XML - Day 2

Admin

POST

1. Are these element names legal? If not, why not?
   • HospitalName
   • hosPitalAddress
   • hospital_director
   • _hospitalWard
   • number of beds
   • 9th_floor
   • Floor_7
2. What information goes on the first line of an xml document?
3. What is the markup for a comment?
4. Start a new file in today's directory called day2post.xml. Write a well-formed xml of 5 lines or less (no need to be creative). How do you know it's well-formed?
5. Leave the parsed xml file on screen to prove you've done it.

• discussion of freeware and commercial editors
  • xmlSpy - screenshot - mindshare, $200/user, email registration
  • Xed - screenshot - unix port, primitive, free
  • xml notepad - class input?
  • other class experiences
• xml document design - 'loser pet shop' example
  • cats
  • dogs
  • wetPets
  • herpPets
  • smallPets
  Some common elements: breed -} name, price, numberOfLegs, suggestedFood?

external DTSs nn|332+

• external DTDs are reusable
• a simple example of an xml document with an external dtd
• external DTDs can be on the same SYSTEM or be PUBLICally available.
• Declaring elements (nn|339):
  • PCDATA: Text not contained in a child element. nn|340-41
  • general format: <!ELEMENT name content> (p.57|338)
  • content = content-model | EMPTY | ANY.
  • Child elements under ANY still must be declared... (works in progress)
  • a content model is like a "regex" (regular expression)
  • Order (aka "Sequences" nn|341): (foo, bar, baz) element foo THEN element bar THEN element baz
  • Alternation (aka "choice" nn|344): (foo | bar | baz) element foo OR element bar OR element baz
  • Grouping ( (foo|bar),(dag|nabbit) ) nn|341
  • Repetition (aka "Occurrence indicators","cardinality" nn|345):
    • ? = 0 or 1 instance of the element
    • + = 1 or more instance of the element
    • * = 0 or more instances of the element
    (foo+, bar?, baz*) at least one instance of element foo THEN one or no instances of element bar, THEN any number (or no) instances of element baz.
  • mixed-content-models nn|340,342: (#PCDATA | foo | bar | baz)* parseable text, THEN element foo, THEN element bar, THEN element baz.
    Note that mixed-content elements are restricted to this content form.
  • can combine internal and external -- the internal is defined first and overrides
  • Ambiguity example: pp60-61|??
  • difficult to specify numbers beyond the wildcards
• Hands-on:
  • make an external DTD for the previous .xml document
  • validate it - add a constraint - break it - change the DTD or xml document
  • here are some constraints to work with
    1. elements in a given order
    2. elements from alternation
    3. element using +
    4. element using ?
    5. element using *
    6. element declaration with mixed content (tricky!)
  • a working example with copy of the DTD
• Declaring attributes nn|349-51:
  • general format: <!ATTLIST element.name attribute.definitions> Why is element's name specified?
  • Attribute defaults nn|356-7
    • #REQUIRED "default text"
    • #FIXED "text" Must be this exactly!
    • #IMPLIED - left up to the receiving app
  • Attribute types
    • strings -
      • CDATA, no built-in entities
    • tokenized - restrict the kind of content
      • ID unique identifier for an element
      • IDREF must be a reference to an existing ID
      • ENTITY/ENTITIES"
      • NMTOKEN/NMTOKENS more strict: [a-zA-Z0-9._-:]
    • enumerated -
      • <!ATTLIST element.name attribute.name (option1 | option2 | etc ) "option2")>
  • Consider this xml fragment:
    <frog color="green"> la la mr. frog</frog>

    and this dtd fragment:
    <!ELEMENT frog (#PCDATA)>
<!ATTLIST frog color #REQUIRED "spotted">

• Hands-on:
  1. add an ID to one of your elements. Validate.
  2. add an IDREF to another element. Validate. Verify it breaks if the IDREF does not reference an existing ID.
• a working example with DTD
• Parameter entities - DTD only
• DTD conditionals - pp 136-7.

Namespace in xml p.139|293-301

• likelihood of collision in an extensible language
• re-use of subdocuments
• namespace as pure extraction
• namespace:name format (c.f. packages or classes)
• problem of unique prefices: solution is to relay on a URI
  • URL: uniform resource locator, familiar to you
  • URN: uniform resource name,
  • declare the namespace in your root element (inheritance):
    <mySpace:myRootElement xmlns:mySpace="http://www.myDomain.com/unique"> (explicit namespace definition, allows multiples and mixed default/explicit) example nn|298-299
    or just <myRootElement xmlns:mySpace="http://www.myDomain.com/unique"> (singular default namespace) example nn|300
    cancel namespace with<myChildElement xmlns:mySpace=""> local to that element and children

Schemas

• DTD is an SGML legacy tool (EBNF grammar) - XML Schema is XML: "schema valid"
• written in XML
• element content validation; nn|385-386.
• able to reuse sections: ElementTypes
• derived datatypes
• associating a document with a schema nn|395
• a partial example
• a simple xml with its schema that uses Microsoft schema. Note the backwards tree construction!

Reading

• CSS - p377-394|61-75
• XSL - 453-477|133
• XSLT nn|133-140

Remember, your login is based on your machine's hostname, not on any other number.
~/[initials] refers to the subdirectory under your homedir, named after your initials. Everything except for .dotfiles will be stored in your ~/[initials] directory.

© 1995-2001 jason carr
Distributed under the terms of the GNU Free Documentation License.