Add option to provide current time when decoding JWT

[ZipFile]

• Makes possible to fake a time without using freezegun:freeze_time or unittest.mock:patch. Less monkey patching == better code.
• Makes possible to reuse current time fetched elsewhere (e.g. from request).

[codecov]

Codecov Report

Merging #267 (7a66496) into master (be8e914) will not change coverage.
The diff coverage is 100.00%.

❗ Current head 7a66496 differs from pull request most recent head 9667b78. Consider uploading reports for the commit 9667b78 to get more accurate results

@@           Coverage Diff           @@
##           master     #267   +/-   ##
=======================================
  Coverage   92.94%   92.94%           
=======================================
  Files          15       15           
  Lines        1418     1418           
=======================================
  Hits         1318     1318           
  Misses        100      100           

Impacted Files	Coverage Δ
jose/jwt.py	98.01% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update be8e914...9667b78. Read the comment docs.

[blag]

Dependency injection to ease testing? 👍
Dependency injection to improve usability? 👍
Code churn for no discernible reason? 👎

If you have a good reason for prepending the now parameter like you did, I'll definitely consider it, but I haven't seen a good reason for it yet. Please point me to it if I missed it. 😅

[blag]

Why change the function signature? Wouldn't it be easier to append now=None to the end of the parameter list and keep the function signature the same? And then use

now = now or datetime.utcnow()

to default to datetime.utcnow(), just in case anybody is using these private methods.

I realize that they're private methods, and people are signing up to keep their code up-to-date when they use these private methods, but there's no point in code churn without a good reason.

[ZipFile]

It was not part of the public API (this particular function does not even have docstring to serve as a documentation), so there was no hesitation to change the signature. Making now (required) positional argument better states the intention to me.

Either way, updated.

[jojolebarjos]

Hi folks, this feature would be very welcome, any chance this gets accepted?

[ZipFile]

Rebased with latest master, just in case.