csv-to-array-browser

csv-to-array-browser

Parse CSV files into JavaScript objects — in the browser and Node.js — with zero dependencies.

• ✅ Works with File, Blob, ArrayBuffer, Buffer, or plain string
• ✅ Handles quotes ("a, b"), escaped quotes (""), CR/LF line endings
• ✅ Auto-detects common delimiters (,, ;, \t, |) or you can set one
• ✅ Optional type coercion ("42" → 42, "true" → true, "null" → null)
• ✅ Strips UTF-8 BOM so your first header isn’t corrupted
• ✅ Tiny surface area, TypeScript-ready

📦 Installation

npm i csv-to-array-browser

🚀 Usage

Browser (with file input)

import { parseCSV } from "csv-to-array-browser";

document.querySelector("#file").addEventListener("change", async (e) => {
  const file = e.target.files[0];
  const rows = await parseCSV(file, { sheet: undefined }); // sheet is ignored; CSV options shown below
  console.log(rows);
});

<input id="file" type="file" accept=".csv" />;

Node.js (with Buffer/ArrayBuffer)

import { readFileSync } from "node:fs";
import { parseCSV } from "csv-to-array-browser";

const buf = readFileSync("./example.csv"); // Buffer
const rows = await parseCSV(buf, { inferTypes: true });

console.log(rows);

📖 API

parseWorkbook(input, options?)

• input

  File | Blob | ArrayBuffer | Buffer

  (Works with browser files, Node Buffers, or raw ArrayBuffers)

• options (optional)

Option	Type	Default	Description
delimiter	string	auto	Field delimiter. If omitted, tries `, ; \t
headers	boolean | string[]	true	Header handling. true: first row is headers. false: synthesize headers (c0, c1, …). string[]: use these headers and skip the first CSV row.
preserveHeaderCase	boolean	false	When false, headers are trimmed + lowercased. When true, keep original casing.
comment	string	undefined	Lines starting with this (e.g. "#") are skipped.
skipEmptyLines	boolean	true	Drop empty rows.
defval	unknown	null	Value to use when a row has fewer cells than headers.
inferTypes	boolean	false	Coerce "true"/"false" → boolean, "null" → null, number-like strings → numbers; otherwise leave as strings.
trimCells	boolean	true	Trims cells when true

The parser automatically strips a UTF-8 BOM at the start of the file.

Examples

1) Basic CSV

const csv = `name,age,city
Alice,30,Paris
Bob,28,Berlin`;

await parseCSV(csv);
// [
//   { name: "Alice", age: "30", city: "Paris" },
//   { name: "Bob",   age: "28",  city: "Berlin" }
// ]

2) With type coercion

const csv = `name,age,active
Alice,30,true
Bob,28,false`;

await parseCSV(csv, { inferTypes: true });
// [
//   { name: "Alice", age: 30, active: true },
//   { name: "Bob",   age: 28, active: false }
// ]

3) Custom delimiter (semicolon)

const csv = `name;age;city
Alice;30;Paris`;

await parseCSV(csv, { delimiter: ";" });
// [{ name: "Alice", age: "30", city: "Paris" }]

4) Custom headers (replace the file’s header row)

const csv = `a,b
1,2`;

await parseCSV(csv, { headers: ["X", "Y"] });
// [{ x: "1", y: "2" }]

5) No headers (synthesize c0, c1, …)

const csv = `a,b
1,2`;

await parseCSV(csv, { headers: false });
// [
//   { c0: "a", c1: "b" },
//   { c0: "1", c1: "2" }
// ]

6) Quoted fields + escaped quotes

const csv = `name,quote
"Alice, Jr.","She said ""Hi"""`;

await parseCSV(csv);
// [{ name: "Alice, Jr.", quote: 'She said "Hi"' }]

7) Comments & CRLF

const csv = `# header
name,age\r\nAlice,30\r\nBob,28\r\n`;

await parseCSV(csv, { comment: "#" });
// [
//   { name: "Alice", age: "30" },
//   { name: "Bob", age: "28" }
// ]

8) Trim cells

const csv = `# header
name,age\r\n Alice ,30\r\nBob,28\r\n`;

await parseCSV(csv, { trimCells: "false" });
// [
//   { name: " Alice ", age: "30" },
//   { name: "Bob", age: "28" }
// ]

FAQ

Q: How big a CSV can it handle?

For typical UX (uploads < ~10–20 MB) it’s great. For very large files or progressive processing, consider a streaming approach (Node: csv-parse; browser: Web Streams + Worker).

Q: How does auto delimiter detection work?

It samples the first lines and tries common delimiters (, ; \t |), picking the one that yields the most consistent, widest table.

Q: Does it parse dates?

Not automatically. You can enable inferTypes for booleans/numbers/nulls and then post-process date fields as needed.

🛠 Development

Clone the repo and run:

npm install
npm run build
npm test

npm run build → Build dist files

npm test → Run unit tests (Vitest)

npm run pack:check → Preview files that will publish