>A method received a primary key like this:
>
>

>PARAMETER tnNumero
>

>
>That is an integer coming in from the table.
>
>The next line is using a PADL syntax:
>
>

>ERASE (gcClientFat+'Images\SchoolPicture\'+PADL(tnNumero,6,'0')+'.jpg')
>

>
>Instead of being something like ..\000005.jpg, it gives something like ..\005.00.jpg. This is not the first time this is happening. I have seen that since VFP 9 at several occasions where I am using PADL(). The only way to be sure this will be properly interpreted, I have to do this:
>
>

>ERASE (gcClientFat+'Images\SchoolPicture\'+PADL(INT(tnNumero),6,'0')+'.jpg')
>

>
>How can I resolve the issue?

Hmm, interesting. I didn't know you could use PADL() with a general expression, I thought it had to be a string. So, I've always done something like

... + PADL( LTRIM( STR( tnNumero ) ), 6, "0" )

I think what you're seeing must be related to the fact that tnNumero cannot be an integer, as that is not a VFP data type (only a field type), so it's coming across as a numeric, with a value slightly different from what you expect. Maybe PADL() has trouble with a long string like 5.0000000000001, perhaps internally it tries to add another 6 sig figs and loses precision, or ??

You'll need to be careful with your code, if the value comes across as 4.999999999999 you'll be deleting using a value of 4 instead of 5. You might actually want to use

PADL( INT( ROUND( tnNumero, 0 ) ), 6, "0" )