XML Problems and Solutions

1
XML Problems and Solutions

• Andy Clark

2
Attributes vs. Elements

• Problem
• I can't decide between using attributes and
  elements.
• Solution
• With DTDs, attributes offer limited "datatype"
  support but are less flexible. XML Schema (among
  other schema languages) have true datatype
  support and can be used for attribute and element
  content.

3
DOM Performance

• Problem
• The DOM tree uses a lot of memory.
• Solution
• Don't use DOM. Instead, use a SAX parser and
  build custom, efficient data structures as
  needed.

4
Java Serialization of DOM

• Problem
• I want to use RMI to serialize my document to a
  remote object but it's not working.
• Solution
• DOM tree implementations are not required to be
  serializable. And since the XML format is a more
  efficient representation than Java serialization,
  use a serializable object that writes XML and
  re-parse the document on the remote end.

5
Relational Databases

• Problem
• I have heard XML is the next "big thing". Do I
  have to throw away my relational database?
• Solution
• All of the major relational database vendors are
  incorporating XML support into their products.
  Your legacy data is safe and will play well with
  all of the new kids in town.

6
Binary Data in XML

• Problem
• I want to embed binary data in my document but
  the parser complains about invalid characters.
• Solution
• The XML specification does not allow all Unicode
  characters to appear in a document. The binary
  data must be encoded using legal XML characters
  (e.g. base64). The application is then
  responsible for decoding the information.

7
Invalid Encoding

• Problem
• I am using ISO Latin 1 characters in my document
  but my parser doesn't like it.
• Solution
• You must specify the proper encoding in the XML
  Declaration line of your XML file. The encoding
  name for ISO Latin 1 is "ISO-8859-1". By default,
  XML parsers assume "UTF-8". Note that ASCII is
  equivalent to UTF-8 chars 0-127.
• e.g. lt?xml version1.0 encodingISO-8859-1?gt

8
Publish as HTML

• Problem
• I want my customers to view an HTML version of my
  XML data. What can I do?
• Solution
• Use XSL stylesheets to transform your XML
  documents into HTML. Using stylesheets separates
  the content from the presentation and makes it
  easy to change the "look feel" of your website
  without changing your data.

9
Browser Support

• Problem
• Can I use the DOM and XSL in browsers on the
  client side?
• Solution
• Some of the latest browsers have support for
  latest DOM and XSL standards (e.g. IE 6.0). Best
  suggestion is to statically generate the pages,
  if applicable, or dynamically generate the HTML
  on the server side (e.g. using servlets).

10
Useful Links

• World Wide Web Consortium (W3C)
• http//www.w3.org/
• XML.com
• http//www.xml.com/

11
XML Problems and Solutions

• Andy Clark