select-csv

The fastest, simplest, and most powerful CSV parser for Node.js. Optimized for high performance and extreme memory efficiency when handling massive datasets (100M+ rows).

select-csv converts .csv files into arrays, JSON objects, or raw lines. It provides two main functions, parseCsv (for local files) and parseText (for raw strings), both sharing the same methods and features.

Key Features

• Ultra-Lightweight: Package size is less than 30 KB.
• Fast Mode: Synchronous execution for maximum speed and zero overhead.
• Memory Efficient: Streams large files using chunks and row offsets instead of loading the entire file into RAM.
• Robust Parsing: Correct handling of complex line-breaks (\r\n, \n) and nested quotations.
• No Dependencies: Zero external dependencies for maximum security and stability.
• Flexible Options: Highly customizable (headers, quotes, delimiters, buffers).

Install:

select-csv is available on npm. It can be installed with the following command:

npm install select-csv

Usage:

Here there are clearly different examples

const {parseCsv,parseText} = require("select-csv");

var parse;

// First create object from .csv file
parse = parseCsv('file_path.csv');

// Or if you just want create object from text
parse = parseText(
`Index,User Id,First Name,Last Name,Sex
1,5f10e9D33fC5f2b,Sara,Mcguire,Female
2,751cD1cbF77e005,Alisha,Hebert,Male
3,DcEFDB2D2e62bF9,Gwendolyn,Sheppard,Male
4,C88661E02EEDA9e,Kristine,Mccann,Female
5,fafF1aBDebaB2a6,Bobby,Pittman,Female
6,BdDb6C8Af309202,Calvin,Ramsey,Female
7,FCdfFf08196f633,Collin,Allison,Male
8,356279dAa0F7CbD,Nicholas,Branch,Male
9,F563CcbFBfEcf5a,Emma,Robinson,Female
10,f2dceFc00F62542,Pedro,Cordova,Male`,
{ 
    linebreak: '\n',
    header: true 
}
);

• If you want to get all rows :

const result = parse.get(); //Return all rows
/*
	{
	Time: "0 ms",
	header:["Index","User Id","First Name","Last Name","Sex"],
	rows:[
		["1","5f10e9D33fC5f2b","Sara","Mcguire","Female"],
		["2","751cD1cbF77e005","Alisha","Hebert","Male"],
		["3","DcEFDB2D2e62bF9","Gwendolyn","Sheppard","Male"],
		["4","C88661E02EEDA9e","Kristine","Mccann","Female"],
		["5","fafF1aBDebaB2a6","Bobby","Pittman","Female"],
		["6","BdDb6C8Af309202","Calvin","Ramsey","Female"],
		["7","FCdfFf08196f633","Collin","Allison","Male"],
		["8","356279dAa0F7CbD","Nicholas","Branch","Male"],
		["9","F563CcbFBfEcf5a","Emma","Robinson","Female"],
		["10","f2dceFc00F62542","Pedro","Cordova","Male"]
		],
	row_count:10
	}
*/

• If you want to get a chunks of rows :

var result;
result = parse.chunk(c) 
//The 'c' parameter must be an integer and greater than or equal to 1

//Examples:
result = parse.chunk(2) //Return row 0 and 1 
/*
{
  Time: "0 ms",
  "Header": [ "Index", "User Id", "First Name", "Last Name", "Sex" ],
  "Rows": [
    [ "1", "5f10e9D33fC5f2b", "Sara", "Mcguire", "Female" ],
    [ "2", "751cD1cbF77e005", "Alisha", "Hebert", "Male" ]
  ],
  "row_count:": 2
}
*/

result = parse.chunk(3) //Return row 2,3 and 4 (Get rows from last offset saved)
/*
{
  Time: "0 ms",
  "Header": [ "Index", "User Id", "First Name", "Last Name", "Sex" ],
  "Rows": [
    [ "3", "DcEFDB2D2e62bF9", "Gwendolyn", "Sheppard", "Male" ],
    [ "4", "C88661E02EEDA9e", "Kristine", "Mccann", "Female" ],
    [ "5", "fafF1aBDebaB2a6", "Bobby", "Pittman", "Female" ]
  ],
  "row_count:": 3
}
*/

result = parse.chunk(1) //Return row 5 (Get rows from last offset saved)
/*
{
  Time: "0 ms",
  "Header": [ "Index", "User Id", "First Name", "Last Name", "Sex" ],
  "Rows": [
    [ "6", "BdDb6C8Af309202", "Calvin", "Ramsey", "Female" ]
  ],
  "row_count:": 1
}
*/

• If you want to get specific rows :

var result
result = parse.rowOffset(from) 
// The 'from' parameter must be an integer and greater than or equal to 0

// Or
result = parse.rowOffset(from,to)
// The 'to' parameter must be an integer and greater than or equal to 1

//Examples:
result = parse.rowOffset(6) //Returns all rows from the sixth row to the last row
/*
{
  Time: "0 ms",
  "Header": [ "Index", "User Id", "First Name", "Last Name", "Sex" ],
  "Rows": [
    [ "7", "FCdfFf08196f633", "Collin", "Allison", "Male" ],
    [ "8", "356279dAa0F7CbD", "Nicholas", "Branch", "Male" ],
    [ "9", "F563CcbFBfEcf5a", "Emma", "Robinson", "Female" ],
    [ "10", "f2dceFc00F62542", "Pedro", "Cordova", "Male" ]
  ],
  "row_count:": 4
}
*/

result = parse.rowOffset(5,8) //Returns all rows from 5th to 8th row
/*
{
  Time: "1 ms",
  "Header": [ "Index", "User Id", "First Name", "Last Name", "Sex" ],
  "Rows": [
    [ "6", "BdDb6C8Af309202", "Calvin", "Ramsey", "Female" ],
    [ "7", "FCdfFf08196f633", "Collin", "Allison", "Male" ],
    [ "8", "356279dAa0F7CbD", "Nicholas", "Branch", "Male" ]
  ],
  "row_count:": 3
}
*/

• If you want to change the row offset :

parse.setRowOffset(offs) 

// The 'offs' parameter must be an integer and greater than or equal to 0.

// If the offset exists, return [offset,row_number].
result = parse.setRowOffset(5) 
/*
[236,5]
*/
result = parse.chunk(1) // Get rows from last offset saved
/*
{
  Time: "0 ms",
  "Header": [ "Index", "User Id", "First Name", "Last Name", "Sex" ],
  "Rows": [
    [ "6", "BdDb6C8Af309202", "Calvin", "Ramsey", "Female" ]
  ],
  "row_count:": 1
}
*/

// If not , returns false and the offset not changed.
result = parse.setRowOffset(20) 
/*
false
*/
result = parse.chunk(1) // Get rows from last offset saved
/*
{
  Time: "0 ms",
  "Header": [ "Index", "User Id", "First Name", "Last Name", "Sex" ],
  "Rows": [
    [ "7", "FCdfFf08196f633", "Collin", "Allison", "Male" ]
  ],
  "row_count:": 1
}
*/

• The default object option :

{
	'header': true,
	'quote': false,
	'linebreak': '\r\n',
	'delimiter': ",",
  'bufferSize':1024*1024 
}
	// delimiter: (String: get rows containing columns, false: get lines without columns)
  //bufferSize: It only works with a CSV file, which is the maximum number of characters that can be read at a time, the minimum value is 1024

• If you want to use specific option :

var option = {
	'header': false, 	/* or true */
	'quote': true, 		/* or false */
	'linebreak': '\n', 	/* '\n' or '\r' or any other string  */
	'delimiter': ","	/* ';' or any other string or false */
  'bufferSize':2000 /* It only works with a CSV file */
}

var parse;
// Create object from .csv file
parse = parseCsv('file_path.csv',option);

// Or if you just want create object from text
parse = parseText(
`Index,User Id,First Name,Last Name,Sex
1,5f10e9D33fC5f2b,Sara,Mcguire,Female
2,751cD1cbF77e005,Alisha,Hebert,Male
3,DcEFDB2D2e62bF9,Gwendolyn,Sheppard,Male`
, option);

option = {
	'header': false,
}

result = parse.rowOffset(2)
/*
{
  Time: "0 ms",
  "Header": false,
  "Rows": [
    [ "2", "751cD1cbF77e005", "Alisha", "Hebert", "Male" ]
  ],
  "row_count:": 1
}
*/

option = { 
	'header': true,
	'delimiter': false 
}
	// delimiter: (String: get rows containing columns, false: get lines without columns)
/*
{
  Time: "0 ms",
  "Header": false,
  "Rows": [
    [ "2,751cD1cbF77e005,Alisha,Hebert,Male" ] // No columns, just string (all line)
  ],
  "row_count:": 1
}
*/

• If you want to reset option after multiple uses of your code :

const option = {       // Just an exapmle
	'header': false,
	'quote': true,
	'linebreak': '\n'
}

parse.resetOption(option); // All saved values are erased and the object is restared again

• If you want to get information of your object :

const result = parse.getInfo();
/*
{
  "offset": 275,
  "rowOffset": 7,
  "option": {
    "header": false,
    "quote": false,
    "linebreak": "\n",
    "delimiter": false
  }
}
*/

• Examples of parsing a large CSV file: (https://www.kaggle.com/datasets/zanjibar/100-million-data-csv)

const parse = parseCsv('100-million-data.csv',{"header": false});
var result;
result = parse.chunk(100000)
/*
{
  time: 222,
  header: false,
  rows: [
    [ '198801', '1', '103', '100', '000000190', '0', '35843', '34353' ],
    [ '198801', '1', '103', '100', '120991000', '0', '1590', '4154' ],
    [ '198801', '1', '103', '100', '210390900', '0', '4500', '2565' ],
    .
    .
    .
    [ '198801', '1', '103', '100', '391590000', '0', '95000', '7850' ],
    [ '198801', '1', '103', '100', '391620000', '0', '1000', '404' ],
    [ '198801', '1', '103', '100', '391723000', '0', '545', '479' ],
    [ '198801', '1', '103', '100', '391732100', '0', '24', '393' ],
    [ '198801', '1', '103', '100', '391732900', '0', '60', '758' ],
    [ '198801', '1', '103', '100', '391810100', '0', '1935', '1042' ],
    [ '198801', '1', '103', '100', '391910200', '0', '510', '1303' ],
    [ '198801', '1', '103', '100', '391910300', '0', '133', '379' ],
    [ '198801', '1', '103', '100', '391990300', '0', '450', '1668' ],
    [ '198801', '1', '103', '100', '391990500', '0', '942', '1721' ],
    [ '198801', '1', '103', '100', '391990900', '0', '40', '235' ],
    [ '198801', '1', '103', '100', '392030000', '0', '406', '652' ],
    ... 99900 more items
  ],
  row_count: 100000
}
*/

result = parse.chunk(3) // Return row 100001,100002 and 100003 (Get rows from last offset saved)
/*
{
  time: 1,
  header: false,
  rows: [
    [ '198801', '1', '326', '500', '841330000', '90', '81', '246' ],
    [ '198801', '1', '326', '500', '841510000', '0', '35', '1366' ],
    [ '198801', '1', '326', '500', '841582100', '0', '6', '334' ]
  ],
  row_count: 3
}
*/

const from = 1000*1000*30;
const to = from + 5;
result = parse.rowOffset(from,to)
/*
{
  time: 3743,
  header: false,
  rows: [
    [
      '199804',    '2',
      '213',       '502',
      '848130000', '16035',
      '746',       '8380'
    ],
    [ '199804', '2', '213', '502', '848140000', '168', '152', '1891' ],
    [ '199804', '2', '213', '502', '848180010', '77', '404', '1366' ],
    [ '199804', '2', '213', '502', '848190000', '0', '131', '570' ],
    [ '199804', '2', '213', '502', '848230000', '300', '4', '882' ]
  ],
  row_count: 5
}
*/

const from = 1000*1000*90;
const to = from + 4;
result = parse.rowOffset(from,to)
/*
{
  time: 44126,
  header: false,
  rows: [
    [ '201412', '1', '125', '400', '283525000', '0', '160000', '6492' ],
    [ '201412', '1', '125', '400', '390740100', '0', '17500', '5579' ],
    [ '201412', '1', '125', '400', '390950000', '0', '36000', '21423' ],
    [ '201412', '1', '125', '400', '392329000', '0', '520', '1413' ]
  ],
  row_count: 4
}
*/

result = parse.chunk(3) // Get rows from last offset saved ( row to,to+1 and to+2 )
/*
{
  time: 29,
  header: false,
  rows: [
    [ '201412', '1', '125', '400', '400932000', '0', '18', '526' ],
    [ '201412', '1', '125', '400', '401110000', '173', '1735', '1197' ],
    [ '201412', '1', '125', '400', '401120000', '133', '1707', '1099' ]
  ],
  row_count: 3
}
*/

result = parse.getInfo() // Get all the information
/*
{
  offset: 3599945660,
  rowOffset: 90000008,
  option: {
    header: false,
    quote: false,
    linebreak: '\r\n',
    delimiter: ',',
    bufferSize: 1048576
  }
}
*/

📄 License

MIT License