Xquery Tutorial

Knoblock

Description

Xquery TutorialCraig KnoblockUniversity of Southern California„„„„„„„„„„ReferencesXQuery 1.0: An XML Query Language www.w3.org/TR/xquery/XML Query Use Cases www.w3.org/TR/xmlquery-use-casesXquery Interpreter:http://prometheus2.isi.edu:9001/xquery.htmlAdvice: Case mattersWrite your queries incrementally!It is hard to debug entire queriesExample XML Document: bib.xml StevensW. Addison-Wesley 65.95 65.95 Example XML Document: bib.xml AbiteboulSergeBunemanPeterSuciuDan Morgan Kaufmann Publishers 39.95 GerbargDarcy CITI Kluwer Academic Publishers 129.95 OOOOOODoes the world need a new query language?Most of the world's business data is stored in relational databases.The relational language SQL is mature and well-established. Can ...

Informations

• Publié par: Phuwyug
• Nombre de lectures: 21
• Langue: English

Extrait

Xquery

Tutorial

Craig Knoblock

University of Southern California

„ „ „

References

XQuery 1.0: An XML Query Language „ ro/gRTx/wwww..3query/ XML Query Use Cases „ www.w3.org/TR/xmlquery-use-cases Xquery Interpreter: „ uexq.hryltmuehtemorp//:ptth1/00:9du.esi.is2 „ Advice: „ Case matters „ Write your queries incrementally! „ It is hard to debug entire queries

Example XML Document: bib.xml

<bib> <book year="1994"> <title>TCP/IP Illustrated</title> <author><last>Stevens</last><first>W.</first></author>  <publisher>Addison-Wesley</publisher>  <price> 65.95</price> </book> <book year="1992"> <title>Advanced Programming in the Unix environment</title> <author><last>Stevens</last><first>W.</first></author>  <publisher>Addison-Wesley</publisher>  <price>65.95</price>  </book>

utho></airstr</f sa>t>rl<tuoh >r a<ir<ft>as/lu<ciSu/<>tsrif/<naD>ts </first>st>Serge < > uahtro /<uahtro emun</anla><>Bst>tsretePtsalif<> 9.<5 >93ec >p/ir ok> </bo k ye<boo9991"=ra ltit<  >" Ece The>s icomon uahtro > p<builhser>Morgan KaufmaP nnilburehsp/<sliubersh <p >  ceri srif/<ycraD>tsri<ft>as/lg<arrbGetaoiifil/<faICITion>liataffit> < of tiD roC dnetnogolany  Tofhnec  >l<sa>t<   detiro </title>gital TV oo/b   < e> ib/b   < k> .921>ecicirp/<59  >

Example XML Document: bib.xml

shli>KerwelucarA <  >n ide/>rot bup<</publisher>  <pr edim cuPlbsiehsr ">  < 2000ar="k yeb<oo ast><fireboul</lsa>tbAtituoh>rl< leit <a > Wee /tb<o atht nltitaD>e

Does the world need a new query language?

The Web

Structured Databases

O

O

O

O

Most of the world's business data is stored in relational databases.

The relational language SQL is mature and well-established.

Can SQL be adapted to query XML data? O Leverage existing software O Leverage existing user skills

How is XML data different from relational data?

5

O

O

O

O

O

Nesting

Relational data is "flat"rows and columns

XML data is nestedand its depth may be irregular and unpredictable

Relations can represent hierarchic data by foreign keys or by structured datatypes

In XML it is natural to search for objects at unknown levels of the hierarchy: "Find all the red things."

XPath is a compact and convenient notation for this type of query: //*[@color = "Red"]

6

O

O

Metadata

Relational data is uniform and repetitive

O All bank accounts are similar in structure

O Metadata can be factored out to a system catalog

XML data is highly variable

O Every web page is different

O Each XML object needs to be self-describing

O Metadata is distributed throughout the document

O Queries may access metadata as well as data: "Find elements whose name is the same as their content"

//*[name(.) = string(.)]

7

O

O

O

Heterogeneous Sequences

Relational queries return uniform sets of rows

The results of an XML query may have mixed types and complex structures.

O "Red things": a flag, a cherry, a stopsign, . . .

O Elements can be mixed with atomic values ("mixed data")

XML queries need to be able to perform structural transformations O Example: invert a hierarchy

8

Ordering

O The rows of a relation are unordered O Any desired output ordering must be derived from values

O The elements in an XML document are ordered

O Implications for query:

O Preserve input order in query results O Specify an output ordering at multiple levels O "Find the fifth step" O "Find all the tools used before the hammer"

<book>

<project>

<name>

<step>

<tool>

9

O

O

Missing Information

Relational data is "dense" O row has a value in every columnEvery O A "null" value is needed for missing or inapplicable data

XML data can be "sparse" O Missing or inapplicable elements can be "empty" or "not there" O This gives XML a degree of freedom not present in relational databases

01

Xquery Overview

„

„

Xquery is an expression language

„ Every statement evaluates to some result „ := 5 let $y := 6 return 10*$x+$ylet $x „ Evaluates to 56

Primitive types

„ Number, boolean, strings, dates, times, durations, and XML types