LAE

Welcome to the LAE community!  Please feel free to start a discussion in the discussion tab or join in a conversation.

Discussions

Members

Resources

Events

 View Only

  • 1.  Delimited File node without record delimiter

    Employee
    Posted 03-25-2010 14:29

    Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

    Originally posted by: JimmyT

    Delimited File node without record delimiter. How do i import an input file with a pipe "|" delimiter without the record delimiter? the Delimited file node doesn't run without it


  • 2.  RE: Delimited File node without record delimiter

    Employee
    Posted 03-26-2010 08:04

    Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

    Originally posted by: timonk

    Jimmy,
    Good Day. I am not sure I fully understand your situation. Are you saying that you have a datafile which is field-delimited by a pipe "|", but does not have any distinct record-delimiter?

    Many record data files simply separate records by line, so your Record Delimiter in the delimited file node would be "\n" (the control character for newline). Else they use some character if the data is a single constant stream. (A|B|C@D|E|F, and you would use "@" as the record delimiter)

    If neither of these situations quite match yours, could you perhaps post an example of the data you are trying to import?

    Regards,
    Timon Koufopoulos
    MDA Support.


  • 3.  RE: Delimited File node without record delimiter

    Employee
    Posted 03-26-2010 08:11

    Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

    Originally posted by: ejones

    Questions:
    Is it a one record file? If you open it with a a text editor Excel see just one record or multiple records?

    If it is a one record file, then you could put something in the RecordDelimiter field of the Delimited File node type that will not appear in the file. If a text editor indicates that there are multiple records, then try one of these:
    \n - if the file is on or came from a Unix system
    \r\n - if the file is on or came from a Windows system
    \r - might also be a possibility but I've never seen it.

    Is there some other way that records are identified for example, maybe there are a certain number of characters per record? An IBM Mainframe might produce a file like this and you would need to use the Fixed Format File node type and then split up the fields in filter node using split() and "|" as the second parameter.

    Ernest