Tests: Automatically normalize line ends

[RunDevelopment]

Also discussed here, The parser for .test files will now automatically convert all file contents to CRLF line ends. This ensures that all languages are tested with the more difficult line ends.

This lead to the discovery (and fix) of a bug in False.

This PR contains 4 things:

• The parser now converting line ends.
• .test files no longer being declared as binary in .gitattributes.
• The fix for False.
• The changes for all changed test files. Some test files used LF line ends which were reflected in the JSON.

This PR does not convert the line ends of existing .test files to LF. This means that we now have a bunch of CRLF files in the repo. Should I convert them?

@mAAdhaTTah @joshgoebel

[github-actions]

JS File Size Changes (gzipped)

A total of 1 files have changed, with a combined diff of +6 B (+2.3%).

file	master	pull	size diff	% diff
components/prism-false.min.js	260 B	266 B	+6 B	+2.3%

Generated by 🚫 dangerJS against 6f540bf

[joshgoebel]

This means that we now have a bunch of CRLF files in the repo. Should I convert them?

I'm personally in favor of unix line endings - whichever one that is - and being consistent seems appropriate.

[RunDevelopment]

Advice for reviewers: Don't open the "Files" tab on Github, your tab will freeze.

[RunDevelopment]

Ok, no. I reverted that. I did a small test and these EOL changes will conflict will literally every PR that changes a test file.