modified std.math.cast to support .ComptimeInt as well as .Int

math.zig

const std = @import("std");
const assert = std.debug.assert;
const maxInt = std.math.maxInt;
const minInt = std.math.minInt;

pub fn cast(comptime T: type, x: anytype) (error{Overflow}!T) {
    comptime assert(@typeInfo(T) == .Int); // must pass an integer
    comptime assert(@typeInfo(@TypeOf(x)) == .Int or
        @typeInfo(@TypeOf(x)) == .ComptimeInt); // must pass an integer or comptimeint
    if (x > maxInt(T)) {
        return error.Overflow;
    } else if (x < minInt(T)) {
        return error.Overflow;
    } else {
        return @intCast(T, x);
    }
}

const testing = std.testing;

test "cast" {
    { // .Int
        try testing.expectEqual(@as(i8, -128), try cast(i8, @as(i32, -128)));
        try testing.expectError(error.Overflow, cast(i8, @as(i32, -129)));
        try testing.expectError(error.Overflow, cast(u8, @as(u32, 300)));
        try testing.expectError(error.Overflow, cast(i8, @as(i32, -200)));
        try testing.expectError(error.Overflow, cast(u8, @as(i8, -1)));
        try testing.expectError(error.Overflow, cast(u64, @as(i8, -1)));

        try testing.expect((try cast(u8, @as(u32, 255))) == @as(u8, 255));
        try testing.expect(@TypeOf(try cast(u8, @as(u32, 255))) == u8);
    }
    { // .ComptimeInt
        try testing.expectEqual(@as(i8, -128), try cast(i8, -128));
        try testing.expectError(error.Overflow, cast(i8, -129));
        try testing.expectError(error.Overflow, cast(u8, 300));
        try testing.expectError(error.Overflow, cast(i8, -200));
        try testing.expectError(error.Overflow, cast(u8, -1));
        try testing.expectError(error.Overflow, cast(u64, -1));

        try testing.expect((try cast(u8, 255)) == @as(u8, 255));
        try testing.expect(@TypeOf(try cast(u8, 255)) == u8);
    }
}

maxInt(@TypeOf(x)) > maxInt(T) and minInt(@TypeOf(x)) < minInt(T) in the current std.math.cast
https://github.com/ziglang/zig/blob/3ee44ce949117e8e91348ef870b18b23571a408d/lib/std/math.zig#L1008-L1018
were added in the commit ziglang/zig@5b00dee
in order to handle signed integers.

It seems unnecessary since tests in the modified version of std.math.cast passes.