Sign of result of "remainder"

[jakobkummerow]

The proposal currently reads:

1.1.6 BigInt::remainder (x, y)
If y is 0n, throw a RangeError exception.
Return the BigInt representing x modulo y.

where "modulo" links to https://tc39.github.io/ecma262/#sec-algorithm-conventions (side note: that should be https://tc39.github.io/ecma262/#sec-mathematical-operations instead), which says (emphasis mine):

The notation “x modulo y” (y must be finite and nonzero) computes a value k of the same sign as y (or zero) such that abs(k) < abs(y) and x-k = q × y for some integer q.

This is notably different from the % operator, which says:

The sign of the result equals the sign of the dividend.

I would assume that % on BigInts should mirror the behavior for integer-valued Numbers. Having differing behavior would be quite surprising, I think:

 7  % -4  ===  3
 7n % -4n === -1n
-7  %  4  === -3
-7n %  4n ===  1n

(I don't think this is a bug in the existing spec; modulo is mostly used with positive powers of two as the right-hand side there, where ToUint behavior is desired.)

[littledan]

Yes, this should have semantics similar to the % operator on Numbers with respect to the sign. Thanks for pointing out the mismatch; I will fix this soon.

[lars-t-hansen]

Because of the matter of the sign, Scheme distinguishes between "remainder" and "modulo" and provides standard procedures for both: http://www.schemers.org/Documents/Standards/R5RS/HTML/r5rs-Z-H-2.html#%_toc_%_sec_6.2.5. I suggest you do the same.

[littledan]

@lars-t-hansen We discussed this in #37 . It seems that % is neither modulo nor remainder, if I understand correctly. Name suggestions are welcome!

[lars-t-hansen]

It might be useful to provide correct (in the number-theoretic sense) remainder and modulo operations, then, since BigInt is an integer type. (Presumably as static methods.)

[littledan]

@lars-t-hansen Seems not so hard for a user to implement out of % and a few conditionals, right? Would it be OK to leave this for v2?