Felix Karpfen <[EMAIL PROTECTED]> writes:
> I have installed the namazu package and its subsidiaries and also
> mknmz-wwwoffle.
>
> And, as far as I can tell, everything works - including (after adjusting
> the paths) the namazu scripts supplied in the WWWOFFLE package.
Yes, that looks like a silly mistake on my part, the path to a
temporary copy of mknmz-wwwoffle was fixed in the script.
> Calling the supplied "lasttime" script creates an index of the keywords
> of relevant webpages in about 50 seconds and using namazu from the
> command line to search the said index works perfectly.
This is good news, part of the process is complete succesfully.
Does the log file in /var/spool/wwwoffle/search/namazu confirm that it
was successful in indexing?
> However, when I attempt to do a search via the link in the WWWOFFLE
> Cache Search page, the called link (the wwwoffle-namazu script?) was
> still running after 40 minutes and - according to the statistics
> displayed by the web browser - had downloaded (the index?) more than 60
> MB when I threw in the towel.
The wwwoffle-namazu script should only be calling the namazu.cgi CGI
script. I don't know why it should be giving you 60 MB of junk. Do
you have any files of that size in or around the namazu directory?
> Is this the way that this script is supposed to work?
No, I haven't used namazu for a long time here, but calling up the
search CGI form worked without giving lots of extra data. I use the
unmodified version of the wwwoffle-namazu script.
> And, if so, just where is that created index and how does it get
> <revised|updated> when the WWWOFFLE cache is purged?
The document README.namazu explains how to use namazu with WWWOFFLE.
But you seem to have done everything correctly, the supplied scripts
should work without any modification.
--
Andrew.
----------------------------------------------------------------------
Andrew M. Bishop [EMAIL PROTECTED]
http://www.gedanken.demon.co.uk/
WWWOFFLE users page:
http://www.gedanken.demon.co.uk/wwwoffle/version-2.7/user.html
