The Read JSON files activity reads JSON files using DuckDB’s read_json_auto function and returns the result as workflow data. It can read multiple JSON files, merge their rows, and align columns found across files.
Supported Features
Multiple file reading: Reads each input .json file and appends the rows.
DuckDB JSON auto-detection: Uses read_json_auto.
Configurable DuckDB parameters: Pass parameters such as format, records, columns, or other supported DuckDB JSON reader options.
Default parameters: Adds auto_detect = true and maximum_object_size = uint.MaxValue when they are not supplied.
Column alignment: Merges files and includes all columns discovered across successful reads.
Preview limit support: Stops reading more files when the preview limit is reached.
Input
Type
Required
Description
Files
Yes
One or more .json files. Non-JSON files are skipped with an error.
Non-JSON files are not read and produce an error such as orders.csv is not of type json.
Output
Field
Type
Description
Data
Array
Rows returned by DuckDB from all successfully read JSON files.
Errors
Array
Errors for invalid file types or failed reads.
Example Output
{
"Data": [
{
"CustomerId": 1,
"Name": "Alice",
"City": "Boston"
},
{
"CustomerId": 2,
"Name": "Bob",
"City": "Chicago"
}
],
"Errors": []
}
Configuration Fields
Field Name
Type
Required
Description
Parameters
Dictionary
No
Additional parameters passed directly to DuckDB read_json_auto. If auto_detect is not present, the activity adds auto_detect = true. If maximum_object_size is not present, the activity adds maximum_object_size = uint.MaxValue.
Conditional Field Rendering Rules
No conditional configuration fields are defined for this activity.