• Michael Covington (4/3/2008)


    I have a variable length flat file that is defined by a "record type" value (first two characters in each row), for example:

    01value1value2value3valueINEEDvalue4value5

    02newval1newval2newval3

    03anotherval1

    04differentval1differentval2differentval3differentval4differentval5differentval6

    I can read the source file using a "flat file source", use a "script component" to break the file row into 2 columns ( rectype and rowdata), I can then use a "conditional split" to push the data to different destination sources (tables - flat files).

    The problem is that the file has to be processed sequentially, there is a value, or column, in the "01" record type rows that I need to relate the following record type rows to the "parent" (01 record type rows). How can I grab a value from one column in the "01" type rows, and add that value as an new column in the Non 01 rows?

    column I need is listed as "valueINEED"

    example output:

    01value1value2value3valueINEEDvalue4value5

    02newval1newval2newval3valueINEED

    03anotherval1valueINEED

    04differentval1differentval2differentval3differentval4differentval5differentval6valueINEED

    You don't necessarily need a script for that. Assuming there's a way to identify the pattern (your test data doesn't give me a hint on how to), take a look at what I posted over in this thread.

    http://www.sqlservercentral.com/Forums/FindPost478822.aspx

    You can then use the derived column transform to parse the stuff, and then your data "stays" with the identity column.

    ----------------------------------------------------------------------------------
    Your lack of planning does not constitute an emergency on my part...unless you're my manager...or a director and above...or a really loud-spoken end-user..All right - what was my emergency again?