[test_url_query] Add failing test for str()

[behnam]

Add failing test for URL.__str__() on an instance with embedded URL
as query value.

[behnam]

This test demonstrates the main blocker we have on yarl. I was trying to get this out as port of #220, but since it's taking longer to land that, adding it here in a minimal diff.

This asymmetric behavior is different from the issue with the current API design for query value getter/setter, which returns/expects encoded values.

What do you think?

[codecov]

Codecov Report

Merging #233 into master will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master     #233   +/-   ##
=======================================
  Coverage   99.65%   99.65%           
=======================================
  Files           2        2           
  Lines         584      584           
  Branches      140      140           
=======================================
  Hits          582      582           
  Partials        2        2

---

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 84a9a2b...f646565. Read the comment docs.

[serhiy-storchaka]

It is easy to make this test passing. But is not it too strict? What is wrong with the current behavior? If preserve encoded ? in raw_query_string, which other encoded characters should be preserved? And in which other parts of URL? If preserve %3F, should %3f also be preserved?

The following test is passed.

assert URL(str(u)) == u

[mjpieters]

I'm not sure there is real value in this PR, and nothing has happened with it in nearly 2 years so I'm closing this. Feel free to recreate or request reopening if this is still something to pursue.