error when running customized batch input

[Former Member]

i recorded a batch input for transaction SM30 - just simply entering data,pressing 'save' and thats all...

afterwards i created a program using this record. it runs fine with the values i entered originally.

now i modified the program so that it runs in a loop, replacing my original values with the data structures corresponding fields.

for fields with type character it works fine. But when i run the program with numerical data fields in my table (and of course the table-field i want to enter it is also numerical) i receive a CONVT_NO_NUMBER error which comes from the 2nd line of this:

FORM BDC_FIELD USING FNAM FVAL.

IF FVAL <> NODATA.

CLEAR BDCDATA.

BDCDATA-FNAM = FNAM.

BDCDATA-FVAL = FVAL.

APPEND BDCDATA.

ENDIF.

ENDFORM.

this method was created by my record,i didnt change it...

so my question finally is: is there any -smart- other way to replace it other than:

FORM BDCFIELD USING FNAM FVAL.

data _fval(255) type c.

_fval = fval.

IF _FVAL <> NODATA.

CLEAR BDCDATA.

BDCDATA-FNAM = FNAM.

BDCDATA-FVAL = FVAL.

APPEND BDCDATA.

ENDIF.

ENDFORM.

[Former Member]

Why don't u move ur numerical data into a char field before doing the further processing?

data: l_char type c length ....

l_num type i.

l_char = l_num.

perform BDC_FIELD XXXX l_char.

FORM BDC_FIELD USING FNAM FVAL.

IF FVAL NODATA.

CLEAR BDCDATA.

BDCDATA-FNAM = FNAM.

BDCDATA-FVAL = FVAL.

APPEND BDCDATA.

ENDIF.

ENDFORM.

Regards,

Joy.

[Former Member]

Joyjit Ghosh:

thats exactly what my workaround is doing, just you do it outside the method and i do it inside for just defining it once (but needing to change the method's name everywhere).

i wouldnt prefer one of those two 'workarounds' to the other but i thought there was some mistake in my program or things i didnt consider or maybe a smarter solution....

(if i, btw when running the program choose a nodata-sign 0 instead of / it doesnt give the error,of course)