Here's a post I made recently to usenet's comp.lang.perl.modules. It's
probably more likely to interest the subscribers of the htdig developer's
list, however, so I'm copying it for your review. I'm especially interested
in opinions on what features should be included, what the GUI should look
like, what level of browser you want supported (e.g.: do you want
lynx-support or DHTML).
Also, I need your comments on the terminology I'm introducing, such as
calling individual htdig configurations "Sites", even though an htdig "site"
might include URLs from numerous web sites. Would "Corpus" be preferable,
or "Moniker". I've used both previously, but they don't seem to apply here,
since corpus implies the database and moniker implies the name itself, not
the configuration.
James Tillman
[EMAIL PROTECTED]
Advanced Systems Design of Tallahassee
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
Sent: Monday, December 27, 1999 8:26 PM
To: [EMAIL PROTECTED]
Subject: RFC: Perl ht://Dig admin interface
I recently won a bid to develop an administrative interface to the open
source site indexer called ht://Dig (www.htdig.org).
My plans are to develop a group of perl modules under the HtDig perl
namespace which provide programmatic access to ht://Dig configuration and
maintenance, and then to create some mod_perl-compatible CGIs which will
provide a GUI for the admin modules.
Although it isn't required, I want what I create to be placeable in CPAN.
So I need comments not only on the interface itself, but also on the
namespace choices and appropriateness of the modules themselves.
I've already got a pretty good idea of what I'll be having these modules do,
but wanted to get some opinions from anyone else who might be interested.
Because htdig allows the configuration to be defined when indexing starts,
the modules will maintain multiple configuration files, each of which will
be usable when an indexing run begins. I'm currently distinguishing between
multiple "conf" files by calling them "Sites", so each Site object will
indicate a separate "conf" file. I'm not in love with this by any means, so
any other suggestions are welcome.
Here's my current interface plan, subject to enormous change:
Module HtDig::Config
* >sites - provides named access to, or enumeration of, all the site
configurations that are currently known for the htdig installation
Module HtDig::Site - will provide add, delete, update, and auto-detect
of htdig configuration files (.conf files), as well as control
of the indexing process
* >properties - all properties that can be set in the .conf file will be
accessible, see the docs for a list. Additional settings will include log
file locations and email addresses of people to be notified when the
database is updated.
* >logs - access to logs of previous indexing runs (not quite sure how
access will be provided, though, probably just locations)
* >merge(site_object or conf filename) - merge another htdig Site's database
with the current one
* >index() - begin indexing a site (I might provide a delayed index, with a
time parameter)
There will probably be hooks into cron (or "at" for NT) for automating the
indexing runs. Other methods and properties will probably present
themselves as time goes by, but if you've got anything you want to toss in,
please feel free to post it here for comment.
Thanks!
Jamie
------------------------------------
To unsubscribe from the htdig3-dev mailing list, send a message to
[EMAIL PROTECTED]
You will receive a message to confirm this.
