Question

My stored procedure looks like follows:

  sqlQuery := 'DROP INDEX idArchivoIndex';
  EXECUTE IMMEDIATE sqlQuery;

  EXCEPTION --En caso de que no exista el índice capturamos la excepcion
    WHEN index_not_exists THEN NULL; --y la ignoramos

  sqlQuery := 'CREATE INDEX idArchivoIndex'||
              ' ON '||qusuario||' (id_archivo)';
  EXECUTE IMMEDIATE sqlQuery;

  doresetvalidacion(qusuario, idarchivo);

  IF (tipoDependencia = 'PEC') THEN
    dovalidapec(qusuario,qaniofiscal,idarchivo,imprimirMensajes);
    COMMIT;
  ELSIF (tipoDependencia = 'SAGARPA') THEN
    dovalidacionpec(qusuario,qaniofiscal,idarchivo,imprimirMensajes);
    COMMIT;
  END IF;

If the exception is not raised the procedure just drops the index but no index is recreated ! I thought that this part of the code

EXCEPTION
    WHEN index_not_exists THEN NULL;

Handled the error and then continue with the code below it. Now that I see the results what’s after the EXCEPTION is executed if and only if the exception was raised.

What I want is to simplify my code, I don’t want to copy-paste the same block of code before the EXCEPTION clause just to make it work as I expect. Is there a way to achieve it? Maybe with a nested BEGIN ... END block? Or will I have to make a separate procedure to reuse code?

Cheers.

UPDATE

create or replace
PROCEDURE DOVALIDAINFORMACION 
(
  QARCHIVO IN VARCHAR2
, QUSUARIO IN VARCHAR2
, QANIOFISCAL IN VARCHAR2
) AS
  imprimirMensajes CHAR;
  tipoDependencia VARCHAR2(25);
  idArchivo NUMBER;
  sqlQuery VARCHAR2(100);
  index_not_exists EXCEPTION;
  PRAGMA EXCEPTION_INIT(index_not_exists, -1418);
BEGIN

  sqlQuery := 'DROP INDEX idArchivoIndex';
  EXECUTE IMMEDIATE sqlQuery;
  ----------------------
  EXCEPTION --En caso de que no exista el índice capturamos la excepcion
    WHEN index_not_exists THEN --y la ignoramos
      NULL;
  END;
  ----------------------
  sqlQuery := 'CREATE INDEX idArchivoIndex'||
              ' ON '||qusuario||' (id_archivo)';
  EXECUTE IMMEDIATE sqlQuery;

  doresetvalidacion(qusuario, idarchivo);

  IF (tipoDependencia = 'PEC') THEN
    dovalidapec(qusuario,qaniofiscal,idarchivo,imprimirMensajes);
    COMMIT;
  ELSIF (tipoDependencia = 'SAGARPA') THEN
    dovalidacionpec(qusuario,qaniofiscal,idarchivo,imprimirMensajes);
    COMMIT;
  END IF;

END DOVALIDAINFORMACION;

But can’t compile the procedure.

Error(32,3): PLS-00103: Se ha encontrado el símbolo "SQLQUERY"
Error(33,48): PLS-00103: Se ha encontrado el símbolo ";" cuando se esperaba uno de los siguientes:     ) , * & = - + < / > at in is mod remainder not rem    <an exponent (**)> <> or != or ~= >= <= <> and or like LIKE2_    LIKE4_ LIKEC_ between || member SUBMULTISET_ 

Answer 1

I suspect that are just missing an extra BEGIN in your updated code. An EXCEPTION clause always matches to a BEGIN and an END. In the code that you posted, the EXCEPTION matches the procedure’s BEGIN. You need it to match the BEGIN of the nested PL/SQL block.

create or replace
PROCEDURE DOVALIDAINFORMACION 
(
  QARCHIVO IN VARCHAR2
, QUSUARIO IN VARCHAR2
, QANIOFISCAL IN VARCHAR2
) AS
  imprimirMensajes CHAR;
  tipoDependencia VARCHAR2(25);
  idArchivo NUMBER;
  sqlQuery VARCHAR2(100);
  index_not_exists EXCEPTION;
  PRAGMA EXCEPTION_INIT(index_not_exists, -1418);
BEGIN
  BEGIN
    sqlQuery := 'DROP INDEX idArchivoIndex';
    EXECUTE IMMEDIATE sqlQuery;
  EXCEPTION --En caso de que no exista el índice capturamos la excepcion
    WHEN index_not_exists THEN --y la ignoramos
      NULL;
  END;

  sqlQuery := 'CREATE INDEX idArchivoIndex'||
              ' ON '||qusuario||' (id_archivo)';
  EXECUTE IMMEDIATE sqlQuery;

  doresetvalidacion(qusuario, idarchivo);

  IF (tipoDependencia = 'PEC') THEN
    dovalidapec(qusuario,qaniofiscal,idarchivo,imprimirMensajes);
    COMMIT;
  ELSIF (tipoDependencia = 'SAGARPA') THEN
    dovalidacionpec(qusuario,qaniofiscal,idarchivo,imprimirMensajes);
    COMMIT;
  END IF;
END DOVALIDAINFORMACION;

As an aside, it seems odd to drop and then immediately re-create an index in a PL/SQL block. If this is somehow related to your question about recreating an index after a load, I’m afraid that you may have misunderstood my answer. In my earlier answer, I was pointing out that it may be more efficient to drop the index, load your 10 million rows of data, and then re-create the index. Assuming that the loads are happening in the stored procedure calls you are making in this code, you would want the index to be re-created after the loads are complete.