https://www.npmjs.com/package/byte-base64 - Encode JS Uint8Array, simple array of bytes or native JS string to base64 and back

base64.js

/*
MIT License

Copyright (c) 2020 Egor Nepomnyaschih

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
*/

/*
// This constant can also be computed with the following algorithm:
const base64abc = [],
	A = "A".charCodeAt(0),
	a = "a".charCodeAt(0),
	n = "0".charCodeAt(0);
for (let i = 0; i < 26; ++i) {
	base64abc.push(String.fromCharCode(A + i));
}
for (let i = 0; i < 26; ++i) {
	base64abc.push(String.fromCharCode(a + i));
}
for (let i = 0; i < 10; ++i) {
	base64abc.push(String.fromCharCode(n + i));
}
base64abc.push("+");
base64abc.push("/");
*/
const base64abc = [
	"A", "B", "C", "D", "E", "F", "G", "H", "I", "J", "K", "L", "M",
	"N", "O", "P", "Q", "R", "S", "T", "U", "V", "W", "X", "Y", "Z",
	"a", "b", "c", "d", "e", "f", "g", "h", "i", "j", "k", "l", "m",
	"n", "o", "p", "q", "r", "s", "t", "u", "v", "w", "x", "y", "z",
	"0", "1", "2", "3", "4", "5", "6", "7", "8", "9", "+", "/"
];

/*
// This constant can also be computed with the following algorithm:
const l = 256, base64codes = new Uint8Array(l);
for (let i = 0; i < l; ++i) {
	base64codes[i] = 255; // invalid character
}
base64abc.forEach((char, index) => {
	base64codes[char.charCodeAt(0)] = index;
});
base64codes["=".charCodeAt(0)] = 0; // ignored anyway, so we just need to prevent an error
*/
const base64codes = [
	255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255,
	255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255,
	255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 62, 255, 255, 255, 63,
	52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 255, 255, 255, 0, 255, 255,
	255, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14,
	15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 255, 255, 255, 255, 255,
	255, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40,
	41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51
];

function getBase64Code(charCode) {
	if (charCode >= base64codes.length) {
		throw new Error("Unable to parse base64 string.");
	}
	const code = base64codes[charCode];
	if (code === 255) {
		throw new Error("Unable to parse base64 string.");
	}
	return code;
}

export function bytesToBase64(bytes) {
	let result = '', i, l = bytes.length;
	for (i = 2; i < l; i += 3) {
		result += base64abc[bytes[i - 2] >> 2];
		result += base64abc[((bytes[i - 2] & 0x03) << 4) | (bytes[i - 1] >> 4)];
		result += base64abc[((bytes[i - 1] & 0x0F) << 2) | (bytes[i] >> 6)];
		result += base64abc[bytes[i] & 0x3F];
	}
	if (i === l + 1) { // 1 octet yet to write
		result += base64abc[bytes[i - 2] >> 2];
		result += base64abc[(bytes[i - 2] & 0x03) << 4];
		result += "==";
	}
	if (i === l) { // 2 octets yet to write
		result += base64abc[bytes[i - 2] >> 2];
		result += base64abc[((bytes[i - 2] & 0x03) << 4) | (bytes[i - 1] >> 4)];
		result += base64abc[(bytes[i - 1] & 0x0F) << 2];
		result += "=";
	}
	return result;
}

export function base64ToBytes(str) {
	if (str.length % 4 !== 0) {
		throw new Error("Unable to parse base64 string.");
	}
	const index = str.indexOf("=");
	if (index !== -1 && index < str.length - 2) {
		throw new Error("Unable to parse base64 string.");
	}
	let missingOctets = str.endsWith("==") ? 2 : str.endsWith("=") ? 1 : 0,
		n = str.length,
		result = new Uint8Array(3 * (n / 4)),
		buffer;
	for (let i = 0, j = 0; i < n; i += 4, j += 3) {
		buffer =
			getBase64Code(str.charCodeAt(i)) << 18 |
			getBase64Code(str.charCodeAt(i + 1)) << 12 |
			getBase64Code(str.charCodeAt(i + 2)) << 6 |
			getBase64Code(str.charCodeAt(i + 3));
		result[j] = buffer >> 16;
		result[j + 1] = (buffer >> 8) & 0xFF;
		result[j + 2] = buffer & 0xFF;
	}
	return result.subarray(0, result.length - missingOctets);
}

export function base64encode(str, encoder = new TextEncoder()) {
	return bytesToBase64(encoder.encode(str));
}

export function base64decode(str, decoder = new TextDecoder()) {
	return decoder.decode(base64ToBytes(str));
}

@vijtheveg I agree that the name btoa is far from intuitive. I would never expect a function with "to ASCII" in its name to return a string. ASCII is an encoding for converting text into bytes, so I would expect a "to ASCII" function to return bytes, not a string. Of course, once you realize that "ASCII" in this context really means "Base64", it makes sense why it returns a string, since Base64 is meant to encode binary data as text, but naming it btoa (binary to ASCII) is definitely confusing.

However, stoa is an even less intuitive name for what it actually does. Typically when we think of a string, we think of something that holds text. So, if I have a function essentially called "string to ASCII", I would assume that it would take a string as text that contains only ASCII supported characters and encode it into ASCII, returning an array of bytes. Again, I would not expect it to encode anything to Base64 based on its name.

Even if btoa was renamed to "string to Base64", it would still be confusing considering what it really does. I would expect it to take a string as text and encode it with a scheme like UTF-8 and then encode that as Base64. Which is exactly what this gist's base64encode does.

What I'm getting at is that the B in btoa is what tips you off to the fact that the string it takes as input is not text. It's a "binary string", meaning that it must only contain characters with code units in the range of 0-255. Probably the best name for it would be binaryStringToBase64.

---

I also agree that this gist's base64encode is not a good name for what it does. In my opinion, it should be called textToBase64 since that's its purpose, and its documentation should mention that it first encodes the text as UTF-8 before encoding it to Base64.