ROADS and Z39.50
Using Zebra to provide Z39.50 access to ROADS servers
The Zebra system is a fielded free-text indexing and retrieval
engine with a Z39.50 frontend, developed by
Index Data. By using a script (roads2gils.pl, see below)
to convert ROADS records into GILS records suitable for loading
into Zebra, a ROADS database can be made available to Z39.50 clients.
Typically, a cron job will be used to regularly convert the ROADS
records to GILS and index them using Zebra.
Obtaining the Software
Zebra is available from
The roads2gils.pl script is available from
Setting up Zebra
- Obtain Zebra and install it as described in
the Zebra documentation.
Create a new database to hold the converted copies of your ROADS records.
Let's assume that your converted records will be stored in
- Obtain the roads2gils.pl script and install somewhere (e.g.
- Run the roads2gils.pl script by hand to generate your GILS
records for use with Zebra. The script takes four arguments:
For example, given the above Zebra directory and a ROADS 'source' directory of
/opt/roads/source, you might type:
must be supplied. It tells the script where your ROADS records are stored.
may be supplied. It tells the script where to put the generated GILS
records. If this argument is not supplied then the current directory is assumed.
forces creation of a new GILS record for each ROADS record.
Unless this argument is given, roads2gils.pl will only convert those records
that are new or that have changed since the last time the script was run.
will generate some debugging output.
/opt/bin/roads2gils.pl -s /opt/roads/source -g /opt/zebra/data/ROADS/records
Index the new records using Zebra as desribed in the Zebra documentation.
Once you have confirmed that this procedure works, you may want to set up
a cron job to automatically convert your ROADS records and re-index the
Zebra database on a regular basis.
To be supplied
GRS-1, USMARC or SUTRS
Powell of the UKOLN Metadata
Last updated: 12-June-1998