I was planning on grabbing the webmin source (it's really just a glorified tarball), figuring out what the author did, and reimplement (possibly better) in awk and/or python. awk is /vastly/ quicker than any other regex-capable language I've ever used, incl. perl. Using awk to produce an intermediate format, and then Python or another tool for 'interpreting' (human display) is probably a good bet. I /was/ going to consider output to XML, thus allowing one to use xsltproc with the right xsl stylesheet to produce nice html or what-not.
I have a log analyzer for Courier that I've written in Python. One of these days, I'll release it, but I want to get the code to a non-embarrassing state (and to output XML and CSV as well as ASCII).
However, it would probably not meet the requirements stated above. I think going to a "thin" file format would only take a memory problem with heaps and swap space and turn it into a memory problem for the file buffers. You'd probably need some sort of "database" to manage the problem. A homegrown own would probably be too time consuming to engineer and a traditional RDBMS would probably be too heavyweight or thick for your purposes. Perhaps BerkleyDB or GDB would be good.
-andy
------------------------------------------------------- This SF.net email is sponsored by: The Definitive IT and Networking Event. Be There! NetWorld+Interop Las Vegas 2003 -- Register today! http://ads.sourceforge.net/cgi-bin/redirect.pl?keyn0001en _______________________________________________ courier-users mailing list [EMAIL PROTECTED] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
