I am building a factorization program and I would like to change each BigInt type to regular Numbers when number <= Number.MAX_SAFE_INTEGER.
Instead of coding two functions for each case, it would be nice if I could keep it all into one function which could vary variables types accordingly (something like let myVar = 3n || 3 I guess).
function Factorize(dividend) {
let divisor = 2n;
//if number <= Number.MAX_SAFE_INTEGER then let divisor = 2. Same for all other bigInts.
let method1 = [], method2 = [];
while (dividend > 1n) {
if (dividend % divisor === 0n) {
method1.push(`${divisor}`);
method2.push(`${dividend} / ${divisor}`);
dividend /= divisor;
} else {
divisor++
};
};
return {
default: method1,
detailed: method2,
get isPrime() {
return this.default.length === 1 && this.default[0] !== 2;
}
};
};
const number = parseInt(prompt());
console.log(Factorize(BigInt(number)));Thanks for your help.
What's the difficulty? Your comment contains half the required code already:
if (dividend <= Number.MAX_SAFE_INTEGER) {
divisor = 2;
dividend = Number(dividend);
}
And then you only need to replace the two strict equality comparisons === 0n and !== 2 with their non-strict variants. 0 == 0n returns true, 0n === 0n returns false.
Some other things worth mentioning:
(1) This method for factorization is extremely slow. There are prime numbers well below Number.MAX_SAFE_INTEGER for which this will take months. Depending on your use case, limiting the input size or implementing some sort of timeout (e.g., returning an error if a certain number of iterations wasn't enough to find the complete result) may be more important than supporting BigInts at all. (For inputs that have only small prime factors, even extremely huge inputs will still terminate quickly, so it's certainly possible to exceed Number range (even Number.MAX_VALUE) while still only taking a few milliseconds.)
(2) Using parseInt to get your input means that you're limiting yourself to Number precision; converting that Number to a BigInt afterwards doesn't bring the lost bits back. For example, if someone enters '12157665459056928801' (which is 3n ** 40n), parseInt will truncate that and your program will hence compute the wrong result. To avoid that, use the fact that the BigInt() constructor can convert strings directly, i.e.: BigInt(prompt()).
(3) While it's sometimes possible to write code that can work on both Numbers and BigInts, doing so is generally not recommended (and often not even useful), because the two types of values (intentionally!) behave differently in a number of ways (otherwise we wouldn't need both of them), so there is a large risk of such code not doing what you think it'll do. In this particular case it should be okay; I'm just advising not to generalize from this example.