Advanced preprocessor features

[Redar13]

• Allows the use of the operators ==, !=, <=, <, >= and >.
• Fixed an error of taking variables 0 and false as true.
• Fixed an error with empty expression body.
• Added SemVer support.
• Added an option to call an error.

[NeeEoo]

Version check uses string comparisons instead of semver comparison, makes it have higher risk of false matches

[Redar13]

Oh, I guess then I'd have to redo the method of getting variables in dynamic instead of string and do a version of the conversion via semver, only then you might have to duplicate the converter in hscript if the user doesn't have the semver library, but I don't know.

[NeeEoo]

You can include the semver in the code directly just credit the original source

also i recommend checking the haxe source code because it assumes its a version iirc

[TBar09]

This is peak.

[TBar09]

Okay so I tried out the pull request, and overall, everything works fairly well! preprocessors that
compare versions and such or the usage of different operators works as intended.

Though there is a bug that i've noticed when trying to do the basic method of checking, similar to
this:

...this doesn't trace anything,but the target is a c++ target. While doing
this:


...does trace something. It's a minor bug, but a fix I found for this is doing #if(cpp == '1'), but I'm
pretty sure it's just a matter of changing the if statement in the preprocessor condition functions.

(Also a little kerfuffle that I found while testing was that doing something like this:

gives an Unexpected token: "#end" error.)

Sorry for the giant text blurb, just thought it was nice to give a good test and give any results I found.

[NeeEoo]

Fyi you need the defines macro, you might be missing it, check codename for that

[Redar13]

Okay so I tried out the pull request, and overall, everything works fairly well! preprocessors that compare versions and such or the usage of different operators works as intended.

Though there is a bug that i've noticed when trying to do the basic method of checking, similar to this: ...this doesn't trace anything,but the target is a c++ target. While doing this: ...does trace something. It's a minor bug, but a fix I found for this is doing #if(cpp == '1'), but I'm pretty sure it's just a matter of changing the if statement in the preprocessor condition functions.

(Also a little kerfuffle that I found while testing was that doing something like this: gives an Unexpected token: "#end" error.)

Sorry for the giant text blurb, just thought it was nice to give a good test and give any results I found.

Thanks for testing the pull request.
As for the #if (cpp == "1") trick, it looks like the definitions are saved as strings, there are no strings in the parser check yet.

[TBar09]

Fyi you need the defines macro, you might be missing it, check codename for that

I used that in this test build. I'm pretty sure that the reason why it was not being ran was because of the different types (the defines macro usually puts stuff as strings, while the parser checks for both booleans and integers. I think doing something like this should fix the issue: