eXtensible Markup Language:

eXtensible Markup Language:
The Least You Need To Know

Rohit Khare

Adam Rifkin
4K Consulting

Who We Are

Rohit Khare

UC Irvine, MCI Telecommunication,
World Wide Web Consortium/MIT, Caltech

rohit@uci.edu

Adam Rifkin

Caltech, Hewlett-Packard, Rome Labs (USAF), NASA/Langley, College of William and Mary

adam at xent dot com

eXtensible Markup Language:
The Least You Need To Know

The Origin of Document Species

The Ascent of XML

XML: Specifications

XML-LANG, XML-LINK, CSS & XS Styles

XML: Applications

Why Every Bit You Write Should Be XML

Examples...

The Origin of (Document) Species

To be presented at
WWW7, Brisbane, Australia

The Origin of
(Document) Species

The Document Ecology

Evolutionary Adaptations of:

Syntax SGML

Style CSS/XSL

Structure HTML

Semantics XML

The Fossil Record

The Document Ecology

The World Wide Web is the "universe of network-accessible information" [Tim Berners-Lee, 1996]

Openness and Content-Neutrality of Documents

HTTP can adapt to any document format

URL can represent links to any document format, from within many

"Natural Selection" Favors a Few Document Formats

Preferential adoption of SGML, CSS, HTML, and now XML

Each embodies the evolutionary strategy of parsimony

Evolution: Capture Info --> Represent Knowledge

Can leverage Web reflexively to capture structure and semantics

XML-based document formats represent an ecosystem of interdependent (rather than competing) document "species"

Evolution: Syntax

Issues of Concrete Representation

Binary (machine) vs. Text (human) formats

Mission-specific vs. Generic formats

Context-free vs. Turing-complete formats

From Turing-Complete to Declarative

Evolution: Style

Externalized Formatting over Embedded Directives

{\keepn\par\pard\sb240\sl-264 \b1\hyppar0 Warning: do not …}

<P CLASS="WARNING"> …
WARNING { font : bold } ...

<?XML-stylesheet TYPE="text/css" HREF="warning.css"?> <WARNING> …
WARNING { font: bold } ...

Cascading to eXtensible Style Sheets

Rendering: displays, Braille, audio, ...
From Inline Formats to Style Sheets

Evolution: Structure

Anatomy of a Newspaper Article

Logical: headline, byline, body, footer, …

Descriptive: bold, italic, indented, …

Declarative: title, address, keyboard-input, …

Custom declarative: <dateline>, <byline>, ...

Automatic Information Collection

Resource Discovery: For providing useful hints to search engines

Classifying: For cataloguing information content and relationships

Content Rating: For aggregation and filtering

Knowledge Codifying, Sharing, and Exchanging: For processing

From Presentational to Declarative

Evolution: Semantics

How well does the document support the potential uses of its contents?

Scenario: To-Do List Manager

Natural-language Scratchpad

HTML Definition List (Datebook)

XML <DEADLINE AT="iso-date"> Element

Composable, Networked DTDs

Reuse by Linking

From Operational to Well-defined

Citation Encodings (1/2)

Presentational Text:

"XML, Java, and the Future of the Web", Jon Bosak, World Wide Web Journal, 2(4):219--228, (1997)

J. Bosak, World Wide Web Journal, "XML, Java, and the Future of the Web", Autumn 1997, Vol. 2, No. 4, pp. 219--228.

Presentational HTML:

<I>XML, Java, and the Future of the Web</I>,<I>World Wide Web Journal</I>,<TT>Jon Bosak</TT>,<B>2(4):219-228</B>, <I>1997</I>.

Structural HTML:

<CITE>XML, Java, and the Future of the Web</CITE>
<H3>World Wide Web Journal</H3>
<H4>Jon Bosak</H4>
<UL> <LI> 2(4):219-228
<LI> 1997
</UL>

Citation Encodings (2/2)

Customized XML:

<BIB><TITLE>XML, Java, and the Future of the Web</TITLE>
<JOURNAL>World Wide Web Journal</JOURNAL>
<AUTHOR> <FIRSTNAME>Jon</FIRSTNAME> <LASTNAME>Bosak</LASTNAME></AUTHOR>
<VOLUME>2</VOLUME> <NUMBER>4</NUMBER> <YEAR>1997</YEAR> <PAGES>219-228</PAGES> </BIB>

XML Metadata Format, using DTDs from the Web:

<?namespace href="http://library.org/bibliography-info" as="BIB"?>
<?namespace href="http://w3.org/schemas/rdf-schema" as="RDF"?>
<?namespace href="http://oclc.org/DublinCore/RDFschema" as="DC"?>
<RDF:serialization> <RDF:assertions href="http://assertions.org/bib-doc">
<BIB:TITLE href="http://sunsite.unc.edu/pub/sun-info/standards/xml/why/xmlapps.htm">
XML, Java, and the Future of the Web </BIB:TITLE>
<BIB:JOURNAL href="http://www.w3j.com/">World Wide Web Journal </BIB:JOURNAL>
<DC:Creator>Jon Bosak</DC:Creator>
<BIB:VOLUME>2</BIB:VOLUME> <BIB:NUMBER>4</BIB:NUMBER>
<BIB:YEAR>1997</BIB:YEAR> <BIB:PAGES>219-228</BIB:PAGES>
</RDF:assertions></RDF:serialization>

The Fossil Record

The Ascent of XML

Joint with Dan Connolly
Fall 97 Web Journal

World-Wide Markup Language

"The HyperText Markup Language is an SGML format." [Tim Berners-Lee, 1991, "About HTML"]

Standard, Open Document Formats since the 1960s (GCA’s GenCode, IBM’s GML)

SGML (ISO 8879:1986) standardized DTDs, doc validity proofs, complex structuring, extensions

1990: Berners-Lee picked tags from a sample SGML DTD and added "killer feature": links

World-Wide HTML tagset embodies 80/20 rule

XML allows Community-Wide Markup Languages,
combining SGML’s power with HTML’s simplicity

Mission of XML

XML was designed to provide an easy-to-write, easy-to-interpret, and easy-to-implement subset of SGML. It was not designed to provide a "one Markup Language fits all" DTD, or a separate DTD for every tag. It was designed so that certain groups could create their own particular markup languages that meet their needs more quickly, efficiently, and (IMO) logically. It was designed to put an end once and for all to the tag-soup wars propagated by Microsoft and Netscape.
---- Jim Cape, June 3, 1997, in a post to

comp.infosystems.www.authoring.html

Community-Wide Markup Languages

Community size is inversely proportional to shared context

Millions agree that <B> means bold, but 2/6/98 reflects local culture

XML decentralizes control of specialized markup languages,
making it cost-effective to capture community ontologies

HTML is not unilaterally extensible (new tags potentially have ambiguous syntax, style, and semantics)

XML is a strict (but simplified) subset of SGML, offering:

Extensibility -- can define new elements, containers, attribute names

Structure -- a DTD can constrain the information model of a document

Validation -- every document can be validated; also, well-formedness can establish conformance to the structure mandated by the DTD

XML includes extensible linking and style formatting also

"Node content must be left free to evolve." [Tim Berners-Lee, 1991, "About Document Formats"]

Evolution of a Standard

W3C Role has Evolved

First guerillas (Generic SGML), then group

Now a movement (GCA, XML ‘98 in March)

Fall ‘97 WWWJ - http://www.w3j.com/xml/

W3C Site is EXTENSIVE

Timeline, Publications, WG Notes, Bookmarks
http://www.w3.org/XML/

List of All Decisions Ever Made, c. Dec ‘97
http://www.w3.org/XML/9712-reports.html

Other Useful XML Links

Our XML Links and Papers Page

http://www.ifindkarma.com/attic/local/xml.html

Robin Cover's Extensive XML Page

http://www.sil.org/sgml/xml.html

Peter Flynn's XML FAQ

http://www.ucc.ie/xml/

XML-Dev Jewels

http://www.vsms.nottingham.ac.uk/vsms/xml/jewels.html

Future Evolution of XML

Coevolution of HTML and XML

All XML is SGML

All HTML is SGML

But all HTML is not XML...

XML profiles SGML:

No Markup Minimization

No Optional Features

Technical Corrigendum harmonizes SGML86

XML: Specifications

Details, Details, Details…

XML-LANG

Self-description with Document Type Defn (DTD):

Structural rules of the document’s markup

External entities, internal entities, non-XML resource entities

Non-Minimization (All Markup is Explicit)

XML Processor - customizable info structures

Most XML users will not know they are using a DTD!

XML-aware software works with custom XML apps without help

Document Validity = Conforms to DTD

Document well-formedness = structurally sound, but no DTD

Required Markup Declaration (RMD) indicates need for DTD

Well-formedness for delivery format, validity for authoring

XML-LANG: Markup Types

Elements - start-tags, end-tags, empty tags

<joke>Take my XML. Please.<applause/></joke>

Attributes - name-value pairs in tags

Entity References - spec chars, macros, external

Comments - not passed along to application

Processing Instructions - passed along to app

<?pi-target-name pi-data?>

CDATA Sections - parser ignores markup

*p = &q; b = (i <= 3);

XML-LANG: DTDs (1/2)

Element Declarations - name and content model

<!ELEMENT email (from, to+, cc*, subject, body, sig?) >
<!ELEMENT body (#PCDATA | image)* >
<!ELEMENT image EMPTY >

<!ELEMENT node (desc, node*)>
<!ELEMENT desc (#PCDATA)>

Attribute Declarations - which elements may have what attributes, default and possible values

<!ATTLIST joke
name ID #required
label CDATA #implied
status ( funny | notfunny ) ‘funny’ >

XML-LANG: DTDs (2/2)

Attribute Declarations - name, type, default

Attribute types: CDATA, ID, IDREF/IDREFS, ENTITY/ENTITIES, NMTOKEN/NMTOKENS, a list of names

Default values: #REQUIRED, #IMPLIED, "value", #FIXED "value"

Entity Declarations - associate name with chunk

Internal entities: < > & ' "
or locally defined: &fork; <!ENTITY fork "4K Consulting">

External entities: <!ENTITY forkfooter SYSTEM "fork/footer.xml">
<!ENTITY pic SYSTEM "http://www.xent.com/4k.gif" NDATA GIF87A>

Parameter entities: <!ENTITY % html.ver "-//W3C//DTD HTML 4.0//EN">

Notation Declarations - external binary data

<!NOTATION GIF87A SYSTEM "GIF">

XML-LINK

Start with HTML, add HyTime and TEI concepts

Simple links point to a single target resource

URL schemes: ftp, http, file, mailto, telnet, nntp, ...

Links can have roles (machine-processible) and human-readable labels associated with them

Can specify default behavior of a link

SHOW - embed in current context, replace it, or start new one

ACTUATE - user must take action (or not) before anything done

Locators - # separates resource name and part id

Connectors - | to return only relevant element

XML-LINK: Advanced

Extended links - can be multidirectional, need not live in the resources they point, link groups

<related-term-group>Hamlet
<related-term HREF="Othello.xml"/>
<related-term HREF="KingLear.xml"/>
<related-term HREF="Macbeth.xml"/>
</related-term-group>

<!ELEMENT related-term-group (#PCDATA | related-term)* >
<!ATTLIST related-term-group
XML-LINK CDATA #FIXED "EXTENDED"
INLINE CDATA #FIXED "TRUE"
CONTENT-ROLE CDATA #FIXED "RT"
… >

Extended Pointers - locate resource by traversing the element tree of its containing document

XPTRs allow links without modifying the containing document

XML-STYLE

CSS1 - rules in terms of element names, Ids, …

XS - based on DSSSL (ISO/IEC 10179:1996)

Formatting specification derived from active style sheets

With XML doc structure, create flow objects (paragraphs, tables, ...) with characteristics (font-name, font-size, …)

Merging, flow object tree determines document layout

Scheme - core expression (math) language in XS

Construction rules declare element style

font-size: 18pt, first-line-start-indent: 20pt, quadding: left

Typically, formatting declared as functions

(element CODE (UNDERLINED-PHRASE))

XML: Applications

July 1997 IEEE Internet Computing

Fall 1997 Web Journal

Data Archaeology

Meishi, or Business Cards

Different shapes, sizes, scripts, demarcations

Two sided, magnetic, photos, public keys

Airline Passenger Name Records

"NQSS5A" means something to airline

Must be manipulated throughout its lifetime

Need stable data format, stable grain of exchange, and common definitions

Every Bit in XML

Across Time

Save Self-Description with Data State

"Future Proofing" of Documents and Data

Across Space

For Exchange as well as Storage of Data

Across Organizations

Building Community-Specific Ontologies

Key to Knowledge Representation

Across Time

Tensions Create Brittle Data Formats

Inertia, Efficiency, Tools, and Context

Future-Proofing Strategies

Machine-Readable - for parsers & generators

Human-Readable - robustness and simplicity

Self-Descriptive - for extraction and validation

XML as Basis to Execute Strategies

e.g., Capturing Database Schema as DTDs

Across Space

Tradeoffs of Marshaling Data

Distributed Systems with Centralized Data

Need: Decentralized, Isolated Subsystems

Strategy: Defer Marshaling Decisions

The Web’s Lesson: Link Instead

Download Networked Resources as Needed

Leveraging the XML-LINK Model

SHOW and ACTUATE; XPTR identifiers

Across Cultures

Challenges to Collaboration

Organizations Defined by their Language

Ontological Problem: Matching Vocabulary

Strategy: Use Documents

"Put It In Writing" Defining Common Terms

Popular Ontologies Can Emerge Organically

Solution: XML-enhanced Documents

Let a Thousand DTDs Bloom on the Web

XML Example Applications

Mathematical Markup Language

Chemical Markup Language

Microsoft’s CDF, Netscape’s MCF

Web Interface Definition Language (WIDL) - Automation

Resource Description Format (RDF) - Composition of schema

Some Software Tools

JSXML - Java Object Stream to XML packages

http://www.camb.opengroup.org/~laforge/jsxml/

Lark - a non-validating XML parser in Java

http://www.textuality.com/Lark/

MSMXL - Microsoft’s XML parser in Java

http://www.microsoft.com/xml/

NXP - a validating parser in Java

http://www.edu.uni-klu.ac.at/~nmikula/NXP/

Many, many others

http://www.ifindkarma.com/attic/local/xml.html

Anatomy of a DTD

From Richard Light,
Presenting XML

Anatomy of a Memo DTD

Creation of Linkable Memos

Name of sender and recipients, date sent, main heading, optional subject keywords, text

Text can contain sections/subsections with optional headings, paragraphs, lists, images and diagrams, quotations from other memos

<!ELEMENT memo (header, links, text)>
is a good start for the DTD, but then what?

Adding Header Declaration

<!ENTITY % common-elements "list | image | quote | anchor">
<!ENTITY % a.global "ID ID #IMPLIED">

<!ELEMENT header (sender+, recipient+, date, main-heading, subject*)>
<!ATTLIST header %a.global;>
<!ELEMENT sender (#PCDATA)>
<!ELEMENT recipient (#PCDATA)>
<!ATTLIST recipient %a.global;>
<!ELEMENT date (#PCDATA)>
<!ATTLIST date %a.global;
SORTFORM CDATA #IMPLIED>
<!ELEMENT main-heading (#PCDATA)>
<!ELEMENT subject (#PCDATA)>

Elements within the Text

<!ELEMENT text (section | p | %common-elements;)+>
<!ATTLIST text %a.global;>
<!ELEMENT section (head?, (section | p | %common-elements;)+)>
<!ATTLIST section %a.global;>
<!ELEMENT p (#PCDATA | %common-elements;)*>
<!ATTLIST p %a.global;>
<!ELEMENT list (item)+>
<!ATTLIST list %a.global;
TYPE (bullet|number|plain) "bullet">
<!ELEMENT item (#PCDATA)>
<!ATTLIST item %a.global;>
<!ELEMENT image EMPTY>
<!ATTLIST image %a.global;
SRC ENTITY #REQUIRED
DESC CDATA #IMPLIED
ALIGN (TOP|MIDDLE|BOTTOM) "middle">
<!ELEMENT anchor (#PCDATA)>
<!ATTLIST anchor %a.global;>


<!NOTATION GIF SYSTEM "http://www.viewers.org/gview.exe">
<!NOTATION JPEG SYSTEM "http://www.viewers.org/jview.exe">

Quotation Links: Simple to Extended

<!ELEMENT quote EMPTY>
<!ATTLIST quote
XML-LINK CDATA #FIXED "simple"
ROLE CDATA #IMPLIED
HREF CDATA #REQUIRED
TITLE CDATA #IMPLIED
SHOW (EMBED|REPLACE|NEW) "EMBED"
ACTUATE (AUTO|USER) "AUTO"
BEHAVIOR CDATA #IMPLIED>

<!ELEMENT quotation (citation, source+)>
<!ATTLIST quotation
XML-LINK CDATA #FIXED "extended"
ROLE CDATA #IMPLIED
TITLE CDATA #IMPLIED
SHOW (EMBED|REPLACE|NEW) "REPLACE"
ACTUATE (AUTO|USER) "USER"
BEHAVIOR CDATA #IMPLIED>

Out-of-Line Links

<!ELEMENT citation EMPTY>
<!ATTLIST citation
XML-LINK CDATA #FIXED "LOCATOR"
ROLE CDATA #FIXED "citation"
HREF CDATA #REQUIRED
TITLE CDATA #IMPLIED
SHOW (EMBED|REPLACE|NEW) "EMBED"
ACTUATE (AUTO|USER) "AUTO"
BEHAVIOR CDATA #IMPLIED>

<!ELEMENT source EMPTY>
<!ATTLIST source
XML-LINK CDATA #FIXED "LOCATOR"
ROLE CDATA #FIXED "source"
HREF CDATA #REQUIRED
TITLE CDATA #IMPLIED
SHOW (EMBED|REPLACE|NEW) "REPLACE"
ACTUATE (AUTO|USER) "USER"
BEHAVIOR CDATA #IMPLIED>

Linking Example: A Link Farm

DTD with quotation, source, citation, and the following links cover element:
<!ELEMENT links (quotation)*>

<?XML version="1.0"?>
<!DOCTYPE links SYSTEM "links.dtd">
<links>
<quotation>
<citation HREF="memoex2.xml#a1"/>
<source HREF="memoex1.xml#p1"/>
</quotation>
<quotation>
<citation HREF="memoex2.xml#a2"/>
<source HREF="memoex1.xml#p7"/>
</quotation>
</links>

Making the Memo DTD Links-Aware

In the memo DTD, we need an extended link group with a single, fixed document locator pointing to the link farm.

<!ELEMENT links (memolinks)>
<!ATTLIST links
%a.global;
XML-LINK CDATA #FIXED "group">
<!ELEMENT memolinks EMPTY>
<!ATTLIST memolinks
XML-LINK CDATA #FIXED "document"
HREF CDATA #FIXED "memo-links.xml">

Sample Memo XML-file

<?XML version="1.0"?>
<!DOCTYPE memo SYSTEM "memo.dtd">
<memo>
<header>
<sender>Adam Rifkin</sender>
<recipient>Rohit Khare</recipient>
<date sortform="19980114">January 14, 1998</date>
<main-heading>XML: The Least You Need to Know</main-heading>
</header>
<links><memolinks/></links>
<text>
<p id="p1"><anchor id="a1"></anchor>I’m almost finished with the slides.</p>
<p id="p2"><anchor id="a2"></anchor>Call me at (626) 449-4123 in an hour.</p>
</text></memo>

Adding Style

<!doctype style-sheet PUBLIC "-//James Clark//DTD DSSSL Style Sheet//EN">
(define memo-style (style font-family-name: "Arial"))
(define para-style (style line-spacing: 13pt font-size: 12pt))
(define header-line-style (style line-spacing: 17pt font-size: 16pt))
(define heading-style (style font-size: 24pt line-spacing: 26pt quadding: 'center font-weight: 'bold))
(define (para-with-prefix prefix)
(make paragraph use: header-line-style
(make line-field field-width: 50pt (literal prefix)) (process-children-trim)))
(default (process-children-trim))
(element MEMO (make sequence use: memo-style (process-children-trim)))
(element P (make paragraph use: para-style space-before: 10pt))
(element main-heading (make paragraph use: heading-style space-before: 20pt))
(element sender (para-with-prefix "From:"))
(element recipient (para-with-prefix "To:"))
(element date (para-with-prefix "Date:"))
(element sender (para-with-prefix "From:"))
(element recipient (para-with-prefix "To:"))
(element date (para-with-prefix "Date:"))
(element links (empty-sosofo))
(element anchor (make paragraph (literal "[fragment of original message was here]")))

XML Editors and Browsers

Useful XML Editor Features

Data validation (with external database)

Insertion of current date on request

Automatic assignment of unique ID to each element

Automatic insertion of elements and attributes required by DTD

Automatic insertion of element data when "reply" button hit

Creating out-of-line links in a user-friendly manner

Useful XML Browser Features

Support for Xpointers (including traversal)

Built-in Support for XS style sheets

Rendering XML fragments embedded in other XML documents

The Philosophy of XML

Future of DTD’s: Maybe…

XML Data - XML all the way down

Metadata collapses into Meatdata

DTD’s become more composable

Legally cascadable DTD’s: RDF

"XML gives Java something to do." [Jon Bosak, 1996]

DOM, XAPI-J, and XML Typing, to name a few

XML: Recommended Reading

XML Books:

Richard Light, Presenting XML, Sams.Net, August 1997.

Web Journal - XML issue, O’Reilly, Fall 1997, http://www.w3j.com/xml/

XML Papers from Khare/Rifkin:

X Marks the Spot: An Introduction to XML http://www.ifindkarma.com/attic/papers/xml/x-marks-the-spot.html

The Ascent of XML (with Dan Connolly) http://www.ifindkarma.com/attic/papers/xml/ascent-of-xml.html

Capturing the State of Distributed Systems with XML http://www.ifindkarma.com/attic/papers/xml/xml-for-archiving.html

The Origin of (Document) Species
http://www.ifindkarma.com/attic/papers/www/origin-of-species.html

XML - World Wide Web Links Galore
http://www.ifindkarma.com/attic/local/xml.html