I'm planning to write WWW dictionary server/cilents.
Maurice Cinquini (mauricec@tplrd.tpl.oz.au)
Wed, 5 Oct 1994 15:12:52 +1000 (EST)
NOTE: libwww is only minimally involved in the following, but
In the welcome message to this mailing list Roy T. Fielding
<fielding@avron.ICS.UCI.EDU> writes:
> What is appropriate discussion for this mailing list
> ----------------------------------------------------
..
> If you are working on a project of general usefulness, let everyone else
> know about it.
I'm working on a set of programs to serve up dictionary entries across
a distributed platform. The entries I have at the moment are an in house
glossary of terms (mostly acronymns) and the following two files.
Hacker's Jargon file, v3.00, 523845 bytes
<ftp://prep.ai.mit.edu/pub/gnu/plaza.aarnet.oz.au/jarg300.info.gz>
Internet Users' Glossary, RFC 1392, 104624 bytes
Q. Any other public domain dictionaries out there?
My current implementation is a single grep-like perl script and
a single ascii file of entries on a unix machine.
Now I want to make this service available to all of the networked PC
users in our company as well, and still keep central control over the
dictionaries (the in house list needs regular updates). I thought I'd
use a Perl && HTTP && HTML && (Mosaic || libwww) solution to achieve this.
Details:
1) Write perl scripts to transform existing dictionary sources into a
single common dictionary file format.
My current dictionary file format proposal is:
* For a dictionary per subject (eg. computer jargon, common english)
* Each dictionary is an ascii text file of dictionary entries.
* Each dictionary entry is begins with "\n:" and consists of a term
and definition.
* Terms are one line.
* Definitions are many lines, can include HTML markup.
Eg. "HEADER\n:TERM1\nDEFN\nMORE DEFN\n:TERM2\n..."
As a shortcut, cross references within a dictionary can be enclosed
within curly braces (ie. {}).
Q. Any suggestions for the common format of the raw dictionary entries?
2) Write a server CGI perl script callable from a HTTPD server that outputs:
a) a HTML+ query form to allow selection of the search phrase and
search options (eg. dictionaries, upper/lower case sensitivity) or;
b) HTML formatted output of the dictionary entries found when search
phrases have been entered.
I plan to use the cgi-lib.pl routines for this.
Q. Has the name ``dictcgi'' been used yet?
3) Write a cilent perl script that takes command line requests (for
those that won't/can't use a WWW cilent like Mosaic) and retrieves
the entries from the server CGI perl script via the libwww functions
(possibly striping out any HTML commands?).
Q. Has the name ``dictwww'' been used yet?
Q. Has anything like this been done yet?
Surely there's something out there. Note that provision of an
immediately up to date company-wide onl-line telephone database
across a variety of platforms is similar to this problem.
My company only runs WWW software internally, so pls don't direct me to
URLs I can't access. (FTP retrieval is possibly though).
--Maurice
Maurice Cinquini __T_e_l_e_c_t_r_o_n_i_c_s__|/\ __
Senior Design Engineer Pacing Systems \/
E-mail: mauricec@tplrd.tpl.oz.au Voice:+61 2 413 6702 Fax:+61 2 413 6060