Comparing Character, Integer and similar types in Java: Use equals or ==?

Question

I wanted to make sure about something in Java: If I have a Character or an Integer or a Long and those sort of things, should I use equals or is == sufficient?

I know that with strings there are no guarantees that there is only one instance of each unique string, but I'm not sure about other boxed types.

My intuition is to use equals, but I want to make sure I'm not wasting performance.

Answer 1

Java Language Spec 5.1.7:

If the value p being boxed is true, false, a byte, a char in the range \u0000 to \u007f, or an int or short number between -128 and 127, then let r1 and r2 be the results of any two boxing conversions of p. It is always the case that r1 == r2.

and:

Discussion

Ideally, boxing a given primitive value p, would always yield an identical reference. In practice, this may not be feasible using existing implementation techniques. The rules above are a pragmatic compromise. The final clause above requires that certain common values always be boxed into indistinguishable objects. The implementation may cache these, lazily or eagerly.

For other values, this formulation disallows any assumptions about the identity of the boxed values on the programmer's part. This would allow (but not require) sharing of some or all of these references.

This ensures that in most common cases, the behavior will be the desired one, without imposing an undue performance penalty, especially on small devices. Less memory-limited implementations might, for example, cache all characters and shorts, as well as integers and longs in the range of -32K - +32K.

So, in some cases == will work, in many others it will not. Always use .equals to be safe since you cannot grantee (generally) how the instances were obtained.

If speed is a factor (most .equals start with an == comparison, or at least they should) AND you can gurantee how they were allocated AND they fit in the above ranges then == is safe.

Some VMs may increase that size, but it is safer to assume the smallest size as specified by the langauge spec than to rely on a particular VM behaviour, unless you really really really need to.