Support API_URLs that include path

[smudge]

/domain @samandmoore @effron
/platform @samandmoore @effron

Will annotate below.

Build is currently failing intermittently due to a test order thing. Will debug that but wanted to get the code up in the meantime.

[nanda-prbot]

Needs somebody from @samandmoore and @effron to claim domain review
Needs somebody from @samandmoore and @effron to claim platform review

Use the shovel operator to claim, e.g.:

@myname << domain && platform

_{HOW TO: Claim a Review}

[smudge]

Not sure if we're parsing URLs all over the place and could consolidate onto the config. Thoughts?

[samandmoore]

yea, it might be nice to have the config expose a path_prefix method instead of doing this here. it won't be used anywhere else, but it would be tidy.

[smudge]

Minor API change but it's an internal class right?

[smudge]

This was required to prevent global state from mucking up the test run and making this test flake:
https://github.com/Betterment/webvalve/pull/41/files#diff-4f60a691d33035e7b6c85487ba4937efR55

I was surprised that we leave existing WebMock stubs around even when we clear out the local registry of stubbed URLs. But maybe this shouldn't go here? I could put it in that one test's before hook instead? 😅

[samandmoore]

seems fine. i think reset is only used internally.

[samandmoore]

<<domain platform tafn

[nanda-prbot]

@smudge needs to incorporate feedback from @samandmoore. Bump when done.

_{HOW TO: Resolve Feedback}

[smudge]

bump @samandmoore (and @jmileham)

I got cute and wrote this test (and it passed) but then I deleted it and changed the ^ to \A:

      it 'is immune to newlines that are injected in the path' do
        with_env 'DUMMY_API_URL' => 'http://dummy.dev/gg' do
          WebValve.register subject.name
          WebValve.setup

          expect { Net::HTTP.get(URI('http://dummy.dev/gg/banana/%0Agg/widgets')) }
            .to raise_error(RuntimeError, /route not defined for GET/)
          expect { Net::HTTP.get(URI("http://dummy.dev/gg/banana/\ngg/widgets")) }
            .to raise_error(URI::InvalidURIError, /bad URI\(is not URI\?\)/)
        end
      end

[nanda-prbot]

Needs @samandmoore to provide platform review
Needs @samandmoore to provide domain review

When you finish a round of review, be sure to say you've finished or sign off on the PR, e.g.:

TAFN or DomainLGTM

If you're too busy to review, unclaim the PR, e.g.:

@myname >> domain

_{HOW TO: Give Feedback}

[jmileham]

i assume you deleted the test because you finally proved to yourself once and for all that it doesn't make sense to use an operator that means "match the front of the current line," in a construct that doesn't support multiple lines

[samandmoore]

just that one last suggestion

then domainlgtm platformlgtm

i think we can cut this as the official 1.0.0 release :)

[nanda-prbot]

Approved! 👏 🙌 🌮

[smudge]

I’m sort of testing URI::parse but I wanted a loose demonstration of basic URL/path expectations in case the underlying parsing implementation ever changes.

[samandmoore]

released! https://rubygems.org/gems/webvalve/versions/1.0.0