Data REST API pagination

You can paginate the JSON response that is called from the REST API. The order of the data is retained from page to page. Given the ability to paginate, you can quickly populate tables and make new REST calls every time you go to the next page of the data on the table. There is significant load time if you want to populate the data table with many rows (greater than 1000) from the Data REST API.

To paginate results in your API response, you must add new parameters to the query:

PageSize determines the number of rows to be included.

{
    "name": "pagesize",
    "description": PageSize: The number of rows.",
    "defaultValue": "-1",
    "type": "integer"
}

Offset determines the starting point.

{
    "name": "offset",
    "description": Offset: The starting point",
    "defaultValue": "-1",
    "type": "integer"
}

PageNumber is an alternate way to determine the offset. You must make a call with pageNumber = 1 first. Then you can access any page. Calling with pageNumber != 1 as the initial call will fail. pageNumber = 0 is not a valid value.

{
    "name": "pagenumber",
    "description": PageNumber: This is an alternate way to set offset. This is 1-based
                   indexing. Offset = (pageNumber - 1) * pageSize.
    "defaultValue": "-1",
    "type": "integer"
}

FormatType is the JSON format type.

{
    "name": "formattype",
    "description": FormatType: This sets the JSON format type. Values that are allowed are
                   FULL and COMPACT.
    "defaultValue": "COMPACT",
    "type": "string"
}

COMPACT is the default type, and is formatted as follows: ['col1', 'col2'] [1, 'a']. While FULL is formatted like this: {'col1': 1 'col2': 'a'}

Example

The following example shows ThoughtSpot data that is being populated in a table:

 /**
 * Sample response for Page-1.
 */
{
    "totalRowCount": 1500,
    "pageSize": 100,
    "pageNumber": 1
        "data":
    [
       {
            "key1": "value1",
            "key2": "value2",
        },
        {
            "key1": "value1",
            "key2": "value2",
        },
    ]
}

Was this page helpful?