This repo aims to show some strange results that occur when trying to generate a baseline.
The problem was first noticed on some AngularJS code.
It seems that the problems occur due to missing comments and differences in whitespace.
The source in the lib directory should be the same apart from comment and whitespace differences.
The lib directory is instrumented by nyc and put in the instrumented directory.
The command nyc --nycrc-path=<nycrc> /usr/bin/true 1> <report_file>.txt is run over both the raw and instrumented code.
I'm assuming that we expect to see the same coverage output at this point, however the two sets of baseline coverage differ at this point.
To generate the instrumented code and baseline coverage reports run npm i then npm t, but have a look at package.json to see what it's doing first.