.Overview "Database/Dynamic_Hashing"
.
Functions in this category are related to manipulating the dynamic hashing
database that lq-text relies upon.
A dynamic hashing database provides a key to value mapping; the key can be
any binary data, and so can the value.
.P
Two dynamic hashing databases are used by lq-text: the first is used to
map a word into a \*[+c]WID\*[-c], that is, into a Word IDentifier number.
The second is used to map a filename into a \*[+c]FID\*[-c], that is, into
a File IDentifier.
.P
You can configure lq-text to use any of a number of different dynamic hashing
packages; ndbm is supplied with most Unix systems; Berkeley's `db' package
is included with lq-text, along with Ozan Yigit's `sdbm' package.
Whichever package you use, the result is essentially the same, except that
some packages are faster or more reliable than others.
For large databases (say, several hundred megabytes), you will probably need
to use the db package, since it has fewer size limits than most others.
.P
The individual dynamic hashing packages provide documentation on the various
routines, such as \*[+fn]dbm_fetch\*[-fn] and \*[+fn]dbm_store\*[-fn], that
you can use with the databases.
The lqword sample client uses routines that iterate over all entries in
a database, one by one.
.P
./Overview
