File Manipulation - Convert JSON to CSV

Overview

This script converts a .json file to .csv format and optionally drops specified columns. It is designed to run within Alli Workflows and can be used to prepare structured data for downstream processing or handoff to external systems.

Variables

Name	Reference	Type	Required	Default	Description
Input Folder	FOLDER_NAME	Alphanumeric			Name of the local folder on Workflows where the target file lives. If left blank, will look in the home directory.
Input Filename	FILE_NAME	Alphanumeric	✅		Name of the target file on Platform. Can be regex if "Match Type" is set accordingly.
Remove Columns From Output	ENABLE_DROP	Boolean	✅	True	Remove Columns from CSV Output
Columns to Drop	DROP_COLUMNS				Comma-Separated List of Columns To Remove from the Output

Script Workflow

Read Configuration
- Grabs FOLDER_NAME and FILE_NAME from environment variables.
- Verifies that the file is a valid .json.
Load JSON Input
- Parses the .json file.
- Supports either a single JSON object or a list of records.
Drop Columns (Optional)
- If ENABLE_DROP is "true", removes any keys listed in DROP_COLUMNS from each record.
Write Output
- Creates a .csv file in the same folder with the same base filename.
- Writes the resulting data with headers based on the keys of the first record.
Logging
- Prints confirmation of conversion.
- Prints dropped columns if any were removed.

Example

Given: products.json

CODE

[   {"id": "123", "name": "Shirt", "price": "29.99", "internal_notes": "New"},   {"id": "456", "name": "Hat", "price": "19.99", "internal_notes": "Old"} ]

With:

ENABLE_DROP=true
DROP_COLUMNS=internal_notes

Output: products.csv

CODE

id,name,price 123,Shirt,29.99 456,Hat,19.99

Alli Workflows Best Practices

Pair this step after a data API or transformation step that outputs JSON.
Use DROP_COLUMNS to sanitize or redact internal metadata.
Chain this step into data pipelines feeding into AI models, dashboards, or storage.

⚠️ Notes

The script expects well-formed JSON and will raise errors if the input is malformed.
Column order is determined by the keys in the first record.
Output file will have the same name as input but with a .csv extension.

🧾 Sample Log Output

CODE

Converted ./products.json to ./products.csv Dropped columns: ['internal_notes']