CSV to JSON

Two formats for two different jobs

CSV is what spreadsheets and exports produce: rows and columns, flat and simple. JSON is what APIs and applications consume: nested, typed and structured. Moving between them is one of the most common small tasks in development, usually when data arrives from a business team as a spreadsheet and has to be loaded by something that expects JSON.

Where CSV parsing goes wrong

CSV looks trivial and is not, because commas appear inside data as well as between fields. A properly formatted file wraps those fields in quotes, and a field can itself contain quotes, which then have to be escaped. Add line breaks inside a field and naive splitting on commas falls apart entirely. This is why hand written parsers fail on real exports while working perfectly on the sample.

Before you convert

• Make sure the first row is a genuine header row, since those become your JSON keys.
• Watch for duplicate column names, which will silently overwrite each other.
• Everything from CSV arrives as a string, so numbers and booleans may need converting after.
• Empty cells become empty strings rather than null, which is not always what you want.