Fix #428

[Hackerpilot]

This fixes a few logic problems in the handling of the trueStyle and falseStyle fields. The worst of these is that the old code called advance to skip over a token before then checking what the token was by looking at the new current token. It also had logic to set the field to DeclarationListStyle.single inside of an if that checked that there was a { or :, which makes no sense.

[codecov]

Codecov Report

Merging #429 into master will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##           master     #429   +/-   ##
=======================================
  Coverage   81.85%   81.86%           
=======================================
  Files          11       11           
  Lines        8443     8446    +3     
=======================================
+ Hits         6911     6914    +3     
  Misses       1532     1532           

Impacted Files	Coverage Δ
src/dparse/astprinter.d	94.77% <100.00%> (ø)
src/dparse/parser.d	91.20% <100.00%> (+<0.01%)	⬆️

---

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 36a3c4b...d34f57c. Read the comment docs.

[WebFreak001]

want to add tests or merge as is?

[Hackerpilot]

I think that our options are

1. Merge as-is
2. Or
   a. merge Add an XPath-based testing system for the AST generated by the parser #425 first
   b. change the XML output for the node to actually say what kind of compile condition it is, so that the XPATH can check it
   c. write a test

[Hackerpilot]

I merged the AST testing system and now this has a regression test.