CSV to JSON Conversion Guide: Beginner to Expert

Converting CSV to JSON is common for APIs, configs, and data pipelines. This guide covers simple conversions to expert handling of edge cases.

Beginner: Basic CSV to JSON

CSV has rows and columns. The first row is often the header; each following row becomes an object with keys from the header.

1name,age,city
2Alice,30,NYC
3Bob,25,LA

Becomes:

1[
2  {
3    "name": "Alice",
4    "age": "30",
5    "city": "NYC"
6  },
7  {
8    "name": "Bob",
9    "age": "25",
10    "city": "LA"
11  }
12]

Note: CSV has no types; numbers and booleans are strings unless you parse them.

Intermediate: Parsing Correctly (Commas and Quotes)

CSV allows quoted fields; commas and newlines inside quotes are part of the value. Use a proper parser (e.g. Papa Parse, csv-parse) instead of simple split(','):

1// Node: csv-parse
2const { parse } = require('csv-parse/sync');
3const rows = parse(csvString, { columns: true, skip_empty_lines: true });
4const json = JSON.stringify(rows, null, 2);

Options: columns: true uses first row as keys; trim: true trims spaces; relax_column_count handles ragged rows.

Advanced: Nested JSON and Types

For nested JSON in cells, parse the cell value with JSON.parse() when safe, or use a convention (e.g. dot-notation user.name → nested object).

Types: Coerce numeric columns (e.g. age) to numbers; detect booleans or dates if your CSV follows a convention.

Expert: Large Files and Streaming

After conversion, validate and format with our [JSON Formatter](/tools/json-formatter/). For other data tasks, see our [Text Transformer](/tools/text-transformer/) and developer tools.