deprecate mp_tc_(and|or|xor) in favor of mp_(and|or|xor)

[minad]

deprecate mp_tc_(and|or|xor) in favor of mp_(and|or|xor)

• same behavior for positive numbers
• generalisation for negative numbers, treating them as two complement
• improve algorithm, iterate once over the digits, manually perform two complement
• simplify mp_add_d, mp_sub_d
• functions are safe in case of a==c or b==c
• renamed mp_tc_div_2d to mp_signed_rsh (signed right shift)

[czurnieden]

(I would have used the old names but that's more a matter of taste, so…)
Please update the documentation, too.
Otherwise too many complains/feature requests of the form "And where is/can you implement mp_xor?" will come in.
The mp_complement has a different naming scheme, rename it to mp_tc_complement?

[minad]

@czurnieden we could also deprecate mp_tc_x in favor of mp_x with changed behavior for negative mp_ints. I think that would be ok since no one uses mp_x on negative numbers.
I would also prefer that to be honest.

[minad]

@czurnieden I changed the PR here - all bitwise functions are called mp_bit_x now. I think the names are much nicer like this. Also mp_tc_div_2d didn't make much sense as a name. It is a sign extending right shift. Furthermore I updated the docs. Do you agree with these changes?

[sjaeckel]

I also ask myself if there's even one user who depends on the old functions' behavior of negative numbers... I can't imagine as that was pretty useless, so I'd say we should use the old names!

[minad]

Ok good! I asked myself the same. I will change everything to the old names then. No unnecessary bit in the name. And I must say tc is also pretty ugly and I am a bit sorry for introducing it...

[minad]

@sjaeckel @nijtmans updated to old names