Keep large integer input out of floating point
A worked example preserves a large decimal integer through binary and hexadecimal representations.
Keep large integer input out of floating point
When an input is an integer identifier or exact counter, parsing it through a floating-point Number can change its value before any formatting starts. The Number safe-integer limit is a useful boundary to test.
Preserve the original decimal text
const raw = "9007199254740993";
const integer = BigInt(raw);
console.log(integer.toString(10)); // 9007199254740993
console.log(Number(raw).toString()); // 9007199254740992
console.log(BigInt("0x" + integer.toString(16)) === integer); // true
The comparison demonstrates two distinct contracts: Number is useful for general numerical calculations, while BigInt represents integers without the same fixed safe-integer bound. Do not insert a Number conversion between the input string and BigInt.
Validate the input language
A converter should say which digits and prefixes it accepts. A decimal-only input can be checked against a signed decimal pattern before parsing. Decimal fractions are not integers; hexadecimal letters are not decimal digits. Returning an error is clearer than silently keeping a valid prefix and discarding the rest.
BigInt values also need an explicit serialization choice. For public JSON examples, use a decimal string and document that choice rather than assuming the default JSON serializer accepts BigInt.
Reference: BigInt.