comparison differentiates absent and empty components

[alandefreitas]

This PR simplifies the URL comparison function. It also fixes a bug where the difference between some absent and empty components was not considered. This fix was part of #425, which is now obsoleted.

[cppalliance-bot]

GCOVR code coverage report https://645.url.prtest.cppalliance.org/gcovr/index.html
LCOV code coverage report https://645.url.prtest.cppalliance.org/genhtml/index.html
Coverage Diff https://645.url.prtest.cppalliance.org/gcovr/coverage_diff.txt

[codecov]

Codecov Report

Merging #645 (3d783b6) into develop (cbcfa6a) will decrease coverage by 0.09%.
The diff coverage is 100.00%.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop     #645      +/-   ##
===========================================
- Coverage    96.48%   96.39%   -0.10%     
===========================================
  Files          138      138              
  Lines         6746     6789      +43     
===========================================
+ Hits          6509     6544      +35     
- Misses         237      245       +8     

Impacted Files	Coverage Δ
include/boost/url/authority_view.hpp	93.93% <ø> (ø)
include/boost/url/impl/authority_view.ipp	85.09% <100.00%> (+4.93%)	⬆️
include/boost/url/impl/url_view_base.ipp	96.68% <100.00%> (-1.28%)	⬇️
include/boost/url/detail/impl/normalize.ipp	90.84% <0.00%> (-1.24%)	⬇️

---

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 cbcfa6a...3d783b6. Read the comment docs.

[vinniefalco]

Seems mostly ok except for style nits but what is the motivation for this?

[alandefreitas]

Seems mostly ok except for style nits

Great. I'll fix that.

but what is the motivation for this?

The main motivation is fixing the bug. The previous function would compare an absent component and an empty component as the same thing, but that's wrong.

BOOST_TEST_EQ(url("https:"), url("https://"));
BOOST_TEST_EQ(url("https://@www.boost.org"), url("https://www.boost.org"));
BOOST_TEST_EQ(url("https:"), url("https:?"));
BOOST_TEST_EQ(url("https:"), url("https:#"));
// ... etc

should fail.

I'll improve coverage with a few more tests and the difference is going to be more explicit.

[vinniefalco]

"simplify URL comparison" doesnt' really tell me that there's a bug in the current release of Boost.URL

[alandefreitas]

"simplify URL comparison" doesnt' really tell me that there's a bug in the current release of Boost.URL

Yes. I chose that because it's not the only thing happening in the PR and because of the trend of "refactor xx" being a synonym for "fix xx".

[cppalliance-bot]

GCOVR code coverage report https://645.url.prtest.cppalliance.org/gcovr/index.html
LCOV code coverage report https://645.url.prtest.cppalliance.org/genhtml/index.html
Coverage Diff https://645.url.prtest.cppalliance.org/gcovr/coverage_diff.txt