You could use 'RESIZE' to convert your second input 'columns to rows'. But it is not so clear what you are trying to do. From your last post it looks like there is more than one record in your input-2 as opposed to what you have shown in your first example. Also what is the logic around this - 'A', 'C', 'I', 'P'....? Will you have other values in this field?
Can you show a more realistic sample input and expected output out of it.
Will the record count in Input2 always be TWICE the number of records in Input1? If not, include such an example and explain how you want to handle that. Will Input-2 always have 3 and 2 byte fields in alternating sequence as shown in your example, or can those have random lengths? I would suggest you re-read this thread and answer the questions posted (including the ones you ignored).
To me it looks like the OP wants to replace the last character in Input2 with a RESIZEd version of Input1 (one character each in each record). But I will let him confirm.
Joined: 10 May 2007 Posts: 2180 Location: Hampshire, UK
No. It looks like you consider the whole of data set 1 (ds1) as one long string and that there should be a 1-to-1 relationship with data set 2 (ds2). The first character in ds1 replaces the last character of the first record in ds2, the second character (ignoring the commas) in ds1 replaces the last character of the second record in ds2, etc, etc.
It looks like you consider the whole of data set 1 (ds1) as one long string
Not exactly a long string but yes, if we have the Input1 converted like this:
then each character from this, replaces the last character in each record in the other data set. Again this is not a proposed sort solution step, but trying to break up the requirement logically. Still questions unanswered, so hopefully the OP can come back and post some useful info.