Add percent encoding function

[laniakea64]

#1982

This function encodes more characters than Javascript encodeURIComponent because the Examples section of that page mentions some cases that require additional percent-encoding, and in my experience extra percent-encoding like this doesn't hurt anything.

Not happy about the static in light of #2049 (comment) but don't know any way around it ☹️

[casey]

My issue with static is just that it requires navigating to the definition of the static to find out what it actually contains, not that it's static per se. I moved the static inside the function, so it's not an issue here.

Regarding which characters to actually percent encode, I think it might be useful to provide functions which exactly match encodeURI and encodeURIComponent, that way, users aren't surprised by which characters are encoded.

We could also then later add a function which percent encodes even more things, if that winds up being useful. I'd rather percent encode more things based on specific feedback and concrete use-cases.

What do you think?

[laniakea64]

Changed to match encodeURIComponent behavior. Per #1982 (comment) encodeURIComponent would fill the original request. And should also work for how I might use it.

I don't recall ever using encodeURI.

[casey]

LGTM! I renamed it to encode_uri_component, so that we have the ability to add a more flexible percent_encode function later on, perhaps one which takes as input a string containing a custom list of characters to escape.