Collection of small issues about armake HOT 5 OPEN

https://github.com/KoffeinFlummi/armake/blob/master/src/p3d.c#L1259 This is explosion shielding from v72
it actually belongs below invArmor and above special lods struct according to mikero

from armake.

https://github.com/KoffeinFlummi/armake/blob/master/src/p3d.c#L842
Second tileU should be tileV

from armake.

There is currently one level of "sectionsInherit"
https://github.com/KoffeinFlummi/armake/blob/master/src/model_config.c#L443
It should be recursive. If the inherited class has another sectionsInherit entry we need to follow that too.

from armake.

Parser error
https://github.com/KoffeinFlummi/armake/blob/master/src/rapify.l#L117
This rejects the "unquoted string" case if there happens to be a number at the front.
x = 0.0447031 * safezoneW + safezoneX;
Oh look a number at the front! That means this can't possibly be a unquoted string!
Not sure what the workaround is.. Maybe make sure that there aren't any stray non-number characters left after the number that was detected?

Yep that worked for me

    auto res = std::from_chars(trimmedString.data(), trimmedString.data() + trimmedString.size(), val);
    if (res.ec != std::errc::invalid_argument) {//It's a number. They don't have to be quoted

        auto end = trimmedString.data() + trimmedString.size();

        //Basically only whitespace is allowed after that, if we have any stray characters, then this is not just a number but infact a unquoted string
        auto isEndClean = std::all_of(res.ptr, end, [](char x)
            {
                return x == '\n' || x == '\r' || x == ' ' || x == '\t';
            });
       
        if (isEndClean)//If end is not clean, this is not JUST a number.
            REJECT;
    }

from armake.

https://github.com/KoffeinFlummi/armake/blob/master/src/keygen.c#L71
This can fail if the hex number would start with a 0. Like
0x08BF7BD...
The 0 at the start is ommited, meaning you end up with a string length that is not power of two.
len will be rounded down instead of up, resulting in 255 length when you expect 256, which will cut one whole byte off. Ending up with a 127 byte signature here: https://github.com/KoffeinFlummi/armake/blob/master/src/sign.c#L345

Here is a pbo where that happens with v2 signatures

core.zip

from armake.