Extending MOMspider's Functionality
MOMspider has been designed to be easily extensible for use in building
related web indexes and to allow it to drive other programs that care
about the traversal structure of a web. This extensibility can be
found (and used) in four areas:
- All decisions made about what metainformation to store for traversed
and tested nodes (URLs) are made within two simple subroutines
-- store() and recall() -- in
momhistory.pl. It is a trivial task
to add new metainformation names to those being saved.
- All traversal events are handled through the routines in
momevent.pl. This provides a simple
and convenient location to embed hooks that drive new programs,
just as the existing hooks drive the indexing
and email processes.
- The MOMspider index files
can be post-processed to extract the
relevant traversal information and metainfo.
- The MOMspider diagnostic
output also includes all the information
needed to discover the structure of a web.
These points (and more) will be described better in future releases
and (hopefully) by the many people who have expressed interest in
building other tools on top of MOMspider.
Roy Fielding <fielding@ics.uci.edu>
Department of Information and Computer Science,
University of California, Irvine, CA 92717-3425
Last modified: Wed Aug 10 02:55:06 1994