	locus is a fulltext database, i. e. database, to which you put texts too
long to read/remember and then ask which words are in them and where. The core
of locus are two programs: cruncher and grazer. See file INSTALL on how to make
them (basically make; make install). Both cruncher and grazer get their
configuration from text file locus.opt (in their working directory), whose
format is described in file options.

	Basic work with locus goes as follows (for various complications see
file options):
	1. Edit locus.opt - you must specify at least directories with your
documents and target directory for index files (it must exist).
	2. Run cruncher (without parameters) and wait. Indexing is slow; it can
be somewhat optimized by a stop list (see file options), but even then, it
won't be fast.
	3. If cruncher ended OK (did not written anything like "assertion
failed\n aborting...", the database is ready for searching. If it did not end
OK, the database is inconsistent and unusable. Indexed files cannot be removed
nor changed; you can only add new ones to the database.
	4. grazer searches in the database created by cruncher for words
specified as its command-line arguments (fancier input is also possible - see
file search). Its output format is customizable text (if you want HTML, you
can have it - see file options).

	If you have problems using locus, bug reports, suggestions for
improvement etc., mail them to vbar@comp.cz
