XML

1
XML

• An introduction in relation to semistructured data

2
Overview

• Background / History
• Basic syntax
• XML and semistructured data
• Document type definitions
• Extensions for XML
• Paraphernalia

3
Overview

• Background / History
• SGML
• SGML, HTML and XML
• World Wide Web Consortium
• Basic syntax
• XML and semistructured data
• Document type definitions
• Extensions for XML
• Paraphernalia

4
Standard Generalized Markup Language (SGML)

• model information exclusively on basis of its
  inner laws and its function
• ? platform independent storage of structured
  information
• standard ISO 8879 from 1986

5
SGML, HTML and XML

• SGML(web application) HTML (is one special
  instance of SGML)
• XML ? SGML

6
Why XML from SGML?

• SGML
• is exceedingly complex and difficult to
  understand
• is formally so complex, that online-applications
  have difficulties to process it in reasonable
  time
• has many properties which were not designed for
  use in network environments (remember that it is
  a standard from 1986)

7
World Wide Web Consortium

• Nov 1996 initial XML draft
• Dec 1997 XML1.0 Proposed Recommendation
• Feb 1998 W3C Recommendation Extensible Markup
  Language (XML) 1.0
• Oct 2000 XML1.0 2nd edition

8
Overview

• Background / History
• Basic syntax
• Elements
• Attributes
• Well-formed XML documents
• XML and semistructured data
• Document type definitions
• Extensions for XML
• Paraphernalia

9
Elements

• element lttaggt content lt/taggt
• lttaggt, lt/taggt markups
• content structures between markups
• no predefined tags
• basic content (no markups) is treated as text
  PCDATA (Parsed Character Data)
• abbreviation for empty elements lttag /gt

10
Example

• ltpersonnelgt
• ltpersongt
• ltnamegt John Cage lt/namegt
• ltfunctiongt Bearer lt/functiongt
• lt/persongt
• ltpersongt
• ltnamegt Elaine Vassal lt/namegt
• ltfunctiongt chief secretary lt/functiongt
• lt/persongt
• lt/personnelgt

11
Attributes

• sometimes called property in data models
• (namevalue) pairs
• value always a string (type NMTOKEN)
• allows building of groups of elements
• ambiguity information as attribute or element?

12
Example

• ltpersonnelgt
• ltperson sexmgt
• ltnamegt John Cage lt/namegt
• ltfunction departmentcivil rightsgt Bearer
  lt/functiongt
• lt/persongt
• ltperson sexfgt
• ltnamegt Elaine Vassal lt/namegt
• ltfunction departmentadmingt chief secretary
  lt/functiongt
• lt/persongt
• lt/personnelgt

13
Well-formed XML documents

• a XML document is well-formed, if
• tags nest properly
• (not ltt1gtltt2gtlt/t1gtlt/t2gt)
• attributes are unique within one element
• (not lttag atta attbgt)

14
Overview

• Background / History
• Basic syntax
• XML and semistructured data
• Simple transformations
• Differences that make transformation more
  difficult
• Additional constructs
• Document type definitions
• Extensions for XML
• Paraphernalia

15
Simple transformations

• with basic XML syntax (no attributes, tree as
  data structure)
• from XML to ssd
• ltpersongt
• ltnamegt John Cage lt/namegt
• ltfunctiongt Bearer lt/functiongt
• lt/persongt
• ? person name John Cage, function
  bearer

16
Simple transformations II

• from ssd to XML (transformation function T)
• T(atomic value) atomic value
• T(l1 v1, , ln vn)
• ltl1gt T(v1) lt/l1gt
• ltlngt T(vn) lt/lngt

17
Differences that make transformation more
difficult

• different semantic of labels
• element or attribute
• order
• mixing elements and text

18
Semantics of labels

• XML
• graphs with labels on nodes

• ssd
• graphs with labels on edges

ltpersongt ltnamegtAlanlt/namegt ltagegt42lt/agegt ltemail
gtab_at_comlt/emailgt lt/persongt
person name Alan, age 42,
email ab_at_com
19
Element or attribute

• ambiguity between representation of information
  as element or as attribute
• ? different possibilities of encoding
• in particular in combination with references
• ltagt ltb ido123gt some string lt/bgtlt/agt
• lta co123 /gt
• or
• lta bo123 /gt
• ltagt ltc ido123gt some string lt/cgtlt/agt

20
Order

• ssd model based on unordered collections
• XML elements are ordered
• but XML attributes are not
• unordered data can be processed more efficiently
• ? for data exchange applications ignore order of
  XML

21
Mixing elements and text

• XML allows mixing of PCDATA and subelements
• lttalkgt
• XML - An introduction in relation to
  semistructured data
• ltspeakergt Sebastian Bitzer lt/speakergt
• lt/talkgt

22
Additional constructs in XML

• comments
• lt!-- comment --gt
• processing instructions
• lt?application-name instruction-textgt
• CDATA (for escaping)
• lt!CDATA markups wont be processed here gt
• entities
• e.g. auml but also external files can be
  declared as entities e.g. a gif-file as pic-1

23
Overview

• Background / History
• Basic syntax
• XML and semistructured data
• Document type definitions
• DTDs as grammars
• DTDs as schemas
• Attributes
• Valid XML documents
• Limitations
• Extensions for XML
• Paraphernalia

24
DTDs as grammar

• document type definition (DTD) serves as grammar
  for underlying XML document
• is precisely a context-free grammar (non-terminal
  ? ordered list of one or more terminals and
  non-terminals)
• can be recursive

25
Definitions

• DTD
• lt!DOCTYPE root-name element-def.s gt
• element-def.s
• lt!ELEMENT name ( content model )gt
• content model
• ordered list of names of elements which can
  occur in the outer element

26
Variations of content model

• lt!ELEMENT r1 (a?, b, c d)gt
• means that elements of type r1 contain
• 0 or 1 a (a is optional) and
• arbitrary many b (0 - 8) and
• either exactly 1 c (c is obligatory)
• or at least 1 d (d is required)
• groups can be build, too
• lt!ELEMENT r2 ((a, b), c?)gt
• means at least one sequence of a followed by
  b comes in front of the optional c

27
DTDs as Schemas

• DTD
• lt!DOCTYPE db
• lt!ELEMENT db ((r1,r2))gt
• lt!ELEMENT r1 ((a,b,c)(a,c,b) (b,a,c) (b,c,a)
  (c,a,b) (c,b,a))gt
• lt!ELEMENT r2 ((c, d) (d, c))gt
• lt!ELEMENT a (PCDATA)gt
• lt!ELEMENT b (PCDATA)gt
• lt!ELEMENT c (PCDATA)gt
• lt!ELEMENT d (PCDATA)gt
• gt
• can be seen as representation for relational
  schema r1(a,b,c), r2(c,d)

28
Declaring attributes

• lt!ATTLIST el.name att.name1 type1 spec1
• att.name2 type2 spec2
• gt
• el.name element which is modified by att.s
• type often CDATA, but also more restricted
  e.g. (mf) for male or female in att. sex
• spec REQUIRED, IMPLIED, FIXED or default value

29
Unique Identifiers

• e.g.
• lt!ATTLIST person id ID REQUIRED
• mom IDREF IMPLIED
• dad IDREF IMPLIED
• children IDREFS IMPLIED
• instance
• ltperson idjohn momjane dadjames
  childrenjack jimgt

30
Valid XML documents

• a XML document is valid, if
• document is well-formed
• additionally has a DTD
• conforms to that DTD
• elements only nested as described in DTD
• just attributes used which are allowed by DTD
• all attributes of type ID must have distinct
  values
• all IDREFS must be to existing identifiers

31
Limitations of DTDs as schemas (summarized)

• order
• only one atomic type (PCDATA, but no INT etc.)
• names are global (partial solution namespaces)
• IDREFs are not constrained to a certain type
  (mother-reference should point to a person)

32
Overview

• Background / History
• Basic syntax
• XML and semistructured data
• Document type definitions
• Extensions for XML
• DCD
• Document navigation
• Paraphernalia

33
Document Content Definitions

• making typing more precise
• seems to be gone
• recent approach XML Schema which must e.g.
• provide for primitive data typing, including
  byte, date, integer, sequence, SQL Java
  primitive data types, etc.
• allow creation of user-defined datatypes, such as
  datatypes that are derived from existing
  datatypes and which may constrain certain of its
  properties
• mechanism for URI reference to standard semantic
  understanding of a construct
• (http//www.w3.org/TR/NOTE-xml-schema-req)

34
XLink XPointer

• pointing to arbitrary positions in documents
• using IDs or relative position
• links can be defined externally to both source
  and target (files)

35
Overview

• Background / History
• Basic syntax
• XML and semistructured data
• Document type definitions
• Extensions for XML
• Paraphernalia
• RDF
• Stylesheets
• SAX and DOM

36
Resource Description Framework

• for representing metadata
• consists of data model and syntax
• simple form edge-labelled graph
• additionally
• containers (bag, sequence or alternative)
• higher-order statements (John says that )

37
Stylesheets

• to specify presentation of data
• Cascading Style Sheets (CSS)
• associate with each element type a presentation
• Extensible Stylesheet Language (XSL)
• specifies the presentation of a class of XML
  documents by describing how an instance of the
  class is transformed into an XML document that
  uses the formatting vocabulary
• http//www.w3.org/Style/XSL/

38
SAX and DOM

• Application Programming Interfaces
• Simple API for XML (SAX)
• standard for parsing
• Document Object Model (DOM)
• interface that will allow programs and scripts to
  dynamically access and update the content,
  structure and style of documents
• compile whole document and build a tree
  representation for it
• http//www.w3.org/DOM/

39
Outlook

• Database issues
• How are we going to model XML? (graphs).
• How are we going to query XML? (XML-QL)
• How are we going to store XML (in a relational
  database? object-oriented?)
• How are we going to process XML efficiently? (uh
  well..., um..., ah..., get some good grad
  students!)

Raghu Ramakrishnan http//www.cs.wisc.edu/cs784-1
/handouts/intro-ssxml.ppt
40
References

• S. Abiteboul, P. Buneman, and D. Suciu, Data on
  the Web. From relations to Semistructured Data
  and XML, Morgan Kaufmann Publishers, San
  Francisco 2000
• H. Lobin, Informationsmodellierung in XML und
  SGML, Berlin, Heidelberg, 2000
• World Wide Web Consortium, Extensible Markup
  Language (XML), http//www.w3.org/XML/