molt-csv - Blazingly Fast CSV Handler

Top-tier CSV parser with 2-7x performance advantage and WASM acceleration.

Overview

@sylphx/molt-csv is a high-performance CSV parser and stringifier. It's competitive with the fastest libraries while providing excellent type conversion and WASM acceleration options.

Key Features

• 🚀 5.9x faster than papaparse on simple CSVs
• 🏆 2.7x faster than csv-parse on large files
• 🔧 Type conversion - Automatic number, boolean, and date conversion
• ⚡ WASM acceleration - Optional WebAssembly for large files
• 🛡️ Type-safe - Full TypeScript support
• 📊 Streaming support - Process large files efficiently
• 📦 Zero dependencies - Pure implementation
• 🎯 RFC 4180 compliant - Proper CSV specification support

Installation

npm install @sylphx/molt-csv

Quick Start

Parse CSV

import { parse } from '@sylphx/molt-csv'

const csv = `
name,age,city,country
Alice,30,New York,USA
Bob,25,London,UK
Charlie,35,Paris,France
`

const records = parse(csv)
console.log(records)
// [
//   { name: 'Alice', age: '30', city: 'New York', country: 'USA' },
//   { name: 'Bob', age: '25', city: 'London', country: 'UK' },
//   { name: 'Charlie', age: '35', city: 'Paris', country: 'France' }
// ]

Stringify CSV

import { stringify } from '@sylphx/molt-csv'

const records = [
  { name: 'Alice', age: 30, city: 'New York' },
  { name: 'Bob', age: 25, city: 'London' }
]

const csv = stringify(records)
console.log(csv)
// name,age,city
// Alice,30,New York
// Bob,25,London

Type Conversion

import { parse } from '@sylphx/molt-csv'

const csv = `
id,name,active,score
1,Alice,true,95.5
2,Bob,false,87.3
3,Charlie,true,92.1
`

const records = parse(csv, {
  columns: {
    id: Number,
    active: Boolean,
    score: Number
  }
})

console.log(records[0])
// { id: 1, name: 'Alice', active: true, score: 95.5 }

Performance

Benchmarks showing molt-csv competitiveness:

Parsing Performance

Test Case	molt-csv	vs papaparse	vs csv-parse	vs d3-dsv
Simple (5 rows)	775,770 ops/s	5.9x faster	7.0x faster	0.96x
Complex (10r×9c)	105,992 ops/s	2.84x faster	4.29x faster	0.57x
Large (1000 rows)	1,591 ops/s	1.39x faster	2.53x faster	0.40x
Quoted fields	418,311 ops/s	7.65x faster	5.84x faster	1.0x

Serialization Performance

Test Case	molt-csv	vs papaparse	vs csv-parse	vs d3-dsv
Simple (5 rows)	637,435 ops/s	1.52x faster	1.37x faster	0.83x
Complex (10r×9c)	133,509 ops/s	1.83x faster	1.27x faster	0.77x

CSV Features

Basic Parsing

import { parse } from '@sylphx/molt-csv'

const csv = `name,email,age
Alice,alice@example.com,30
Bob,bob@example.com,25
Charlie,charlie@example.com,35`

const records = parse(csv)

Quoted Fields

const csv = `
name,description,price
"Product A","Contains, commas and ""quotes""",19.99
"Product B","Normal description",29.99
`

const records = parse(csv)
console.log(records[0])
// {
//   name: 'Product A',
//   description: 'Contains, commas and "quotes"',
//   price: '29.99'
// }

Custom Delimiters

const tsv = `
name	age	city
Alice	30	New York
Bob	25	London
`

const records = parse(tsv, { delimiter: '\t' })

Headers Configuration

const csv = `
Alice,alice@example.com,30
Bob,bob@example.com,25
Charlie,charlie@example.com,35
`

const records = parse(csv, {
  headers: ['name', 'email', 'age']
})

console.log(records[0])
// { name: 'Alice', email: 'alice@example.com', age: '30' }

Type Conversion

const csv = `
id,username,active,balance,joined
1,alice,true,1500.50,2024-01-15
2,bob,false,2300.75,2024-02-20
`

const records = parse(csv, {
  columns: {
    id: Number,
    active: Boolean,
    balance: Number,
    joined: (val) => new Date(val)
  }
})

console.log(records[0])
// {
//   id: 1,
//   username: 'alice',
//   active: true,
//   balance: 1500.50,
//   joined: Date(2024-01-15)
// }

Custom Type Converters

const records = parse(csv, {
  columns: {
    id: Number,
    tags: (val) => val.split(';'),
    metadata: JSON.parse,
    price: (val) => parseFloat(val)
  }
})

API Reference

parse(input, options?)

Parse CSV into JavaScript objects.

function parse(
  input: string,
  options?: ParseOptions
): Record<string, any>[]

Options:

• delimiter?: string - Field delimiter (default: ',')
• headers?: string[] - Column headers
• columns?: Record<string, Function> - Type converters per column
• skip_empty_lines?: boolean - Skip empty lines (default: true)
• trim?: boolean - Trim values (default: false)

stringify(records, options?)

Stringify JavaScript objects to CSV.

function stringify(
  records: Record<string, any>[],
  options?: StringifyOptions
): string

Options:

• delimiter?: string - Field delimiter (default: ',')
• header?: boolean - Include header row (default: true)
• columns?: string[] - Specific columns to include
• quote?: boolean - Quote fields (default: true)

Common Patterns

Reading CSV Files

import { readFileSync } from 'fs'
import { parse } from '@sylphx/molt-csv'

const csv = readFileSync('data.csv', 'utf-8')
const records = parse(csv)

Writing CSV Files

import { writeFileSync } from 'fs'
import { stringify } from '@sylphx/molt-csv'

const records = [
  { name: 'Alice', age: 30 },
  { name: 'Bob', age: 25 }
]

const csv = stringify(records)
writeFileSync('output.csv', csv)

Data Analysis Pipeline

import { parse } from '@sylphx/molt-csv'

const csv = readFileSync('sales.csv', 'utf-8')

const records = parse(csv, {
  columns: {
    date: (val) => new Date(val),
    amount: Number,
    quantity: Number
  }
})

// Filter and transform
const filtered = records
  .filter(r => r.amount > 100)
  .map(r => ({
    ...r,
    totalValue: r.amount * r.quantity
  }))

const output = stringify(filtered)

Database Bulk Insert

import { parse } from '@sylphx/molt-csv'

const csv = readFileSync('users.csv', 'utf-8')

const records = parse(csv, {
  columns: {
    id: Number,
    age: Number
  }
})

// Insert into database
for (const record of records) {
  await db.users.insert(record)
}

Combining Multiple CSVs

import { parse, stringify } from '@sylphx/molt-csv'

const csv1 = readFileSync('file1.csv', 'utf-8')
const csv2 = readFileSync('file2.csv', 'utf-8')

const records1 = parse(csv1)
const records2 = parse(csv2)

const combined = [...records1, ...records2]
const output = stringify(combined)

Advanced Features

WASM Acceleration

For very large files, WASM can help:

import { parse } from '@sylphx/molt-csv/wasm'

// Uses WebAssembly implementation
const records = parse(largeCSV)

Streaming Large Files

import { createReadStream } from 'fs'
import { createParseStream } from '@sylphx/molt-csv'

const stream = createReadStream('huge.csv')
const csvStream = createParseStream()

stream.pipe(csvStream)

csvStream.on('data', (record) => {
  console.log('Processing:', record)
})

Error Handling

import { parse } from '@sylphx/molt-csv'

try {
  const records = parse(csv, { strict: true })
} catch (error) {
  console.error('CSV parse error at line:', error.line)
  console.error('Message:', error.message)
}

Comparison with Alternatives

molt-csv vs papaparse

// molt-csv is 5.9x faster on simple CSVs
import { parse } from '@sylphx/molt-csv'

const records = parse(csv)

molt-csv vs csv-parse

// molt-csv is 7x faster and has better type support
import { parse } from '@sylphx/molt-csv'

const records = parse(csv, {
  columns: { id: Number }
})

Best Practices

1. Always specify column types - Enables automatic conversion
2. Use headers option - Makes data more readable
3. Handle errors gracefully - CSV can have inconsistencies
4. Consider streaming - For files larger than 10MB
5. Validate data after parsing - Use schema validation

Resources

• Quick Start Guide
• Installation Guide
• Benchmarks
• GitHub Repository

---

Next: Explore other packages or check the Benchmarks