We have started using Ht://dig on the chemisty site

As we have significant number of JCAMP dx spectra files, PDB andVRML We have written parsers which allow Htdig and JObjects to index them. The parser is able to cope with Gzipped pdb files The following lines are indexed in the pdb parser REMARK , TITLE, AUTHOR, JRNL, KEYWDS, EXPDTA, COMPND but the particular search engine may not index all these fields. The JCAMP parser indexed the following fields TITLE , OWNER, DATA TYPE

the VRML parser extracts the following fields

VRML 1.0: AsciiText WWWAnchor WWWInline

VRML 2: Anchor text viewpoint Worldinfo

NB VRML has not been tested at all! (yet)

This is the first release and no major test to date has been done.

HT:DIG

The contents of the pdb and Jcamp files are indexed with a weighting of 1

Get the Classes

Get the Source

These are written in Java and so you need a java installed

unzip these files into the htdig bin directory - they should go in to a directory called chemical

Add the following line to the Ht://dig config file:

external_parsers: chemical/x-pdb "/usr/java/bin/java chemical.Htdigfront" \

chemical/x-jcamp-dx "/usr/java/bin/java chemical.Htdigfront" \

x-world/x-vrml "/usr/java/bin/java chemical.Htdigfront" \

model/x-vrml "/usr/java/bin/java chemical.Htdigfront"

Making sure that they are pointing towards the java binary!

Ht://dig should now be able to cope with the pdb files and jcamp dx files ( if the server has these files under a different mime type these will have to be changed accordingly

JOBJECTS

These files below are for JObjects

The word and comments are indexed as body text. When these files are unzip they will go in to a directory called chemical, this is correctand should not be changed

Note VRML is not avaible for J Objects

Get the Classes

Get the Source

The mime types that have to be set up are:

chemical/x-pdb chemical/x-jcamp-dx

Last updated 1/06/99

For further information please contact Dr H Rzepa