This recipe migrates usage of the legacy buffer.atob() and buffer.btoa() APIs to the current recommended Buffer-based approaches for base64 encoding and decoding.
What It Does
The codemod transforms:
buffer.atob(data) → Buffer.from(data, 'base64').toString('binary')buffer.btoa(data) → Buffer.from(data, 'binary').toString('base64')
Usage
npx codemod nodejs/buffer-atob-btoa
Examples
Migrating buffer.atob()
Decoding base64-encoded data:
const buffer = require('node:buffer');
const data = 'SGVsbG8gV29ybGQh'; // "Hello World!" in base64
const decodedData = buffer.atob(data);
console.log(decodedData); // Outputs: Hello World!
Migrating buffer.btoa()
Encoding data to base64:
const buffer = require('node:buffer');
const data = 'Hello World!';
const encodedData = buffer.btoa(data);
console.log(encodedData); // Outputs: SGVsbG8gV29ybGQh
Why Migrate?
The buffer.atob() and buffer.btoa() methods were convenience wrappers around Buffer operations. Using Buffer methods directly:
- Provides more explicit control over encoding
- Aligns with modern Node.js best practices
- Reduces dependency on the
buffer module import
- Offers better performance in some scenarios
The Buffer API provides more encoding options beyond ‘base64’ and ‘binary’, including ‘hex’, ‘utf8’, ‘ascii’, and more.
References
