String conversions

[phlptp]

Fix for Issue #420.

The issue appears to be that we were checking for is_constructible<std::string, T> from a type, but then using a implicit conversion. So that code path was changed to use is_convertible. A separate path was used for cases with !is_convertible and is_constructible

A test of the situation from #420 was created to verify.

@henryiii you might want to check if this resolves #410 as well.

[codecov]

Codecov Report

Merging #421 into master will not change coverage by %.
The diff coverage is n/a.

@@            Coverage Diff            @@
##            master      #421   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files           12        12           
  Lines         3580      3582    +2     
=========================================
+ Hits          3580      3582    +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 6b7f6a7...98397da. Read the comment docs.

[henryiii]

Thanks! Doesn't seem to solve my problem, though:

https://github.com/GooFit/GooFit/pull/231/checks?check_run_id=419249036

[phlptp]

The errors looked different enough that I wasn't sure. Tstring is complex enough that it could be a number of different things. What I am thinking about doing is adding another test file that is for testing the compatibility of different types with add_option. App test is starting to get to be a bit to big to manage.

[henryiii]

I don't think it's just TString (since a small test shows that TString can be added), but it's something that interacts with TString and the subclassing/new namespace.