After spending quite some time unsuccessfully attempting to get
the listserver to produce a data base search of the parkinsn data
base [using the "job stream" format found in a recent post], I was
about to give up, figuring that some sort of "system
incompatibility" was thwarting my efforts.
But rather than give in, I took the bull by the horns, and through
tenacity [my 13 year old calls it "brute force"] and information
gleaned from scanning hard copies of various listserver manuals
strewn haphazardly at my side, I managed to stumble on a "job
stream" that actually worked for me. Here it is, for any of you
having a similar problem communicating with the machine:
======================SNIP==================
// JOB Echo=Yes
Database Search DD=Rules
//Rules DD *
SEARCH Q-10 in parkinsn
INDEX
SEARCH NADH in parkinsn since November
INDEX
/*
======================SNIP==================
This is, in reality, a modification of the excellent example given by
John Cottingham on 13 Nov 1995 [Subject: Finding information in
Parkinson Archives by E-mail 11/1l/95].
The first three lines, as well as the last line, should be typed
exactly as shown. [The "Echo=Yes" in the first line makes it easier
for me to interpret the output by printing the commands in the
output file - John uses an "Echo=No" to suppress such
information - choose for yourself!]
The line pairs beginning SEARCH and INDEX:
(1) actually define what you're looking for [SEARCH], and
(2) tell the listserver to show you a tabular INDEX of the "hits"
[John C. uses the listserver synonym "SELECT", which does the
same thing as "SEARCH" - either will work - and he puts SELECT
and INDEX on one line].
Notice in my example that two different searches were conducted -
one for items containing the string "Q-10" [no time limit], and one
for items containing reference to "NADH",but only if posted since
November 1 of this year.
Two files are generated by this job. One contains a resource
utilization report [computer time to do the job], and can be ignored
[it's called 'Output from your job xxxxxxxx", where "xxxxxxxx" is your
user id].
The second file, referred to in the "Subject field" of your mailbox
as:
File: "DATABASE OUTPUT"
actually has the info that is desired.
Using the above job stream, the following was produced:
> SEARCH Q-10 in parkinsn
--> Database PARKINSN, 6 hits.
> INDEX
Item # Date Time Recs Subject
------ ---- ---- ---- -------
002023 95/03/01 12:13 1568 Enclosures - Alzheimer's research
002752 95/04/24 09:05 309 Newsletter
002785 95/04/27 22:41 525 PARKINSN Digest - 23 Apr 1995 to 24
Apr 1995
006532 95/11/06 11:30 109 NADH\SOMEWHAT TECHNICAL...
006553 95/11/07 06:54 103 CoEnzyme Q-10/Thanks Delana!
006555 95/11/07 07:32 45 NADH [Q-10] P.S.
> SEARCH NADH in parkinsn since November
--> Database PARKINSN, 58 hits.
> INDEX
Item # Date Time Recs Subject
------ ---- ---- ---- -------
006446 95/11/02 11:06 38 Hello and Introduction
006451 95/11/02 13:29 49 Re: Hello out there... need advice for a
new Parki+
...
... most edited to keep this e-mail small.
...
007070 95/11/25 10:02 66 Re: NADH=Q10?? + MORE
007089 95/11/26 20:24 12 Re: NADH in N.Z.different?
007144 95/11/28 11:47 34 UPF Newsletter
For info on the meaning of the various items in the output file, as
well as a tutorial on how to get at the actual files listed in the
output, see John's excellent post mentioned above.
An important point of order - send the data base search job to the
listserver at:
[log in to unmask]
Otherwise all 1000+ of us will get to see your query!
Jim
"So what if I can't do 2000 things anymore -
just think of all the extra time I have
to do the 1799 things I still can do!"
jea
[ INTERNET e-mail: ]
[log in to unmask]
-or-
[log in to unmask] (for those of you who hate typing like me)
