Getting started

csv-pipe converts between arrays of objects and RFC 4180-compliant CSV, in both directions. It has no dependencies and runs in Node, browsers, Deno, Bun, and edge runtimes.

Install

npm

npm install csv-pipe

pnpm

pnpm add csv-pipe

yarn

yarn add csv-pipe

bun

bun add csv-pipe

Quick start

import { stringify } from 'csv-pipe';

type User = { name: string; email: string; age: number };

const users: User[] = [
  { name: 'Alex Johnson', email: 'alex@example.com', age: 29 },
  { name: 'Carlos Herrera', email: 'carlos@example.com', age: 24 }
];

stringify(users);
// name,email,age
// Alex Johnson,alex@example.com,29
// Carlos Herrera,carlos@example.com,24

The header comes from the record keys, and a field is quoted only when it contains the separator, a quote, or a line break. That is the whole API for the common case.

Parse it back

parse is the mirror of stringify, turning CSV into typed records.

import { parse } from 'csv-pipe';

type User = { name: string; email: string; age: number };

parse<User>('name,email,age\nAlex Johnson,alex@example.com,29');
// [{ name: 'Alex Johnson', email: 'alex@example.com', age: '29' }]

See the parsing guide for streaming, columns, and validation.

Reusing an encoder

createCsvEncoder resolves options once and returns a callable encoder with row and stream. Use it to encode many datasets with the same configuration. stringify(data, options) is shorthand for createCsvEncoder(options)(data).

import { createCsvEncoder } from 'csv-pipe';

const toCsv = createCsvEncoder<User>({ columns: ['name', 'email'] });

toCsv(users); // the whole document as a string
toCsv.row(users[0]); // one line, without a header

Next steps

• Why csv-pipe for the problems it removes.
• Choosing columns to select, order, and rename.
• Parsing to read CSV into typed records.
• Streaming for large datasets and HTTP responses.
• API reference for every export.