You can find a dotted-decimal representation of the same syntax diagrams (meant for screen readers) in the DB2 for zSeries online documentation which should be much easier to parse. Unfortunately, getting to the dotted-decimal representation isn't all that easy... here's a quick summary.
- Open the DB2 for zSeries information center in your Web browser.
- Find the statement or command you're interested in.
- View the source of the frame to find <img src="c.gif" alt="Read syntax diagram" longdesc="syntax.htm" border="0" />. syntax.htm will be some long HTML filename that contains the dotted-decimal version of the syntax.
- View the dotted-decimal syntax by copying the frame URL and changing the filename to point to the dotted-decimal filename. The DB2 for zSeries information center uses frames heavily, so the base URL doesn't change to reflect the content of the page you're looking at.
Here's an example of two "normal" syntax diagrams and their dotted decimal equivalents (1 (subselect) and 2 (select-clause)).
These should be much easier to parse if you can get LWP to grab the dotted-decimal syntax files.
Update: Duh, here's a page that describes the dotted-decimal syntax format.
Update 2: If you need the DB2 for Linux, UNIX, Windows version of the syntax, you can install the DB2 8.2 Information Center locally, unzip the doc.zip files in eclipse/plugins/com.ibm.db2.udb.doc/doc.zip, and work with the HTML files directly rather than going through LWP and the Web. Nice way of avoiding the relative URI problems with the framesets, too. Unfortunately, DB2 8.2 for Linux, UNIX, and Windows won't be released for a little while -- in the meantime, if you can still get your hands on the DB2 Stinger beta information center that will also do the job. Versions of DB2 prior to 8.2 don't include the dotted-decimal syntax diagrams.