Database-Supported XML Processors

Transcription

1 Database-Supported XML Processors Prof. Dr. Torsten Grust Technische Universität München Winter Term 2005/06 Technische Universität München

2 A Word About Myself 2 Torsten Grust Originally from Hannover Student of Computer TU Clausthal Database U Konstanz 1999 Promotion 2000 Visiting IBM, DB2 Everyplace 2004 Habilitation TU Clausthal since 6/2005 TUM Web home:

3 Welcome to this Course... 3 Database-Supported XML Processors? We will use relational database technology to develop a highly efficient, scalable processor for XML languages like XPath, XQuery, and XML Schema. This means that 1 you will get to know these XML technologies quite well, and 2 you can apply and deepen your (rusty?) knowledge of RDBMSs in a challenging, unusual, and fun domain.

4 Relational XML Processing 4 XML Processors Tree Processors This is a course on Relational Tree Processors. Relational Tree Encoding E E Map tree queries into relational queries over tree encodings: Tree E Rel tree query relational query Tree Rel E 1

5 Compiling XQuery To Relational Algebra (1) 5 Input: XQuery Expression Query against an Internet auction database (think ebay): How many auction items are listed in each of the site s [geographical] regions? for $r in doc("auction.xml")/site/regions/* return count($r//item) Tree query: Note how this query uses tree navigation operators / (read: child) and // (descendant) to explore the input XML document auction.xml.

6 Compiling XQuery To Relational Algebra (2) 6 Output: Relational Algebra (MonetDB s Dialect). a0000 := a0004.reverse ().sort ().reverse (); a0000 := a0000.ctrefine (a0003); a0000 := a0000.ctrefine (a0002); a0000 := a0000.mark (0@0).reverse (); a0001 := a0000.leftjoin (a0002); a0005 := a0000.leftjoin (a0004); a0006 := a0000.leftjoin (a0003);. a0003 := count(a0004.reverse ()); a0007 := a0003.reverse ().mark (0@0).reverse (); a0008 := a0003.mark (0@0).reverse ();. [ lines in total... ]

7 Pathfinder (MonetDB/XQuery) 7 For about 3 1 /2 years now, work is underway to design and build the purely relational XQuery processor Pathfinder aka MonetDB/XQuery Joint work with a couple of good guys from (CWI, Amsterdam; U Twente, Netherlands; U Konstanz)

8 Hands On! 8 In a sense, this lecture is an in-depth tour of the techniques and concepts behind Pathfinder. Because Pathfinder has been under development since 2002, the system is already usable and provides an ideal playground for us. Available under the Mozilla OSS License Source code and installers for Linux/Unix, Windows. Please download and use it (and submit bug reports ;-)).

9 These Slides 9 Note: This slide set is developed from scratch during the course of the WS 05/06 semester. I am grateful for any suggestions for improvement as well as reports on typos, glitches, etc. Examples Code/query fragments Definitions Quizzes

10 Further Reading Material the XML standard family: (links marked with are frequently found on the slides) Warning: rather impenetrable on first sight!... on XPath and XQuery: XQuery from the Experts Jonathan Robie et.al. ISBN Addison-Wesley, 2003 The XML Query Language Michael Brundage ISBN Addison-Wesley, various research papers on how database technology can embrace XML, XPath, and XQuery (this is a vivid research area); downloadable from the course web page.

11 Further Reading Material 11 Easily digestable introductions to XML, XPath, and XQuery: The Annotated XML Specification Chapter XPath of XML in a Nutshell (O Reilly) XQuery: A Guided Tour xquery/docs/katz_c01.pdf

12 Organisatorisches 12 Termine Zeit Ort Vorlesung Mo, 9:15 10:00 MI HS 2 Vorlesung Di, 14:15 15:45 MI HS 2 Übung Do, 10:15 11:45 MI Homepage + Forum zur Vorlesung Folien [PDF] oft erst kurzfristig zum Download verfügbar (ca. 2 Stunden vor Termin).

13 Wie profitiert man von dieser Vorlesung? 13 Übungsaufgaben und Klausuraufgaben werden sich sehr ähneln. Aktiv dabei sein! Übungen starten diesen Donnerstag (20. Oktober) Klausur zum Ende des Semesters bestehen. Das Forum zu dieser Vorlesung nutzen! Jens Teubner und ich lesen/schreiben mit. Sprechstunde nutzen Immer, wenn die Türen zu unseren Büros (MI und ) offen stehen. Effektiv sind das 100 % unserer Anwesenheitszeiten.

14