In article <3oe6nv$s5o@cssun.mathcs.emory.edu>,
nnob@loc.gov (Nick Nobbe) wrote:
>On Thu, 4 May 1995, Michael Jatzeck wrote:
>
>> Hello,
>>
>> I am confronted with a usual problem in database applications: I have to
>> prevent the users from inserting items more than once. The primary key of
this
>> item is a text string, therefore my first choice was the Shift-AND
algorithm
>> by S.Wu and U. Manbar to find rows with a similar key before entering a new
>> one.
>>
>----- stuff deleted -----
>
> Michael,
> Hey, wait a minute! Did you say you want to prevent your users from
> entering a duplicate key. How can that be if the key is unique?
> Why bother if you can trap the error condition and return a message, or
> if the users are using Isql, they will most likely receive an Informix
> error message.
> Begging you forgiveness for my fuzzy thinking, but it is late Friday
> afternoon over here.
Well the point is, that this text string contains the name of a real person.
Therefore it might be mispelled by the user. What I want to prevent is that
multiple records are inserted for the same person.
Of course you cannot prevent a user from inserting the above kind of duplicate
record if he wants to. But you can provide the user with a list of similar
entries so that he at least knows about them and can choose by himself. This
is I want to use the Shift AND alg. for.
Yours,
Michael
__________
_/_/_/_/_/ _/_/_/_/ _/_/_/_/ Michael Jatzeck michel@taz.de
_/ _/ _/ _/ die tageszeitung
_/ _/_/_/_/ _/ Germany
_/ _/ _/ _/ D-10969 Berlin Phone: +49-30-25902-131
_/ _/ _/ _/_/_/_/ Kochstr. 18 FAX : +49-30-2510694