There is no economy of whether it's done at the source or before exporting,
the work is still the same.
Any search and replace for characters vs characters is still going to be the
number of total characters to examine times the number of characters to
compare. I think you could speed up the process a bit by creating a win32
dll which would work inline as opposed to calling a SP against the data.
Such a DLL would look like this against the dataset.
Update SomeTable Set SomeCol = (DLCall('MyCustomDll.dll',
'MyCustomSRPLfunction', SomeCol))
I have no idea against the size of the data column you are working, but the
speed of the update would depend entirely on size and the number of
Unprintable characters you wish to replace with spaces.
----- Original Message -----
From: "Steve Hartmann" <[email protected]>
To: "RBASE-L Mailing List" <[email protected]>
Sent: Friday, June 14, 2013 3:09 PM
Subject: [RBASE-L] - Eliminating Unprintable Character
Does anyone have a command file that will analyze variable and automatically
replace unprintable characters with spaces. Much of our data is imported
from client sources and now we have to export it and the receiver of the
data objects to these characters. I know I have written one before myself.
For now we can use such a function in the programs that create the exports,
but ideally I would like to put it at the source. Unfortunately that means a
lot of programs to change. I have not worked with R:Base triggers very much,
just Oracle. Would it be possible to add a function to "clean" the data of
unprintable characters as part of and ON UPDATE OR ON INSERT trigger.
Steve
[email protected]
________________________________
