CtrlK
BlogDocsLog inGet started
Tessl Logo

tessl/npm-ieee754

Read/write IEEE754 floating point numbers from/to a Buffer or array-like object

—

Pending
Quality

Pending

Does it follow best practices?

Impact

Pending

No eval scenarios have been run

SecuritybySnyk

Pending

The risk profile of this skill

Overview
Eval results
Files

IEEE754

IEEE754 is a lightweight JavaScript library for reading and writing IEEE754 floating point numbers from/to Buffer objects or array-like structures. It implements the IEEE Standard for Floating-Point Arithmetic (IEEE 754) with precise control over floating-point number encoding and decoding, supporting different endianness, mantissa lengths, and byte counts.

Package Information

  • Package Name: ieee754
  • Package Type: npm
  • Language: JavaScript
  • Installation: npm install ieee754

Core Imports

const ieee754 = require('ieee754');

For ES modules:

import * as ieee754 from 'ieee754';

Basic Usage

const ieee754 = require('ieee754');

// Create a buffer to work with
const buffer = Buffer.alloc(4);

// Write a 32-bit float (single precision)
const value = 42.42;
ieee754.write(buffer, value, 0, true, 23, 4);

// Read the 32-bit float back
const readValue = ieee754.read(buffer, 0, true, 23, 4);
console.log(readValue); // 42.42000007629395

// Working with 64-bit doubles
const doubleBuffer = Buffer.alloc(8);
const doubleValue = 12345.123456789;

// Write a 64-bit double (double precision)
ieee754.write(doubleBuffer, doubleValue, 0, true, 52, 8);

// Read the 64-bit double back
const readDouble = ieee754.read(doubleBuffer, 0, true, 52, 8);
console.log(readDouble); // 12345.123456789

Architecture

IEEE754 implements low-level binary operations for floating-point number encoding and decoding:

  • Binary Representation: Converts floating-point numbers to/from their binary IEEE754 representation
  • Configurable Precision: Supports different mantissa lengths and byte counts for various precision levels
  • Endianness Control: Handles both little-endian and big-endian byte ordering
  • Standards Compliance: Follows IEEE 754 specification including special value handling (Infinity, NaN, denormalized numbers)

Capabilities

Reading IEEE754 Numbers

Reads an IEEE754 floating point number from a buffer or array-like object.

/**
 * Reads an IEEE754 floating point number from a buffer
 * @param {Buffer|Uint8Array|Array} buffer - Buffer or array-like object containing the data
 * @param {number} offset - Offset into the buffer to start reading
 * @param {boolean} isLE - Whether to use little-endian byte order
 * @param {number} mLen - Mantissa length in bits
 * @param {number} nBytes - Number of bytes to read
 * @returns {number} The parsed floating point value
 */
function read(buffer, offset, isLE, mLen, nBytes);

Usage Examples:

const ieee754 = require('ieee754');

// Read a 32-bit float (single precision)
const buffer32 = Buffer.from([0x42, 0x28, 0xF5, 0xC3]); // Little-endian representation
const float32 = ieee754.read(buffer32, 0, true, 23, 4);

// Read a 64-bit double (double precision)
const buffer64 = Buffer.alloc(8);
buffer64.writeDoubleLE(3.141592653589793, 0);
const float64 = ieee754.read(buffer64, 0, true, 52, 8);

// Read big-endian format
const bigEndianBuffer = Buffer.from([0xC3, 0xF5, 0x28, 0x42]);
const bigEndianFloat = ieee754.read(bigEndianBuffer, 0, false, 23, 4);

Writing IEEE754 Numbers

Writes an IEEE754 floating point number to a buffer or array-like object.

/**
 * Writes an IEEE754 floating point number to a buffer
 * @param {Buffer|Uint8Array|Array} buffer - Buffer or array-like object to write to
 * @param {number} value - The floating point value to write
 * @param {number} offset - Offset into the buffer to start writing
 * @param {boolean} isLE - Whether to use little-endian byte order
 * @param {number} mLen - Mantissa length in bits
 * @param {number} nBytes - Number of bytes to write
 * @returns {void} Modifies the buffer in-place
 */
function write(buffer, value, offset, isLE, mLen, nBytes);

Usage Examples:

const ieee754 = require('ieee754');

// Write a 32-bit float (single precision)
const buffer32 = Buffer.alloc(4);
ieee754.write(buffer32, 42.42, 0, true, 23, 4);

// Write a 64-bit double (double precision)  
const buffer64 = Buffer.alloc(8);
ieee754.write(buffer64, 3.141592653589793, 0, true, 52, 8);

// Write in big-endian format
const bigEndianBuffer = Buffer.alloc(4);
ieee754.write(bigEndianBuffer, 123.456, 0, false, 23, 4);

// Write special values
const specialBuffer = Buffer.alloc(4);
ieee754.write(specialBuffer, Infinity, 0, true, 23, 4);
ieee754.write(specialBuffer, -Infinity, 0, true, 23, 4);
ieee754.write(specialBuffer, NaN, 0, true, 23, 4);

Common Patterns

Standard IEEE754 Formats

const ieee754 = require('ieee754');

// 32-bit single precision float (IEEE 754 binary32)
// mLen: 23 (mantissa bits), nBytes: 4 (total bytes)
const buffer32 = Buffer.alloc(4);
ieee754.write(buffer32, 1.5, 0, true, 23, 4);
const value32 = ieee754.read(buffer32, 0, true, 23, 4);

// 64-bit double precision float (IEEE 754 binary64)  
// mLen: 52 (mantissa bits), nBytes: 8 (total bytes)
const buffer64 = Buffer.alloc(8);
ieee754.write(buffer64, 1.5, 0, true, 52, 8);
const value64 = ieee754.read(buffer64, 0, true, 52, 8);

Endianness Handling

const ieee754 = require('ieee754');
const buffer = Buffer.alloc(4);
const value = 42.0;

// Little-endian (most common on x86/x64)
ieee754.write(buffer, value, 0, true, 23, 4);
const littleEndian = ieee754.read(buffer, 0, true, 23, 4);

// Big-endian (network byte order)
ieee754.write(buffer, value, 0, false, 23, 4);
const bigEndian = ieee754.read(buffer, 0, false, 23, 4);

Working with Typed Arrays

const ieee754 = require('ieee754');

// Works with Uint8Array and other typed arrays
const uint8Array = new Uint8Array(4);
ieee754.write(uint8Array, 3.14159, 0, true, 23, 4);
const result = ieee754.read(uint8Array, 0, true, 23, 4);

Error Handling

The library handles IEEE754 special values correctly:

  • Infinity: Positive and negative infinity values are preserved
  • NaN: Not-a-Number values are handled appropriately
  • Zero: Both positive and negative zero are supported
  • Denormalized numbers: Subnormal values are handled according to IEEE754 specification
const ieee754 = require('ieee754');
const buffer = Buffer.alloc(4);

// Special values
ieee754.write(buffer, Infinity, 0, true, 23, 4);
console.log(ieee754.read(buffer, 0, true, 23, 4)); // Infinity

ieee754.write(buffer, -Infinity, 0, true, 23, 4);
console.log(ieee754.read(buffer, 0, true, 23, 4)); // -Infinity

ieee754.write(buffer, NaN, 0, true, 23, 4);
console.log(ieee754.read(buffer, 0, true, 23, 4)); // NaN
Workspace
tessl
Visibility
Public
Created
Last updated
Describes
npmpkg:npm/ieee754@1.2.x
Publish Source
CLI
Badge
tessl/npm-ieee754 badge