Buffer.byteLength benchmark does not test base64 encoding correctly

[kisugez]

Version

Output of node -v: (e.g. v20.10.0)

Platform

Output of uname -a (for UNIX) or Windows version string (for Windows).

Subsystem

buffers (benchmark/buffers/buffer-bytelength-string.js)

What steps will reproduce the bug?

Run the buffer byteLength benchmark for base64 encoding.
Expected: getInput should return a base64-encoded string when encoding is 'base64'.
Actual: getInput returns the original string, not the base64-encoded string.

How often does it reproduce? Is there a required condition?

Run the buffer byteLength benchmark for base64 encoding.
Expected: getInput should return a base64-encoded string when encoding is 'base64'.
Actual: getInput returns the original string, not the base64-encoded string.

What is the expected behavior? Why is that the expected behavior?

getInput should return the base64-encoded string so the benchmark accurately tests Buffer.byteLength for base64 input.

What do you see instead?

The benchmark tests the original string, not the base64-encoded string.

Additional information

Suggested fix:
In benchmark/buffers/buffer-bytelength-string.js, update getInput to return Buffer.from(original, 'utf8').toString('base64') when encoding is 'base64'.
This will ensure the benchmark is correct.