Greetings,

I have a requirement to create a file consisting of attribute/value pairs.

My input is an Oracle table.

Currently I am using the IB sample database 'EMPDATA' as my model.

Here is what I have at the moment..

Focexec:

APP PREPENDPATH IBISAMP
FILEDEF FSEQ DISK common/fseq.txt
JOIN CLEAR *
-RUN
-*
JOIN BLANK WITH PIN IN EMPDATA TO BLANK IN FSEQ AS J1
-*
DEFINE FILE EMPDATA
 BLANK/A1 WITH PIN = ' ';
 ATTRIBUTE/A25 = DECODE COUNTER(  1 'LASTNAME'
                                  2 'FIRSTNAME');
 VALUE/A25     = IF COUNTER EQ 1 THEN LASTNAME ELSE FIRSTNAME;
END
-*
TABLE FILE EMPDATA
 PRINT ATTRIBUTE
       VALUE
 IF COUNTER EQ 1 OR 2
 IF LASTNAME EQ VALINO
 ON TABLE SAVE AS SVATTRS
END
-RUN
-SET &WORK = &LINES;
-*
-REPEAT ENDLOOP1 &WORK TIMES
-READ SVATTRS &ATTR.A25. &VAL.A25.
-TYPE Attribute: &ATTR Value: &VAL
-ENDLOOP1
-EXIT

Results:

0 NUMBER OF RECORDS IN TABLE=        2  LINES=      2
 ALPHANUMERIC RECORD NAMED  SVATTRS
 0 FIELDNAME  ALIAS         FORMAT LENGTH
   ATTRIBUTE                A25    25
   VALUE                    A25    25
 TOTAL                             50
 Attribute: LASTNAME  Value: VALINO
 Attribute: FIRSTNAME Value: DANIEL  

The attributes are a subset of the fields in the input file. And the attribute values, are the contents of the field in the file.

I am using the McGuyver technique to do a in memory read the incoming record n number of times.

n is equal to 2 in this case, because I want two attribute/value pairs.

My request: Is there a better way to do this, then what I am doing now?

In my real case I need to create 40 attribute/value pairs (from a very wide Oracle table), and I am concerned about the growing complexity of the IF THEN ELSE structure.

Thanks in advance,
Dave

This message has been edited. Last edited by: David Briars,

How about this.

TABLE FILE SYSCOLUM
PRINT NAME USAGE
WHERE TBNAME EQ 'EMPDATA'
ON TABLE SAVE AS COLNAMES 
END

-RUN

-SET &COLS = &LINES ;

FILEDEF SVATTRS DISK svattrs.ftm

TABLE FILE EMPDATA
 PRINT 
-REPEAT COLLOOP FOR &CNTR FROM 1 TO &COLS ;

-READ COLNAMES NOCLOSE &ColName.A66. &ColFmt.A8.
-SET &ColName = TRUNCATE(&ColName) ;
-SET &ColFmt  = TRUNCATE(&ColFmt) ;

 COMPUTE COL&CNTR/I2 = PUTDDREC('SVATTRS',7,LJUST(&ColName.LENGTH,'&ColName','A25') | FPRINT(&ColName,'&ColFmt','A25') ,50,COL&CNTR) ;

-COLLOOP
 WHERE RECORDLIMIT EQ 1
 ON TABLE HOLD AS WRITEOUT 
END

-RUN

-REPEAT ENDLOOP1 &COLS TIMES
-READ SVATTRS &ATTR.A25. &VAL.A25.
-TYPE Attribute: &ATTR Value: &VAL
-ENDLOOP1

Thank you Waz!

This is the 'quantum leap' improvement to my version, that I was hoping for.

The use SYSCOLUM reduces hardcoding.

Function PUTDDREC nicely stacks the attribute-value pairs.

And FPRINT greatly supports my need to reformat the non-text fields.

Once again, many thanks,
Dave