Forum Discussion

IshanT1's avatar
IshanT1
New Contributor II
6 years ago
Solved

CSV Formatter Error

Hi,

I am trying to create a CSV file by querying data from snowflake. I have a parent pipeline which is passing output file name & path as parameter to my child pipeline through pipeline execute.

When I am validating the child pipeline, everything is working normal but when I am running the parent, I am getting below error on the child pipeline at CSV Formatter step.

“Snap errors: {ruuid=58b7a2f5ba6d563aff5d73d9_8faf50a9-40dc-441e-9003-454d2c650b2d, reason=Input document should be a flattened map data, but found: LinkedHashMap, label=CSV Formatter, resolution=Please make sure the input document a flattened map data., failure=Cannot format CSV data}”.

Below is the parent pipeline.

Child Pipeline

Please let me know if you need any further information and as always thanks for your help in advance.

5 Replies

  • dimitri_hristov's avatar
    dimitri_hristov
    Contributor
    6 years ago

    Hi @IshanT1, welcome to the Community 🙂

    So, at first glance, I’d venture to guess that when you run the child pipeline by calling it from the parent pipeline, in the output of the snap called ‘Summary File Data’ you have the $original object. Naturally, the $original object has nested fields, so it’s not a flat structure and the CSV Formatter exits with an error.
    If such is the case, then you should remove the $original object from the input to the CSV Formatter, for instance by means of a Mapper Snap.

    Hope this helps.
    BR,
    Dimitri

    • IshanT1's avatar
      IshanT1
      New Contributor II
      6 years ago

      Hi @dimitri.hristovski ,

      Thanks for your response.

      ‘Summary File Data’ snap is snowflake execute and running a very simple select statement.

      As suggested I tried to add a mapper between “Summary File Data” and CSV formatter but still getting the same error.

      Adding more snapshots in the thread about the settings of each snap in the child pipeline.

      Thanks,
      Ishan

      • dimitri_hristov's avatar
        dimitri_hristov
        Contributor
        6 years ago

        Hi,
        Given that I can’t see the state of the ‘Pass through’ property in the Mapper3 Snap, I’d try and unselect the checkbox if it is selected. That will ensure that if there are any other fields received on the Mapper Snap’s input, they won’t be forwarded to the output.

        BR,
        Dimitri

  • IshanT1's avatar
    IshanT1
    New Contributor II
    6 years ago

    Hi @dimitri.hristovski,

    Unchecking the ‘pass through’ checkbox in the Mapper did the trick.
    Thanks for your help again.

    Best,
    Ishan

    • dimitri_hristov's avatar
      dimitri_hristov
      Contributor
      6 years ago

      Hi,

      I just realized looking at the Settings in a Snowflake Execute Snap, that perhaps if you uncheck the “Pass Through” checkbox in the Snowflake Execute Snap, it might save you from adding the Mapper before the CSV Formatter Snap.

      BR,
      Dimitri