Prefix custom structures with _ to hide them

[tamasgal]

Since the custom stuff is mostly for test coverage, it makes sense to hide it from the public API. Btw. it also collides with downstream code since the structs are usually defined there and otherwise there is a clash, see #95 (comment)

[codecov]

Codecov Report

Merging #104 (b79dc41) into master (703079e) will not change coverage.
The diff coverage is 70.00%.

@@           Coverage Diff           @@
##           master     #104   +/-   ##
=======================================
  Coverage   88.19%   88.19%           
=======================================
  Files          11       11           
  Lines        1330     1330           
=======================================
  Hits         1173     1173           
  Misses        157      157           

Impacted Files	Coverage Δ
src/custom.jl	72.72% <70.00%> (ø)

---

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 703079e...b79dc41. Read the comment docs.

[Moelf]

I thought we already have LorentzVector for test coverage
https://codecov.io/gh/tamasgal/UnROOT.jl/src/master/src/custom.jl

so actually, if you don't need them to be inside UnROOT, feel free to delete them haha. Which will increase our coverage %!

[tamasgal]

I'd rather add more tests 😆 I promise I'll do 😉

[Moelf]

I mean they could be in the test for sure, I'm saying if you don't need experiment-specific structs live inside UnROOT.jl. I'm just unsure how much you've factored out and how much you still rely on what's inside UnROOT.jl for actual work, so we didn't touch it.

[tamasgal]

I think I have factored out everything, so they are really only there for tests. I just realised there is this DAQHit where I experimented with an ultra fast in-place-interpretation by defining it as a primite type and hacking Base.getproperty. That's a huge miss in test coverage but I did not want to forget about it for future implementations of dynamic type and parser generation. I'll probably just remove it for now and create a gist or so...