EXANDA - A web-based utility to explore and analyze data
Technical overview
Exanda uses common open source software like Tomcat (Java servlet container),
Cocoon (web development framework by Apache),
MySQL (database system), and
Apache (web server).
Central principles:
- Modular structure for flexibility and variety in use.
- Flexible output management to generate various output formats.
The Cocoon pipeline concept is used to integrate following steps of the process (Cocoon is running as servlet in Tomcat):
- The Cocoon components XSP (extended server pages) and
ESQL
(XSP logic sheet for SQL) are used to implement the logic. The XSP templates generate Java programs, which connect by JDBC to the database system and query the database tables. The database system MySQL is used as a computation engine. Any database system compliant to SQL (full ANSI SQL-92) could be used for this purpose.
- The results are produced first in an intermediary XML format which uses mainly:
- The content-oriented XHTML 2.0 table model for the result tables. The semantic meaning of each table cell is described by the headers attribute.
- The DDI format for the primary information on the study and on the selected variables.
- Transforming the intermediary XML output format according XSL style sheets generates the various target output formats.
Output formats:
- HTML (XHTML 1.1) for display in web browsers.
- FO (formatting objects) to produce
PDF.
- GMR
(Gnumeric file format) to produce XLS spreadsheets for further processing in Microsoft Excel or Open Office.
- SVG (scalable vector graphics) for display in e.g. Adobe SVG viewer and to produce
PNG (portable network graphics).
Planned features
- Configurable site-specific header and footer for the user interface.
- Integration of string variables of survey data.
- Caching of the dynamic generated results (reverse Squid cache server).
|
