JSON

JSON files can be used as data input by uploading a file, specifying an _s3 _path (s3a://\), or providing an HDFS location.

• Multiline options can be set to false if the JSON structure is in single line. Default value for this is set to True.
• JSON can be flattened, by checking the _Flatten Data _checkbox, if the JSON structure is in hierarchical format.
• JSON elements can be selected by specifying element names in the columns section and can be renamed using them as keywords. E.g. if id, first__name and last_name are the only three elements to be selected from a JSON with many distinct elements: isbn as id, author.firstName as first_name and author._lastName as last_name.

Multi-line False sample:

{ "isbn": "123-456-222","lastname": "Doe","firstname": "Jane"}
{"isbn": "123-456-777","lastname": "Smith","firstname": "Jane"}

Multi-line True sample:

[
    { "isbn": "123-456-222","lastname": "Doe","firstname": "Jane"},
    {"isbn": "123-456-777","lastname": "Smith","firstname": "Jane"}
]

Info: Root level columns can be selected above in a multi-root JSON structure. For selection of nested columns, use the Select Columns shape from the palette.