Connect DataFiles & Apps
JSON File
You connect JSON File from the DataSquares connection wizard. Upload a .json or .ndjson file — records are flattened and staged as a table.
-
In DataSquares, open Data Sources and choose Connect Source. The Connect a data source page opens.
-
Pick JSON File on that page (search or browse by category).
-
Name the source, choose the file, then choose Upload & Create.
Uploading the file
Section titled “Uploading the file”Pick JSON File on the Connect a data source page and upload your file — no credentials
needed. Records (including .ndjson) are flattened and staged as a table.
Choosing records
Section titled “Choosing records”Upload a JSON array of objects, an object containing one record array, or a
single object. When several arrays exist, set Records path to the intended
array, such as result.items.
NDJSON contains one record per non-blank line. Use objects consistently, or use
scalar values consistently; scalar records become a value column. A records
path is unnecessary and is not supported for NDJSON.
Nested object keys become dotted column names. A field a record does not have,
and a JSON null, are NULL; a real empty string ("") stays an empty string in
a text column, and becomes NULL in a number, boolean or date column (it cannot
be one). Later records can add columns. Large integer and decimal values keep their
precision. Duplicate keys, conflicting dotted names and mixed object/scalar
records are refused so the import cannot silently discard data.
File size and replacements
Section titled “File size and replacements”A JSON document has a server parsing limit of 32 MiB by default, separate from the general upload limit, because the whole document must be parsed at once. A larger document is refused with HTTP 413 before parsing; split it or use a pipeline. Your administrator can change the parsing ceiling to match server capacity. NDJSON is read one line at a time, so an NDJSON file may be larger than the ceiling; only a single line may not. Row, cell and column limits apply to both, including empty cells created by records with different keys. A records path always means one document, so it is held to the ceiling.
Use Replace file to update a saved source. The preview checks the proposed file and reports schema changes before publication. An oversized or invalid replacement leaves the existing data in place.
Uploads and replacements are imported in the background, with progress and Cancel in the app — see CSV: importing in the background.
Related
Section titled “Related”- Connector catalog — everything DataSquares connects to.
- Quickstart — from connection to shared dashboard.