Re: Scan and Replace

On 02 Oct 2013 11:56, Kari Zeglin wrote:

On 02 Oct 2013 10:26, Alan Campin wrote:

What release level are you at. V7R1 has a Scan and Replace built
in function that would probably do the entire job.

When I do GO LICPGM, it tells me V7R1M0 under Installed Release.
Does that mean that I could use %SCANRPL?

Yes. Using the v7r1m0 RPG compiler on the IBM i 7.1 OS (v7r1m0) allows using that %SCANRPL (Scan and Replace Characters) Built-in function; i.e. the TGTRLS(*CURRENT) must be utilized on the compile.

AFaIK the following will replace all occurrences of a blank to the empty string:

SUBDTA = %ScanRpl(' ':'':SUBDTA) ; // remove blanks from SUBDTA

http://pic.dhe.ibm.com/infocenter/iseries/v7r1m0/topic/rzasd/sc092508871.htm#bbscanrp
"... %SCANRPL returns the string produced by replacing all occurrences of the scan string in the source string with the replacement string. ..."

On 02 Oct 2013 11:45, Justin Taylor wrote:

To remove spaces, wouldn't %XLATE be a much simpler option?

I tried that as well, but with me trying to learn all the different
ways, I think that I am confusing myself..
An example of %XLATE that I had tried was:
SUBDTA= %XLATE(' ' : '' : SUBDTA)
but I wasn't getting the correct result either.
Did I write that wrong too? <<SNIP>>

The typical concept of character translations is from one character to another; i.e. it does not include the concept of the removal of any characters. The effect from having specified the empty-string for the second argument\parameter caused the result of the %XLATE to be an effective no-op; i.e. the doc for the %XLATE (Translate) Built-in function suggests:

http://pic.dhe.ibm.com/infocenter/iseries/v7r1m0/topic/rzasd/sc092508896.htm
"... If the first parameter is longer than the second parameter, the additional characters in the first parameter are ignored. ..."