Start tuning cmath functions

[ckormanyos]

Now that most <cmath> functions are present, the purpose of this PR is to tune them, check/extend ranges and try to hit more non-covered lines.

[codecov]

Codecov Report

Attention: Patch coverage is 98.07692% with 1 lines in your changes are missing coverage. Please review.

Project coverage is 99.0%. Comparing base (fa54eb5) to head (266715c).

Additional details and impacted files

@@            Coverage Diff            @@
##           develop    #580     +/-   ##
=========================================
+ Coverage     98.9%   99.0%   +0.1%     
=========================================
  Files          200     200             
  Lines        12912   12940     +28     
  Branches      1487    1492      +5     
=========================================
+ Hits         12766   12799     +33     
+ Misses         146     141      -5     

Files	Coverage Δ
include/boost/decimal/detail/cmath/exp.hpp	100.0% <100.0%> (+4.6%)	⬆️
include/boost/decimal/detail/cmath/lgamma.hpp	100.0% <100.0%> (ø)
include/boost/decimal/detail/cmath/tgamma.hpp	100.0% <100.0%> (ø)
test/test_big_uints.cpp	100.0% <100.0%> (ø)
test/test_exp.cpp	100.0% <100.0%> (ø)
test/test_lgamma.cpp	100.0% <100.0%> (ø)
test/test_log.cpp	100.0% <100.0%> (ø)
test/test_tgamma.cpp	100.0% <100.0%> (ø)
include/boost/decimal/detail/emulated128.hpp	97.7% <83.4%> (+1.2%)	⬆️

... and 1 file with indirect coverage changes

---

Continue to review full report in Codecov by Sentry.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update fa54eb5...266715c. Read the comment docs.

[ckormanyos]

Hi Matt (@mborland) I'm gonna merge this one lickety split. I've got a goal to get to $99\%$ coverage soon but not quite on this one.

Based on experiences from other projects, it takes about a few months and another thousand lines of test code to reach $100\%$, which I'd like to put on the table as a reachable goal. Let's party.

[mborland]

Hi Matt (@mborland) I'm gonna merge this one lickety split. I've got a goal to get to $99\%$ coverage soon but not quite on this one.

Based on experiences from other projects, it takes about a few months and another thousand lines of test code to reach $100\%$, which I'd like to put on the table as a reachable goal. Let's party.

Sounds good to me. For the new fast types I've been recycling the same test code so hopefully it doesn't drop the coverage too much.

[ckormanyos]

For the new fast types I've been recycling the same test code so hopefully it doesn't drop the coverage too much.

It's worth the effort. Those fast types will be key for clients requiring fast math on decimal types. This was really a good direction to introduce into this work.

Trivial coverage, ... we will get as we evolve the LIB. This thing is moving.

[ckormanyos]

Well I made a few efforts to get better on funcs and also got a few lines covered on 128/256 bit backends. I did not quite finish everything on the (non-published) TODO list. But it's headed in the right direction. So when this thing cycles green, I'll just go ahead and push it.

The battery of funcions/impl-details is really getting rather stable nowdays. Good stuff.