Compile list of test cases for #define

[LordGolias]

I created a new branch with a working implementation of analysis of #define statements, including the 4 common cases

#define A
#define A (...)
#define A a,b,...
#define A(a,b,...) (a + b * ...)

I pushed this to a branch because I am not so familiar with the exact syntax of #define, and this may raise lots of errors if not done correctly.

I would like to ask for help in compiling a list of cases where the parser is expected to break, and cases which it does not. The goal is to, before moving to the ACE/ACRE/Exile code, have a solid list of simpler examples to test from.

[kymckay]

Some helpful pages:
https://community.bistudio.com/wiki/PreProcessor_Commands
https://community.bistudio.com/wiki/PreProcessor_Errors

Also (the VBS wiki - a surprisingly useful resource for arma):
https://resources.bisimulations.com/wiki/PreProcessor_Commands#.23define

The VBS wiki is probably more helpful as it's actually maintained by the guys who make VBS and know what they're talking about (in theory).

[kymckay]

I'm not even sure if it's possible to break #define, as far as I know it'll just take all of the remainder of the line no matter the contents. With the \ at EOL carrying the definition over to the next line.

I'm not sure what happens if you do:

#define A \
#define B

or any combination of preprocessor commands within a definition. It's probably good to avoid though.

Oh and a quote from the VBS wiki:

The defined name is case sensitive, alphanumeric, and must start with a letter or underscore

[LordGolias]

I compiled and addressed a list of #define test cases. I am sure it is not bullet-proof against everything, but it should cover the relevant use-cases.

I have not included the last one you mentioned because it seems a bit too edgy. Once we know what we should expect, and if it is worth caring about, we can address it.

Thanks a lot for the list of resources!

The relevant commit that addresses this is e0195cd.