I assume this has been done many times before, so I am wondering if you have code around to load a MySQL database into Texis. I would prefer to use the API (I think -- I already have a site written in Perl, so I'm not sure if Vortex is suitable for me).
By the way, if I am misunderstanding something (which is pretty likely) please let me know. But, it seems that you have to use vhttpd to use Vortex, and then I am not sure how one would integrate the Vortex with existing Perl (or ASP or PHP...), so I am shying away from Vortex at the moment since it seems to require major changes to a server/site. If this is not true, a quick overview (in small words) would be very helpful. I am sure that this is in the docs somewhere, but they are over 500 pages long and haven't made a whole lot of sense to me so far.
Don't know the best way to get mysql into texis at the moment (you could generate csv from mysql and pipe that into timport) but as for vortex, you don't need vhttpd. There's a CGI program called "texis" that comes with Texis and runs vortex scripts. During install you were asked your CGI directory and texis would have been installed. Given a standard apache config you would access vortex with something like http://yoursite.com/cgi-bin/texis/yourscript
You might find the tutorial helpful. Check out the "Vortex Introduction" in particular. http://www.thunderstone.com/texis/site/tutorial/
One note, the tutorial hasn't been updated for version 5. The scripts are in /usr/local/morph3/texis/scripts by default rather than document root.
Thanks for the info. I would generate CSV if my data weren't 100GB and full of special characters that I don't trust to work correctly in CSV. Even if it worked, by the time it could export I could probably have another solution figured out (knock on wood). I like Perl, and I just figured out where the DBD::Texis module is, so I hope to use that if I can find documentation for uts use.
Thought I would post a follow-up to this in case anyone else has similar questions. The resolution seems to be:
1) DBD::TEXIS doesn't compile with the current (5.8.3) version of Perl, so that option is out.
2) I ended up using TIMPORT, which worked flawlessly and could process about 100 records/sec from MySQL to TEXIS on an Athlon 3200+ using a separate drive for each database.
By the way, I was controlling TIMPORT with a Perl script, which also did the SQL queries to the MySQL database. TIMPORT can't directly access a MySQL database -- I just wanted to make sure that was clear because perhaps I made it sound like that.
