dbms_HASH...or something like it?...comparing data...

  • From: "Marquez, Chris" <CMarquez@xxxxxxxx>
  • To: <oracle-l@xxxxxxxxxxxxx>
  • Date: Sat, 22 Jan 2005 18:03:22 -0500

Need some help.

Developer has a wide table...20 columns.
He would like to add a 21st column to the table.
Then he would like to (be able to) store the first=20
20 columns data in concatenated form in the 21st column.
For example;
col1, col2, col3

No problem PL/SQL program can handle this, BUT
he would rather store the 20 concatenated columns in=20
a simpler, tighter form. He suggested "hashing" the=20
20 concatenated columns to (I guess) an Alphanumeric string?

And given that this is possible, what would be the=20
best/required data type for the hashed data?

Why you ask?
Because he want to be able to compare rows=20
(against new data) simply and quickly without selecting and=20
concatenating all 20 originally columns.

So he would store the hashed version of a row (data) on Insert.
Then on Update using a function hash the update data  and compare
it to the  already hashed value in the row.
Any help is appreciated?
Any other solution is appreciated, BUT his program is limited and=20
can only S,I,U,D and call PROCEDURES or FUNCTIONS.

Please reply directly to me as well as to the list.

Thanks in advance,

Chris Marquez
Oracle DBA
HEYMONitor(tm) - heymonitor.com
"Oracle Monitoring & Alerting Solution"


Other related posts: