ok, i looked at some of the code. i see that you have basically written the elliptic curve calculates on your own. are you sure they are correct? is it tested?
i didn't understand two things:
1. why are you doing this check? r + s doesn't have to be dividable by 2. and i can't find it anywhere that says this check is needed. do you know any reason i am missing? see 4.1.3 of SEC1
also i believe this mod is completely unnecessary as the numbers you are giving it are 256-bit and mod with that larger number will return the same thing. not to mention that where it is being called the numbers (r and s for instance) must be 32 byte otherwise the implementation would be broken.
2. the other thing i didn't get is why are you using the private key itself for generating a random number for signing. isn't there any better RNG provider to use?
might i also suggest Separation of concerns. some files are too large (as in 1000+ lines) and it seems like they don't all need to be in one place.
nitpicking but DER Format comment should really move to line 252 because that is where you are DER formatting
i didn't understand two things:
1. why are you doing this check? r + s doesn't have to be dividable by 2. and i can't find it anywhere that says this check is needed. do you know any reason i am missing? see 4.1.3 of SEC1
also i believe this mod is completely unnecessary as the numbers you are giving it are 256-bit and mod with that larger number will return the same thing. not to mention that where it is being called the numbers (r and s for instance) must be 32 byte otherwise the implementation would be broken.
2. the other thing i didn't get is why are you using the private key itself for generating a random number for signing. isn't there any better RNG provider to use?
might i also suggest Separation of concerns. some files are too large (as in 1000+ lines) and it seems like they don't all need to be in one place.
nitpicking but DER Format comment should really move to line 252 because that is where you are DER formatting