XOM

1
XOM
2
XOM Design Principles

• XOM (XML Object Model) is yet another XML parser,
  by textbook author Elliott Rusty Harold
• The goals of XOM, in order, are
• Correctness
• Only accepts well-formed XML documents
• Only produces well-formed, round-trippable XML
  documents
• Simplicity
• Easy to use Make easy things easy, hard things
  possible
• Easy to learn Minimum number of classes and
  methods uses Java idioms where they make sense
• Performance
• Fast enough Not as fast as SAX, because it uses
  a SAX parser
• Small enough 2 - 3 times smaller than DOM

3
Node types

• Node is an abstract class that extends Object
• All the contents of an XML document are
  represented as Nodes
• There are eight concrete subclasses
• Element
• Document
• Text
• Comment
• Attribute
• ProcessingInstruction
• DocType
• Namespace

4
ParentNode methods

• A ParentNode is a Node that may have children
• There are two subclasses Document and Element
• Some of the methods are
• void appendChild(Node child)
• void insertChild(Node child, int position)
• Node removeChild(Node child)
• Node removeChild(int position)
• int getChildCount()
• Node getChild(int position)
• int indexOf(Node child)

5
Element methods

• Element(String name) // constructor
• void appendChild(String text)
• int getAttributeCount()
• void addAttribute(Attribute attribute)
• Attribute getAttribute(int index)
• Attribute getAttribute(String name)
• Elements getChildElements()
• Elements is another class, with methods
• Element get(int index)
• int size()

6
Creating an XOM document, I

• Element question, yesChild, noChildElement
  node1 new Element("item") question new
  Element("question") question.appendChild("Does
  it have stripes?") yesChild new
  Element("ifYes") yesChild.appendChild("zebra")
  noChild new Element("ifNo")
  noChild.appendChild("horse") node1.appendChi
  ld(question) node1.appendChild(yesChild) node1
  .appendChild(noChild)

7
Creating an XOM document, II

• The portion of the document created so far, if
  printed, looks likeltquestiongtDoes it have
  stripes?lt/questiongtltifYesgtzebralt/ifYesgtltifNogthor
  selt/ifNogt

8
Creating an XOM document, III

• Heres some additional tree creation, including
  an attributeElement node2 new
  Element("item")question new
  Element("question") question.appendChild("Does
  it live in the water?")yesChild new
  Element("ifYes") yesChild.addAttribute(new
  Attribute("color", "green")) yesChild.appendChil
  d("frog")noChild new Element("ifNo") noChild
  .appendChild(node1)node2.appendChild(question)
  node2.appendChild(yesChild)node2.appendChild(no
  Child)

9
Creating an XOM document, IV

• Heres the XML so farltGameDatagt ltitemgt
  ltquestiongtDoes it live in the
  water?lt/questiongt ltifYes
  color"green"gtfroglt/ifYesgt ltifNogt
  ltitemgt ltquestiongtDoes it have
  stripes?lt/questiongt
  ltifYesgtzebralt/ifYesgt
  ltifNogthorselt/ifNogt lt/itemgt
  lt/ifNogt lt/itemgtlt/GameDatagt

10
Creating an XOM document, V

• Now to finish upElement root new
  Element("GameData")root.appendChild(node2)pret
  typrint(root)
• The result is the same as earlier, with
  lt?xml version"1.0" encoding"ISO-8859-1"?gtadded
  at the top

11
Printing the document

• void prettyprint(Element root) Document doc
  new Document(root) try Serializer
  serializer new
  Serializer(System.out, "ISO-8859-1")
  serializer.setIndent(4) serializer.setMaxLengt
  h(64) serializer.write(doc) catch
  (IOException ex) System.err.println(ex)
  

12
Reading and parsing an XML file

• File file chooseFile()try Builder builder
  new Builder() Document doc
  builder.build(file) // Done reading and
  parsing now print preorderPrint(doc.getRootEleme
  nt(), "")catch (ValidityException ex) ...
  catch (ParsingException ex) ... catch
  (IOException ex) ...

13
My preorderPrint method

• private void preorderPrint(Element root, String
  indent) if (root.getChildCount() 0)
  System.out.print(indent root.getLocalName()
  ) System.out.println(" "
  root.getValue()) else
  System.out.println(indent root.getLocalName())
  Elements children root.getChildElements(
  ) for (int i 0 i lt children.size()
  i) preorderPrint(children.get(i),
  indent " ")

14
Making a validating XML parser

• First, specify a DTD
• Second, replace Builder builder new
  Builder() Document doc builder.build(file)
  with Builder builder new Builder(true)
  Document doc builder.build(file)

15
My DTD

• lt!ELEMENT GameData (item)gtlt!ELEMENT item
  (question, ifYes, ifNo)gtlt!ELEMENT question
  (PCDATA)gtlt!ELEMENT ifYes (PCDATAitem)gtlt!ELEM
  ENT ifNo (PCDATAitem)gtlt!ATTLIST ifYes color
  CDATA "brown"gt

16
Getting and installing XOM

• Go to http//www.xom.nu
• Download either
• The Minimal JAR file, xom-1.0.jar, or
• The Complete distribution, zip format which
  contains source code, Javadoc documentation, and
  xom-1.0.jar (3.8MB)
• Use the same way you would any other JAR file
• In Eclipse, this means Project ? Properties... ?
  Java Build Path ? Libraries ? Add External Jars..
• Note This is the only JAR file you need
• Older documentation may say you also need Xerces,
  but a stripped-down version is now included in
  the XOM file

17
Comments

• The XOM API looks large, but it really is simple
  you can do most of what you need with just a few
  classes
• There are almost no convenience methods
• Building a document could be a lot easier
• You can easily write your own convenience methods
• Problem For me, getChildCount() never returns
  zero
• This is to be expected with a nonvalidating
  parser that assumes any whitespace could be a
  child
• I think my problem is that there are some
  children that I dont know about

18
The End