XML Schema Help

gushal · September 27th, 2010, 11:20 AM

Hello everyone,
I am a newbee in XML, and have a problem in XML schema. First I am writing an example from "Beginning XML" from Wrox.

**---------------- name5.xsd ----------------**

Code:

<?xml version="1.0" ?>
<schema xmlns="http://www.w3.org/2001/XMLSchema" xmlns:target="http://www.example.com/name"
targetNameSpace="http://www.example.com/name" elementFormDefault="qualified">   /* ******* Problem 1 ******* */
	<element name="name">
		<complexType>
			<sequence>
				<element name="first" type="string" />	  /* ******* Problem 2 ******** */
				<element name="second" type="string" />
				<element name="third" type="string" />
			</sequence>
			<attribute name="title" type="string" />
		</complexType>
	</element>
</schema>

**---------------- name5.xml ----------------**

Code:

<?xml version="1.0" ?>
<name xmlns="http://www.example.com/name" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://www.example.com/name name5.xsd" title="Mr.">   /* ******* Problem 3 ******** */
	<first>John</first>
	<middle>Fotzgerald Johansen</middle>
	<last>Doe</last>
</name>

**--------------------------------------------------------------**

Problems:

Problem 1: Why is root element named - "schema", and not something else? And, if the URI in the namespace should be unique (I mean, they don't point to any URL or address; they are just unique name), why are we only using - xmlns="http://www.w3.org/2001/XMLSchema"?

Problem 2: type="string". The book says "string" is a datatype. Now type - "http://www.w3.org/2001/XMLSchema\third:string" is again only a unique name (it is not pointing to any URL); so where is the definition of primitive datatype. And if parser knows the definition, then why can't be it "target:string"

Problem 3: What is "xsi:schemaLocation". How does parser know, it can take namespace-location pair

Extra Problem: This one is hard to explain.
<complextype name="nameType">
....content model....
</complexType>
<element name="name" type="target:nameType" />

Now name="nameType" and ' type="target:nameType" ' belongs to different namespaces, then what does "target:nameType" point to ?

Thanks in advance.

joefawcett · September 29th, 2010, 11:50 AM

Quote:

Originally Posted by gushal

Hello everyone,

Problems:

Problem 1: Why is root element named - "schema", and not something else? And, if the URI in the namespace should be unique (I mean, they don't point to any URL or address; they are just unique name), why are we only using - xmlns=http://www.w3.org/2001/XMLSchema?
Because the schema for schemas defines the root element as being named schema.

Problem 2: type="string". The book says "string" is a datatype. Now type - "http://www.w3.org/2001/XMLSchema\third:string" is again only a unique name (it is not pointing to any URL); so where is the definition of primitive datatype. And if parser knows the definition, then why can't be it "target:string"
Because the target namespace applies to the instance document, not the schema document. The primitive datatypes are defined in the schema for schemas document.

Problem 3: What is "xsi:schemaLocation". How does parser know, it can take namespace-location pair
Where the schema is located (not always used by the XML parser). It knows because it's defined that way in the schema's schema and so it's hardcoded into conformant parsers.

Extra Problem: This one is hard to explain.
<complextype name="nameType">
....content model....
</complexType>
<element name="name" type="target:nameType" />

Now name="nameType" and ' type="target:nameType" ' belongs to different namespaces, then what does "target:nameType" point to ?
Don't get this one.

Thanks in advance.

Hope this helps

gushal · September 29th, 2010, 06:17 PM

Thanks for explanation. It's quiet clear now.