Hi. My student has been working on a java xml project which is not covered in our syllabus. He is doing pretty good but before I can help him I need some general advice.

He is trying to load the XML elements into a Java List so it can be searched and sorted.

Is this a good approach or not? I actually have never used a List is it the same as an array list or vector?

Opinions welcome not looking for code.

Coneninja

Here is the information on the List Class. From my very cursory glance, I gathered that Lists are like ArrayLists and Vectors, but Lists are designed to be viewed in a Java Application or Applet. They also do NOT make use of generics. Lists take only Strings in their List.add() method.

I'm not going to say his approach is "wrong" per se, maybe creative. But, java.awt.List is part of the (heavy weight) Java GUI package. It's not intended to be use as a general list. In fact, there is no general List class, only a List interface. While XML is a text file, it is not really a flat file, and its nested structure makes it less suited for flat list-type structures. The way I would store XML data would be to create a class that models the data. So, for example, if your XML file is storing data about people, you could create a Person class with attributes String firstName, String lastName, String phoneNumber, and Address address. You would also have and Address class having attributes String street, String city, String state, and String zip. This class structure would correspond to the following XML snippet



This is just one simple example of how XML can be turned into a class and vice versa. In this case, you can cave the Person class implement Comparable<Person> and in the compareTo(...) method you can have set this class's natural order as alphabetical by last name, or whatever you want. Once a class properly implements the Comparable is can be sorted by the Java Collections methods.

Now, you didn't mention the kind of data you're storing in XML files. If it's just a flat list, then using XML is overkill. In that case, just go back to a flat file.

This post has been edited by alcdotcom: 25 Mar, 2007 - 07:00 AM

I have seen two different styles of parsing xml. I know there are others, as I have only looked at replicating unvalidating xml.

The first style I saw doesn't try and store the xml in a structure; but rather reads through the xml and the tags generate events in the program. The program then checks the tag, and if it matches fills out the appropriate attribute to the class.

The second style, uses the recursive nature of xml to parse the xml into a tree.

Personally I think lists would be rather hard to use because xml tends to have multiple layers of associativy. For example in the following xml

The first child is distinct from the child located underneath of it. For example, you are your parents child, and your parents are also the child of their parents; however, you are not your parents. So I guess, if you captured the path of the nodes then it could work. So for the above xml document you would have three entries:

This way all the elements maintain their depth position.

Hope that answers your question.

Salindor

This post has been edited by salindor: 25 Mar, 2007 - 11:13 AM