Nested loops and if statements

Question

I have four schemas with a problem category table "gppcat" that
has problem categories and problem descriptions. I need to
output these categories and descriptions to a csv file. while
doing this I need to make sure I don't include duplicate
records. I use schema 1 as a system of record. I want to select
pcode "category" from schema 1 and loop through to compare each
record in the other schemas and write each unique record once to
the file. I have included the code I am writing. I son't have
much experience in PL/SQL, I usually write in ABAP or Java so
the looping and if logic may seem incongruent as I have never
implemented this in PL/SQL before. Thanks in advance, all the
answers I have gotten fomr this forum have been great.
CURSOR C1
IS
SELECT PCODE, PDESC FROM gpcomp1.gppcat;
CURSOR C2
IS
SELECT PCODE, PDESC FROM gpcomp2.gppcat;
CURSOR C3
IS
SELECT PCODE, PDESC FROM gpcomp3.gppcat;
CURSOR C4
IS
SELECT PCODE, PDESC FROM gpcomp4.gppcat;
-- error handling routine. we will handle all file access errors
-- and data access errors in this procedure err_processing
PROCEDURE err_processing (p_errFileHandle UTL_FILE.FILE_TYPE,
p_lineNum PLS_INTEGER,
p_errText VARCHAR2)
IS
BEGIN
lv_errorFile := UTL_FILE.FOPEN(p_errDir,p_errFile,'W');
UTL_FILE.PUT_LINE( lv_errorFile, lv_initErrorLog );
UTL_FILE.FFLUSH(lv_errorFile);
UTL_FILE.FCLOSE(lv_errorFile);
UTL_FILE.FCLOSE_ALL;
END err_processing;
-- Retreive data into cursors
BEGIN
catFileHandle := UTL_FILE.FOPEN(p_dirName,p_fileName,'W');
Here we need to loop through 4 companies and write all Problem
Codes (pcode) and
Problem Descriptions (pdesc) without writing duplicates. In this
case we are using
the same logic in the old extract by using the US company as the
driving table.
We compare the pcode, record by record in eanch 4 companies.
When we find a unique
record we write it to the file.
FOR catLine1 IN C1
LOOP
     FOR catLine2 IN C2
     LOOP
     IF catLine2.pcode = catLine1.pcode THEN
     writeLine:= (catLine1.pcode) || ',' ||
(catLine1.pdesc);
          UTL_FILE.PUT_LINE( catFileHandle, writeLine );
     ELSE
     writeLine:= (catLine2.pcode) || ',' ||
(catLine2.pdesc);
          UTL_FILE.PUT_LINE( catFileHandle, writeLine );
     END IF;
     END LOOP;
     FOR catLine3 IN C3
     LOOP
     IF catline3.pcode = catLine1.pcode THEN
     writeLine:= (catLine1.pcode) || ',' ||
(catLine1.pdesc);
          UTL_FILE.PUT_LINE( catFileHandle, writeLine );
     ELSE
     writeLine:= (catLine3.pcode) || ',' ||
(catLine3.pdesc);
          UTL_FILE.PUT_LINE( catFileHandle, writeLine );
     END IF;
     END LOOP;
     FOR catLine4 IN C4
     LOOP
     IF catline4.pcode = catLine1.pcode THEN
     writeLine:= (catLine1.pcode) || ',' ||
(catLine1.pdesc);
          UTL_FILE.PUT_LINE( catFileHandle, writeLine );
     ELSE
     writeLine:= (catLine4.pcode) || ',' ||
(catLine4.pdesc);
          UTL_FILE.PUT_LINE( catFileHandle, writeLine );
     END IF;
     END LOOP;
     EXIT WHEN C1%NOTFOUND;
     UTL_FILE.PUT_LINE( catFileHandle, writeLine );
END LOOP;
UTL_FILE.FFLUSH(catFileHandle);
UTL_FILE.FCLOSE(catFileHandle);
Thanks again.
-Pat

Answer 1

The easiest way of doing this is to use the UNION operator.
You just have one cursor, viz:
CURSOR C1
IS
SELECT PCODE, PDESC FROM gpcomp1.gppcat
UNION
SELECT PCODE, PDESC FROM gpcomp2.gppcat
UNION
SELECT PCODE, PDESC FROM gpcomp3.gppcat
UNION
SELECT PCODE, PDESC FROM gpcomp4.gppcat;
This will return only unique rows (to get duplicates you use
UNION ALL). Then you can use a single CURSOR LOOP structure to
write out the result set:
FOR lrec IN C! loop
utl_file.put_line(C1.pcode);
END LOOP;
HTH, APC