Several hashcode implementations (including the one within the Java foundation classes) relies on integer overflows to have predictable behavior. This assumption prevents these hashcode implementations from being portable, since integer overflows result in different values across runtime environments. The following hashcode implementation does n…

HashCode.java

@Override
public int hashCode() {
    int hash = 0, buffer = 0;
    int modulo, logicalShiftLeft;

    for(int index = 0, max = this.stringProperty.length(); index < max; ++index) {
        modulo = index % 4;
        logicalShiftLeft = 8 * modulo;
        buffer += this.stringProperty.charAt(index) << logicalShiftLeft;
        if(modulo == 3) {
            hash ^= buffer;
            buffer = 0;
        }
    }
    return hash ^ buffer;
}

HashCode.js

function hashCode(value) {
    var hash = 0, buffer = 0;
    var modulo, logicalShiftLeft;

    for(var index = 0, max = value.length; index < max; ++index) {
        modulo = index % 4;
        logicalShiftLeft = 8 * modulo;
        buffer += value.charCodeAt(index) << logicalShiftLeft;
        if(modulo == 3) {
            hash ^= buffer;
            buffer = 0;
        }
    }
    return hash ^ buffer;
}