Rational in std deprecation

PauSolerValades · April 6, 2026, 9:32pm

Hello!
I wanted to implement a Rational number in order to mess around with some concepts, and I was surprised there used to be a Rational implementation on std in 0.14. I also see there is a comment regarding changing it to use it’s own allocator instead of using two Ints.
Okay, my questions are:

Why this got dropped? Any plans to bringing it back? (did some PR digging but did not find anything) Was it just due to suboptimal performance by storing two Ints?
Now that master is explicitly rejecting the managed variants, is BigInt going to suffer this same fate? Idk if this redesing also applies to the Int type.
If this were to be implemented again (might look into it) which would be the correct approach? I assume it’s not to have it’s own allocator for sure, but also not holding two BigInt again. Is maybe storing a Limb on there?

Thank you and have a nice day :))

squeek502 · April 6, 2026, 10:25pm

github.com/ziglang/zig

Absorb std.math.big.rational logic into std.math.big.int; fix `@intFromFloat` safety check (#24188)

master ← mlugg:intfromfloat-safety

opened 01:40PM - 15 Jun 25 UTC

mlugg

+1341 -1082

First, @jacobly0 rewrote a bunch of bigint logic for me, and that's... actually …most of the lines in this PR. Then is what it was meant to be; see below. This is breaking because it removes `std.math.big.rational`. --- This safety check was completely broken; it triggered unchecked illegal behavior *in order to implement the safety check*. You definitely can't do that! Instead, we must explicitly check the boundaries. This is a tiny bit fiddly, because we need to make sure we do floating-point rounding in the correct direction, and also handle the fact that the operation truncates so the boundary works differently for min vs max. Instead of implementing this safety check in Sema, there are now dedicated AIR instructions for safety-checked intfromfloat (two instructions; which one is used depends on the float mode). Currently, no backend directly implements them; instead, a `Legalize.Feature` is added which expands the safety check, and this feature is enabled for all backends we currently test, including the LLVM backend. The `u0` case is still handled in Sema, because Sema needs to check for that anyway due to the comptime-known result. The old safety check here was also completely broken and has therefore been rewritten. In that case, we just check for 'abs(input) < 1.0'. I've added a bunch of test coverage for the boundary cases of `@intFromFloat`, both for successes (in `test/behavior/cast.zig`) and failures (in `test/cases/safety/`). Resolves: #24161

PauSolerValades · April 6, 2026, 10:43pm

thanks m8! so it does not exist anymore because all internal logic that was rellevant to the compiler is into math.big.int?