[TOC]
On the command line, it is possible to use `curl` to get a dictionary
reference. We can first list all the dictionaries available at
[dict.org](http://dict.org) like so:
curl dict://dict.org/show:db
This lists all available dictionaries and their shortnames:
gcide "The Collaborative International Dictionary of English v.0.48"
wn "WordNet (r) 2.
moby-thes "Moby Thesaurus II by Grady Ward, 1.0"
elements "Elements database 20001107"
vera "Virtual Entity of Relevant Acronyms (Version 1.9, June 2002)"
jargon "Jargon File (4.3.1, 29 Jun 2001)"
foldoc "The Free On-line Dictionary of Computing (27 SEP 03)"
easton "Easton's 1897 Bible Dictionary"
hitchcock "Hitchcock's Bible Names Dictionary (late 1800's)"
bouvier "Bouvier's Law Dictionary, Revised 6th Ed (1856)"
devils "THE DEVIL'S DICTIONARY ((C)1911 Released April 15 1993)"
world02 "CIA World Factbook 2002"
.
.
.
eng-fra "English-French Freedict Dictionary"
To look up a word or phrase from a certain dictionary, issue:
curl dict://dict.org/d:word_goes_here:dictionary_short_name_here
For instance, I'll attempt to look up India in the CIA World FactBook
(2002 edition):
curl dict://dict.org/d:India:world02 > India.Reference.txt
And obtain this (partial dump):
20 miranda.org dictd 1.9.15/rf on Linux 2.6.26-2-686 <62887065.988.1256582810@miranda.org>
250 ok
150 1 definitions retrieved
151 "India" world02 "CIA World Factbook 2002"
India
Introduction India
------------------
Background: The Indus Valley civilization, one
of the oldest in the world, goes
back at least 5,000 years. Aryan
tribes from the northwest invaded
about 1500 B.C.; their merger with
the earlier inhabitants created the
classical Indian culture. Arab
incursions starting in the 8th
century and Turkish in 12th were
followed by European traders
beginning in the late 15th century.
By the 19th century, Britain had
assumed political control of
virtually all Indian lands.
Offline Dictionaries
--------------------
The example above uses online resources. You might, however, need to use
these offline. Like everything else, there's a tool for precisely this
purpose on Linux called [**dict**](http://www.dict.org/bin/Dict) and is
quite flexible in terms of its features. Think client-server
architecture, multiple dictionary references, configurable hosts on the
client side, etc.
The utility is split into two tools: `dictd` (the server daemon) and
`dict` (the client application.)
The `dict` client, by itself, can search the resources at dict.org if
you don't want to install the server. What follows is a guide to
installing the `dictd` daemon and configuring the client app to look at
'itself' (i.e. localhost)
#### Step 1: Obtain and install dict
wget ftp://ftp.dict.org/dict/dictd-1.9.15.tar.gz
Where 1.9.15 is the latest version at the time of this writing. Untar it
and follow the [typical installation
instructions](ftp://ftp.dict.org/dict/INSTALL) to compile this from
source.
**Note**: You might get compiler warnings when running `make` against
lower version numbers [since this is a
bug](http://bugs.gentoo.org/81211).
A few things:
* If you did not use `--prefix` when running `configure`, the path to
`dictd` (server) should be `/usr/local/sbin`
* Configuration files for client and server will be in
`/usr/local/etc`
#### Step 2: Obtain and set up dictionaries
The dictionary or reference files the `dictd` daemon requires have two
parts: the database and the index files. In this step, we will get the
Webster's 1913 dictionary (which is in public domain).
wget ftp://ftp.dict.org/dict/pre/dict-web1913-1.4-pre.tar.gz
Once you untar the archive, place the two files `web1913.dict.dz` and
`web1913.index` in a convenient location. I placed mine at
`/usr/share/dictionaries`. Make sure these are readable!
chmod a+r web1913.dict.dz web1913.index
You can obtain other dictionaries by perusing [the FTP site you got the
Webster's dictionary from](ftp://ftp.dict.org/dict/pre/)
**Note**: `/usr/share/dict` is *not* an appropriate location!
`linux.words` is the file used by the `passwd` command, for instance, to
determine weak passwords.
#### Step 3: Configure dict
You now need to create two configuration files for the dict server and
client: `dictd.conf` and `dict.conf` respectively. These need to go in
`/usr/local/etc`
Configure dictd server
----------------------
`dictd.conf` will contain references to the database(s) from Step 2.
database web1913 { data "/usr/share/dictionary/web1913.dict.dz"
index "/usr/share/dictionary/web1913.index" }
If you installed another reference database from [the FTP
site](ftp://ftp.dict.org/dict/pre/) in Step 2, append it to this file
using similar syntax. For instance, if I downloaded a list of chemical
elements, I would add this:
database elements { data "/usr/share/dictionary/elements.dict.dz"
index "/usr/share/dictionary/elements.index" }
After this stage, start the dictd server
` /usr/local/sbin/dictd start`
Test your installation with something
` dictd --test perspicacious`
Configure dict client
---------------------
`dict.conf` can contain a single server or a list of servers which have
the `dictd` daemon running. In this example, we'll use our own machine
` server localhost`
And that's it! Test this with
` dict surreptitious`
Other notes
-----------
You might need an init script if `dict word` does not work.
```bash
DICTD=/usr/local/bin/dictd
# DICTD_OPTIONS="-put -command_line -options -for -dictd -here"
DICTD_OPTIONS=""
DICTD_PID_FILE=/etc/dictd.pid
case "$1" in
'start')
if [ -x $DICTD ]; then
echo "dictd starting."
$DICTD $DICTD_OPTIONS
else
echo "dictd.init: cannot find $DICTD or it's not executable"
fi
;;
'stop')
if [ ! -f $DICTD_PID_FILE ]; then
exit 0
fi
dictdpid=`cat $DICTD_PID_FILE`
if [ "$dictdpid" -gt 0 ]; then
echo "Stopping the dictd server."
kill -15 $dictdpid 2>&1 > /dev/null
fi
rm -f $DICTD_PID_FILE
;;
*)
echo "Usage: dictd.init { start | stop }"
;;
esac
exit 0
```
### To Do
* Add information on doing this securely (i.e. dictd should use only
one port in iptables)