Improve performance, fix bug in authorization schema

[Uzlopak]

While programming I realized we have a bug in the code. We did not check if between Bearerand key was atleast one space. This bug is fixed, see unit tests.

before:
Missing Header:

autocannon http://127.0.0.1:8000/foo
0 2xx responses, 271167 non 2xx responses
271k requests in 10.01s, 59.9 MB read

Invalid BearerType:

autocannon -H authorization='Beaver key' http://127.0.0.1:8000/foo
0 2xx responses, 295334 non 2xx responses
295k requests in 11.01s, 65.3 MB read

Invalid key:

autocannon -H authorization='Bearer invalid' http://127.0.0.1:8000/foo
0 2xx responses, 274488 non 2xx responses
274k requests in 11.01s, 60.7 MB read

Valid Request:

autocannon -H authorization='Bearer key' http://127.0.0.1:8000/foo
272k requests in 11.01s, 52.6 MB read

after:
Missing Header:

autocannon http://127.0.0.1:8000/foo
0 2xx responses, 300072 non 2xx responses
300k requests in 11.01s, 66.3 MB read

Invalid BearerType:

autocannon -H authorization='Beaver key' http://127.0.0.1:8000/foo
0 2xx responses, 291809 non 2xx responses
292k requests in 11.01s, 64.5 MB read

Invalid key:

autocannon -H authorization='Bearer invalid' http://127.0.0.1:8000/foo
0 2xx responses, 326559 non 2xx responses
327k requests in 11.01s, 72.2 MB read

Valid Request:

autocannon -H authorization='Bearer key' http://127.0.0.1:8000/foo
404k requests in 11.01s, 77.9 MB read

Checklist

• run npm run test and npm run benchmark
• tests and/or benchmarks are included
• documentation is changed or added
• commit message and code follows the Developer's Certification of Origin
  and the Code of conduct

[mcollina]

lgtm

[gurgunday]

Looks great!

[olivierchatry]

I'm quite sure that BearerType should be case insensitive.

[Uzlopak]

@olivierchatry

My PR did not change any behaviour regarding case sensitivity. Probably you wanted to comment it on PR #167 by @dancastillo , who changed bearer to Bearer.

Eitherway, maybe open an issue to discuss it.

[jsumners]

I am quite certain the spec requires Bearer https://www.rfc-editor.org/rfc/rfc6750#section-2.1

[olivierchatry]

lexik/LexikJWTAuthenticationBundle#411

I think this is just a string comparaison, and there will most probably be bearer and Bearer in the wilde. This is really just string comparison, so I'm not sure why not do it.

[Uzlopak]

@olivierchatry

Please open an issue and discuss this there. Please also refer also to RFCs. This PR did not change that behaviour as I already explained.