Add generate_fingerprint_bytes

[AllSeeingEyeTolledEweSew]

Closes #5985

[arvidn]

this doesn't look right. what happened to the A?

[AllSeeingEyeTolledEweSew]

The original test was wrong. The behavior of generate_fingerprint_bytes() is analogous to generate_fingerprint() as expected.

The tests I wrote with @unittest.skip() were not all correct (because they wouldn't run, so I couldn't verify the desired condition would be correct until the fix was made).

I expect that here:

• I wrote and validated the test for fingerprint() (str(fingerprint("A", 1, 2, 3, 4)) == "-A\x001234-")
• I copied that output into some test cases for generate_fingerprint()
• I realized we should have generate_fingerprint_bytes()
• I split the functions into test_generate() (with @unittest.skip()) and test_generate_str()
• I validated test_generate_str() (generate_fingerprint("A", 1, 2, 3, 4) == "---1234-", unlike fingerprint())
• I forgot to analogously update test_generate()

[arvidn]

I expect the "A" that's passed it would be used in the final fingerprint. It looks like a bug that it turns into -- instead. I can take a closer look at what's going on

[arvidn]

I see, there's a check and the name has to be two bytes long