Hello Tomasz,

It's great to see that you are trying to use some new techniques, instead of simply sticking with RPG III. It's always encouraging when you see that people are trying to improve their skills instead of just sticking with what they learned 15 years ago.

Unfortunately, sometimes when we try to mix old techniques with newer ones, we run into trouble. It might be worth taking things a step further and using a few more RPG IV techniques in your code, replacing some of the old RPG III stuff you still use.

There are several problems with your code. The one that's causing you the most grief is the fact that you are using the MOVEL op-code with a VARYING field. MOVEL was invented perhaps 30 years before VARYING, and it doesn't really understand VARYING. MOVEL will never change the length of a VARYING field.

On the last line of your program, you have this:

C Movel TXTO gsqTXTT

As I mentioned, MOVEL never changes the length of a VARYING field. gsqTXTT had a length of 0 when the program started, and MOVEL won't change that, it'll keep it at 0. So nothing actually gets moved into gsqTXTT in this statement. Using RPG III opcodes with VARYING fields will usually leave you frustrated.

The quickest fix to your program is to use EVAL instead. Remove the final MOVEL and use EVAL in it's place as follows:

C eval gsqTXTT = %trimr(TXTO)

That's the error that is causing you problems. However, there are several other problems with your code that you may not have noticed (and probably wouldn't find during testing):

1) You have the sqlFinalFlg flag coded incorrectly. It should be coded as "10i 0" instead of "5i 0".

2) You have specified "allow parallel" which I don't think should be used with code that's not threadsafe. I suggest changing it to "disallow parallel"

3) Your scratchpad parameter (sqlScrPad) is coded too short. The system always adds an Integer (10i 0) containing the length to the start of this field, so you should have coded it at 254 long. Though, I don't understand why you are using a scratchpad in this program. (You don't appear to use it for anything... so you'd never see the error...)

4) Your program starts by moving a varying (Varchar) field into a fixed-length field. That means you lose the ability to detect the original length of the field! That seems silly. If this were my code, I'd use VARYING throughout. (Unfortunately, that also means replacing the remainder of the MOVEL and XLATE opcodes with modern replacements)

5) You repeatedly do MOVEL *BLANKS and then on the next line of code replace the blanked out field with something else. This isn't really a bug, but it's a waste of CPU cycles. why fill it with blanks, then overwrite the blanks with something else? Why not just overwrite the data to begin with?

6) You set on LR in a NOMAIN subprocedure. This isn't really a bug, but it doesn't do anything, since there's no RPG logic cycle to check for the LR indicator. So setting it on does nothing. (It just looks foolish)


If you take all of these things into consideration, the resultant code looks like this: (I won't repeat the D-specs from what you had -- they just be self-explanatory...)

D TXT S like(gsqTXTF)

C eval TXT = %xlate(PO:LAT:gsqTXTF)
C eval TXT = %xlate(LOW:CAP:TXT)
C eval TXT = %xlate(SPC:BLK:TXT)
C eval gsqTXTT = TXT

Or better yet in free-format:

D TXT S like(gsqTXTF)

/free
TXT = %xlate(POL:LAT:gsqTXTF);
TXT = %xlate(LOW:CAP:TXT);
TXT = %xlate(SPC:BLK:TXT);
gsqTXTT = TXT;
/end-free

Good luck!


Tomasz Skorża wrote:
Hi

I created a function ZIPTXT which should "clean" text (change polish national chars to lating chars, and change all special chars to space)

code is:

H NoMain Option(*SrcStmt : *NoDebugIO) D zipTXT pr D gsqTXTF 100 Varying D gsqTXTT 100 Varying D sqlNull1 5i 0 D sqlOutNull 5i 0 D sqlState 5 D sqlFuncNm 517 Varying D sqlSpecName 128 Varying D sqlMsgTxt 70 Varying D sqlScrPad 250 D sqlFinalFlg 5i 0 P zipTXT b Export
D zipTXT pi D gsqTXTF 100 Varying D gsqTXTT 100 Varying D sqlNull1 5i 0 D sqlOutNull 5i 0 D sqlState 5 D sqlFuncNm 517 Varying D sqlSpecName 128 Varying D sqlMsgTxt 70 Varying D sqlScrPad 250 D sqlFinalFlg 5i 0
* Tables for conversion
D POL C 'ąęćśżźńółĄĘĆŚŻŹŃÓŁ' D LAT C 'aecszznolAECSZZNOL' D LOW C 'aąbcćdeęfghijklłmnńoóprsśtuvwxyzźż' D CAP C 'AĄBCĆDEĘFGHIJKLŁMNŃOÓPRSŚTUVWXYZŹŻ' D SPC C '~`!@#$%˘&*()_-+={}\|;:",<>?/' D BLK C ' '
C Movel *blanks TXTI 100 C Movel *blanks TXTO 100 C Movel gsqTXTF TXTI
C POL:LAT XLate TXTI TXTO C Movel *blanks TXTI C Movel TXTO TXTI C Movel *blanks TXTO
C LOW:CAP XLate TXTI TXTO C Movel *blanks TXTI C Movel TXTO TXTI C Movel *blanks TXTO
C SPC:BLK XLate TXTI TXTO
C Movel TXTO gsqTXTT
C SetOn LR C Return
P zipTXT e
Then I created it with command:

1) CRTRPGMOD MODULE(LIB/ZIPTXT) SRCFILE(LIB/SQLSRC) DBGVIEW(*LIST)
2) CRTSRVPGM SRVPGM(LIB/ZIPTXT) EXPORT(*ALL)
3) CREATE FUNCTION LIB/ZIPTEXT (varCHAR(100)) RETURNS varCHAR (100) LANGUAGE RPGLE DETERMINISTIC NO SQL RETURNS NULL ON NULL INPUT NO EXTERNAL ACTION ALLOW PARALLEL FINAL CALL SCRATCHPAD 250 PARAMETER STYLE DB2SQL EXTERNAL NAME 'LIB/ZIPTXT(ZIPTXT)'
When I calling this function it returns blank text with 100-chars length.

What is wrong?



As an Amazon Associate we earn from qualifying purchases.

This thread ...

Replies:

Follow On AppleNews
Return to Archive home page | Return to MIDRANGE.COM home page

This mailing list archive is Copyright 1997-2024 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available on our policy page. If you have questions about this, please contact [javascript protected email address].

Operating expenses for this site are earned using the Amazon Associate program and Google Adsense.