perf: optimizing functions

[H4ad]

Before:

isUint8Array x 30,520,055 ops/sec ±1.32% (93 runs sampled)
uint8ArrayToString x 102 ops/sec ±0.41% (76 runs sampled)
stringToUint8Array x 165 ops/sec ±0.16% (85 runs sampled)
uint8ArrayToBase64: 
stringToBase64: 

The values are empty because it only throws errors.

Now:

isUint8Array x 160,822,491 ops/sec ±1.10% (100 runs sampled)
uint8ArrayToString x 103 ops/sec ±0.19% (76 runs sampled)
stringToUint8Array x 164 ops/sec ±0.21% (84 runs sampled)
uint8ArrayToBase64 x 16.44 ops/sec ±6.55% (37 runs sampled)
stringToBase64 x 8.37 ops/sec ±5.35% (25 runs sampled)

Full Execution:

isUint8Array x 160,822,491 ops/sec ±1.10% (100 runs sampled)
concatUint8Arrays with 2 arrays x 3,125 ops/sec ±7.13% (71 runs sampled)
concatUint8Arrays with 3 arrays x 3,699 ops/sec ±3.41% (82 runs sampled)
concatUint8Arrays with 4 arrays x 1,974 ops/sec ±2.87% (84 runs sampled)
uint8ArrayToString x 103 ops/sec ±0.19% (76 runs sampled)
stringToUint8Array x 164 ops/sec ±0.21% (84 runs sampled)
uint8ArrayToBase64 x 16.44 ops/sec ±6.55% (37 runs sampled)
stringToBase64 x 8.37 ops/sec ±5.35% (25 runs sampled)
base64ToString x 4.30 ops/sec ±0.60% (15 runs sampled)
uint8ArrayToHex x 20.14 ops/sec ±5.41% (38 runs sampled)
hexToUint8Array x 38.27 ops/sec ±0.52% (67 runs sampled)

I added benchmark for all the functions but I didn't necessarily optimize all of them.

The changes were mostly to address the issue with larger arrays and some of the optimizations will only be noticeable on tiny arrays (uint8ArrayToString and stringToUint8Array).

[sindresorhus]

I think you switched the benchmark results, as otherwise, everything got slower.

[sindresorhus]

Thank you :)