R1CS parser continuation

[Vindaar]

This PR continues on from #327.

I tried to stick very closely to the binary layout of the files in the definition of the R1CS object. We could restructure it to be such that it has different types for different sections explicitly. But given the description of the file format, this is more honest.

Still have to:

• [x] add test cases
• [ ] decide if some types should be uint32/64 or int32/64
• [ ] change the type of prime: seq[byte] to something more appropriate. I assume I'll do the same as for the value of the constraints?

[Vindaar]

I've added the test case from https://github.com/iden3/r1csfile/blob/master/test/r1csfile.js using the file from here: https://github.com/iden3/r1csfile/tree/master/test/testutils