I'd like to adapt swish to index text inside a rather large (200
Mbytes, 300000 records) GDBM database, and then write
code to find the records pointed to by a search's results.
I've provided a search interface to this database once already,
by writing files containing the search text for each record,
indexing those files using the fs method, throwing away the
files, and then using the search results to find the original
records (there's a simple relationship between the names of
the files used to build the index and the database records).
While this process worked, it took a loooong time to create
the files and a long time to build the index, so I'd like to save
a few steps and build a custom indexing method that works
directly on the database.
Since the documentation on doing this is very skimpy, I'd like
to know if anyone else has done this before I start banging
my head against this particular wall :)
Any pointers, code snippets, examples, etc, that people can
provide would be greatly appreciated. With or without help, I'll
publish my finished indexing method code when I'm done, so
everyone else can use it.
Steve van der Burg
Technical Analyst, Information Services
London Health Sciences Centre
London, Ontario, Canada
Tel: +1 519 685-8300 x 35559
Received on Tue May 18 07:29:50 1999