WWW7: a comprehensive introduction to XML

by james tauber presented at QUT (Queensland University of Technology)

status of the various specifications related to XML:

• XML recommendation
• XML namespaces working draft
• XLL working drafts (XLink and XPointer)
• XSL note

limitations of HTML:

• extensibility (no own tags)
• structure
• validation (correct syntax not enforced)

benefits of separation of structure and presentation:

• multiple delivery formats
• information reuse
• multiple sources of information
• consistent presentation
• reduce time spent fiddling with presentation
• applications can make use of documents

how XML came about as a solution:

"some people" realized that a limited set of tags can't be the solution. what we need must be:

• a simple but extensible solution
• a subset of SGML
• simple to implement
• designed for the web
• powerful linking
• style sheets with scripting
• familiar to HTML authors

XML design goals:

• straightforwardly usable over the Internet
• support a wide variety of applications
• compatible with SGML
• easy to write programs which process XML documents
• the number of optional features in XML is to be kept to the absolute minimum, ideally zero
• XML documents should be human legible and reasonably clear
• the XML design should be prepared quickly
• the design of XML is formal and concise
• XML documents must be easy to create
• terseness in XML markup is of minimal importance

the next version of PERL will have XML support built in.

the anatomy of an XML document:

• made up of entities
• parsed data (XML text) or unparsed data (not XML: text or binary, has a notion with it)

entity structure:

• storage units (files or data from database)
• may contain parsed and non-parsed data
• piecing together a document with multiple parts
• reuse components (footer etc)
• all entities but document entity have names

types of markup

• start-tags: <NAME ATTR="value"> (quotes or double quotes required)
• end-tags: </NAME>
• empty elements: <NAME></NAME> or <NAME/>
• entity references: &NAME; (allows to "include" text or text fragments)
• character references: &#decimal-number; or &#xhexadecimal-number;
• comments: <!-- any text but not two hyphens -->
• CDATA: <![CDATA[this is not markup]]>
• processing instructions: <?target instructions?>, whereas target might be an application such as SQL

XML declaration:

document prolog: <?xml version="1.0" encoding="UTF-8"?>

document type declarations:

<!DOCTYPE NAME SYSTEM "name.dtd"> ("external subset")
can also be included into the document ("internal subset"), which has preceedence over the external declarations.

XML documents have to be well-formed (syntactically correct e.g. not intermixed tags) and to be valid (has to have a document type definition and must comply to that).

document type definitions:

• markup declarations
• element declarations are implemented through the so called "content model" which defines entities made out of (recoursive) content particles. particles can be a sequence (number of attributes that must appear in this order) or choice of (zero or many, one or many, zero or one) particles.
• attribute lists declarations: <!ATTLIST NAME declarations> (values can be required or optional, default values can be provided)
• entity declarations: <!ENTITY NAME value> (there are a number of pre-definied entities such as &lt; etc)
• notation declarations: <!NOTATION NAME SYSTEM "specifications">
• processing instructions
• comments
• conditional sections, e.g. draft versus final version of the same document

character entities:

• character repertoire: unicode (16 bit)
• character encoding: UTF-8 (variable length: 8 bit for ASCII, 16 bit for 2-byte unicode, more bytes if required) or UTF-16
• character classes: characters, name characters, NMTOKEN (name token), NAME (must start with a letter or an underscore)
• white spaces are one or more SPACE, CR, LF or TAB
• multiple end of lines will be normalized to one line break
• language identification: xml:lang="language code" (e.g. ISO 639 language code)

XML processors:

• not all implementations require valid documents
• most are implemented in Java
• run either on the client or on the server

XML namespaces:

how can two definitions with the same name be unique ? define a namespace based on a unique URI: <?xml:namespace URI PREFIX ?>. use qualified names to refer to such definitions (PREFIX:NAME)

XLL: the eXtensible Linking Language:

XLink: based on HTML, HyTime and TEI

in HTML:

• the link is expressed at one of its ends
• can only be travelled from that end
• the behaviour of a link is impossed by the browser
• a link goes only to one direction

types of links <A xml:link attributes>xxx</A>

• locators: URI (as in HTML)
• connectors: URI#XPointer (similar to HTML) or URI\XPointer
• queries: URI?XML-XPTR=XPointer

links attributes:

• inline vs outline
• multi-directional links
• links to multiple resources

the target document can either replace the current document, it can be included in the current document or it can appear in a different window (SHOW attribute). a link can be followed automatically or on user action (ACTUATE attribute).

XSL: eXtensible Style sheet Language:

XSL is written in XML and is a subset of DSSSL. a stylesheet associates formating objects with elements in an XML document to produce formatted output.

the XSL scripting language is based on ECMAScript (which is the standardized version of JavaScript). a number of built in functions allow to do computations to achieve things such as automatic numbered list generation etc.

to the main page of this WWW7 trip report.

production note: these session notes were taken live on a SHARP HC-4500A running Windows CE with Pocket Word. they were then transfered to a notebook running Windows 95 and were slightly reformatted using HoTMetaL V4.0. this document is supposed to be HTML V4.0 compliant.

tutorial_xml.html / 14-apr-1998 (ra) / reto ambühler