Next: Study to Install an Information System at the University of the Istanbul Astronomy and Space Sciences Department
Up: Archiving and Information Services
Previous: Using a Java Web-based Graphical User Interface to access the SOHO Data Archive
Table of Contents -
Subject Index -
Author Index -
Search -
PS reprint -
PDF reprint
Bell, D. J. 1999, in ASP Conf. Ser., Vol. 172, Astronomical Data Analysis Software and Systems VIII, eds. D. M. Mehringer, R. L. Plante, & D. A. Roberts (San Francisco: ASP), 257
Webmail: An Automated Web Publishing System
David Bell
NOAO1, P.O. Box 26732, Tucson, AZ 85726
Abstract:
A system for publishing frequently updated information to the World Wide
Web will be described. Many documents now hosted by the NOAO Web server
require timely posting and frequent updates, but need only minor changes
in markup or are in a standard format requiring only conversion to HTML.
These include information from outside the organization, such as electronic
bulletins, and a number of internal reports, both human and machine
generated. Webmail uses procmail and Perl scripts to process incoming email
messages in a variety of ways. This processing may include wrapping or
conversion to HTML, posting to the Web or internal newsgroups, updating
search indices or links on related pages, and sending email notification
of the new pages to interested parties. The Webmail system has been in
use at NOAO since early 1997 and has steadily grown to include fourteen
recipes that together handle about fifty messages per week.
Publishing a document on the World Wide Web often involves several steps, which
may be performed by multiple parties:
- Authoring - The document's content is created.
- Editing - Textual markup is added, and document may be split up.
- Posting - The document is copied to a Web server.
- Referencing - Links to the document are added or updated on other
pages.
- Advertising - The document's presence is announced.
In many cases most or all of these steps can be completed in an automated
manner. By reducing the need for direct human involvement, such documents
can be posted and updated quickly and accurately, even by individuals without
any experience with HTML.
The Webmail software is installed in a dedicated account on the Web server
which has been given write access to any directories or files it must update.
On the author's end, posting consists merely of sending
an electronic mail message to this account.
In some cases, a pre-agreed subject line is used to identify the document.
For messages originating from interactive programs, this step is performed
with a simple ``Submit'' button.
Procmail first filters the message by running a MIME-decoder, if needed.
This has been particularly useful for handling messages sent from some
PC-based mailers which unnecessarily encode plain text attachments,
often without the sender's knowledge.
Procmail then checks against a series of recipes to identify the message
type. These include a list of addresses permitted to post or update the
document, and
usually information about the expected subject line or contents of the
message. Procmail logs and archives all submissions and passes the message
contents and type to a Perl script for further processing. Unidentified
messages are forwarded for human review.
The Perl script includes recipes for each message type. Standard methods
include Web posting, posting to newsgroups, archiving previous versions,
and sending email announcements to distribution lists.
Specialized recipes are used for the more sophisticated documents.
There is not always a one-to-one
correlation between email messages and Web documents - a single message might
be split into many pages or used to update just small sections of existing
pages. Links may be added or updated on several related pages. For
documents that are part of a searchable Web database, index files
are also updated.
Webmail was launched in early 1997 for NOAO's internal posting of IAU
Circulars, but has since grown to include fourteen recipes that together
process about fifty messages per week. Recipes range from simple HTML-wrappers
to those performing sophisticated parsing and reformatting. Some
examples are described below.
- IAU Circulars - Message are indexed for searching, posted to an
internal newsgroup, and saved as plain text. A CGI script on the NOAO Intranet
is used to search circulars and add HTML wrappers and links to related
circulars. A useful feature of this approach is that the links may point to
circulars that arrived at a later time with updated information.
- Telescope Schedules - CTIO and KPNO schedules are sent from
a MS Access database in a column-formatted table encompassing six months
of dates for all telescopes. Individual months are split out of the message,
formatted into HTML tables and cross-linked. For each scheduled project, links
are provided to a page with detailed information including a full
list of investigators and the project's title and scientific abstract.
Minor schedule changes may occur several times each semester, so any changed
months are announced by email to a distribution list. Example schedules
may be found at http://www.noao.edu/kpno/forms/tel_sched/
or
http://www.noao.edu/ctio/forms/tel_sched/.
- Conference Room Schedules - These are maintained on a PC running
Schedule+ and are output as HTML and mailed. Webmail fixes some
of the markup and splits and recombines weekly schedules into more useful
formats for posting on the NOAO Intranet. A single message updates over
twenty Web pages.
- ACM Colloquium page - This is one of the simplest recipes Webmail
processes, but may be of interest since the author is from outside
NOAO. A pre-formatted HTML document is simply posted to the NOAO
server at http://www.noao.edu/noao/acm.html.
We have found the Webmail approach to be quite useful and powerful, yet
perhaps so simple
as to be easily overlooked. For pre-formatted documents needing little
or no markup, new recipes can be set up with just a few lines of procmail
and Perl code. The resulting process is more efficient than passing each
document through a human Webmaster, and more secure than giving all authors
direct access to the Web server. For documents requiring extensive parsing
and markup, the benefits of an automated system are even greater.
Figure 1:
A telescope schedule page produced by Webmail.
|
Footnotes
- ...NOAO1
- National Optical Astronomy Observatories, operated by the
Association of Universities for Research in Astronomy, Inc. (AURA) under
cooperative agreement with the National Science Foundation
© Copyright 1999 Astronomical Society of the Pacific, 390 Ashton Avenue, San Francisco, California 94112, USA
Next: Study to Install an Information System at the University of the Istanbul Astronomy and Space Sciences Department
Up: Archiving and Information Services
Previous: Using a Java Web-based Graphical User Interface to access the SOHO Data Archive
Table of Contents -
Subject Index -
Author Index -
Search -
PS reprint -
PDF reprint
adass@ncsa.uiuc.edu