Namespaces

• XML applications process XML documents according to element and attribute names.

• Unlike HTML, XML allows you to create your own tags. Therefore, in a distributed environment like the Web, documents may contain any kind of element and attribute names.

• An XML application might want to process elements from various types of documents. For example:
  

• In the example above, the elements that define books in the three different bookstores might have a different structure.

• For example, the catalog for City Lights might look like this:

  <?xml version="1.0"?> <catalog> ... <book> <isbn>2-07-071966-9</isbn> <author> <first>Albert</first> <last>Camus</last> </author> <title>Le Mythe de Sisyphe</title> <type>Essay</type> <language>French</language> <publisher>Gallimard</publisher> </book> <book> <isbn>65-01265434-7</isbn> <author> <first>Alejandro</first> <last>Reyes</last> </author> <title>The Gods are Postmodern and Drink Coca-Cola</title> <type>Short Stories</type> <language>English</language> <publisher>Bureau Inc.</publisher> </book> ... </catalog>

• Our virtual bookstore should be able to process information gathered from the City Lights catalog and process its contents correctly along with items from our own catalog.

• Namespaces are a way to do this:

  <?xml version="1.0"?> <catalog xmlns:cl="http://www.citylights.com/xml/catalog"> ... <book isbn="ISBN-0-552-99670-X"> <title>The Mistress of Spices</title> <author>Chitra Banerjee Divakaruni</author> <publisher>Black Swan</publisher> <year>1997</year> <description> A dazzling tale of misbegotten dreams and desires, woven with poetry and storyteller magic. </description> </book> <cl:book> <isbn>65-01265434-7</isbn> <author> <first>Alejandro</first> <last>Reyes</last> </author> <title>The Gods are Postmodern and Drink Coca-Cola</title> <type>Short Stories</type> <language>English</language> <publisher>Bureau Inc.</publisher> </cl:book> ... </catalog>

• Namespaces are a way of creating "universal" names. In the example above, the <cl:book> element is effectively transformed into:

  <{http://www.citylights.com/xml/catalog}book>

• By defining the namespace, our application can detect XML elements that belong to a different schema or DTD and process them accordingly.

• Namespaces are "inherited" by sub-elements. Therefore, all elements within cl:book belong to the cl namespace.

• Multiple namespaces may be defined in the same document.

• The xmlns attribute can appear within any element, not just the root element:

  <book xmlns:cl="http://www.citylights.com/xml/catalog">

• Elements with no namespace are considered "local", not "universal".

• A default namespace can be declared:

  <catalog xmlns:cl="http://www.citylights.com/xml/catalog" xmlns="http://www.keller.com/xml/books"> ... </catalog>

  This declares the cl namespace, and specifies that, for unprefixed elements, the default namespace is http://www.keller.com/xml/books.

• NOTE: The URLs specified do not have to point to any real location. They are simply used as unique strings.

• In an early version of its XML website, Microsoft provided a particularly cogent explanation of the basic concepts of namespaces.

Lab 1G