements nicely and always have the context of any element.

As for writing XML: very simple XML documents you can just write, they do not differ from regular text files (open a stream, write to it, close it). However, it is usually better to operate on the DOM tree and then save it; there are many tutorials about it.

If your software is going to be used often, and you rely on the structure of your XML document a lot, and want to avoid nasty crashes caused by misplaced elements, be sure to have a Schema (XSD) or a DTD for it and validate each XML document you load against it. It helps a lot ;)

Miki
  • 7,052
  • 2
  • 29
  • 39