CSV to JSON and JSON to CSV: What Can Go Wrong

CSV looks simple because rows and commas are familiar, but there is no single shape that can represent every JSON document.

CSV works best for tables A conventional CSV file has one header row followed by records with matching columns. Cells may contain commas, quotes or new lines when they are correctly quoted. A converter needs to understand those rules instead of splitting every line on commas.

JSON can be nested JSON supports objects inside objects and arrays inside records. A flat CSV table does not have a universal way to preserve that structure. Some systems flatten nested keys with dots, others store nested data as JSON strings, and others create multiple files.

Choose the flattening rule yourself If your JSON has nested values, decide how the receiving system expects them. Flattening `customer.address.city` may be useful for analytics, while a database import might need separate tables.

Data types can change CSV cells are text. A value such as `00123` may need to stay text even though a parser might interpret it as a number. Dates, booleans and null values also need an agreed convention.

Safe conversion habit Test with a small sample, inspect the headers, check quoted fields, confirm leading zeros and then compare record counts before using the converted output.