if two objects hashcode is a same, why doesnot mean that o1.equals(o2)?

Question

The implementation of equals() and hashCode() should follow these rules.

If o1.equals(o2), then o1.hashCode() == o2.hashCode() should always be true.
If o1.hashCode() == o2.hashCode() is true, it doesn’t mean that o1.equals(o2) will be true.

If o1.hashCode() == o2.hashCode() is true . i know, that If two objects are equals then these two objects should return same hash code. if two objects hash code is a same, why does it not mean that o1.equals(o2)?

For the same reason that all people with the name "Rezo" are not you. — Kayaman, Dec 11 '19 at 14:43
The correct statement is "not all people with the name Rezo are you". We have empirical evidence that at least of them is. ;-) — , Dec 12 '19 at 03:48

score 6 · Accepted Answer · answered Dec 11 '19 at 14:44

6

Consider Long. There are 2^64 possible values for this type. hashCode returns an int, which only has 2^32 possible values.

For identity hash code (availble from System.identityHashCode), objects move around in memory on many modern JVM implementations. There is no reasonable way of keeping track of which hash codes are still in use. Even with a (thread-safe) counter, after 2^32 allocations there will need to be some kind of reuse.

answered Dec 11 '19 at 14:44

Tom Hawtin - tackline

139,906
30
206
293

hashcode not unique? If the hashcode of the o1 and o2 objects is equal, what does this mean? – Rezo Joglidze Dec 11 '19 at 14:49
Can two objects have the same hashcode? – Rezo Joglidze Dec 11 '19 at 14:49
Actually all the objects could have the same hash code and it would be a legal hash function. Not very practical though. – Andres Dec 11 '19 at 14:50
@Andres There is/was an option to build HotSpot such that every identity hash code is 1 - for debugging purposes. – Tom Hawtin - tackline Dec 11 '19 at 14:52

Andres · Answer 2 · 2019-12-11T14:54:49.903

0

The purpose of a hash function (the clue is in the name) is to return values distributed uniformly in a range for different objects. Not to return a different value for each object.

This means a hash function normally has collisions. Though ideally their number should be as small as possible.

edited Dec 11 '19 at 14:54

answered Dec 11 '19 at 14:47

Andres

9,722
4
37
57

WJS · Answer 3 · 2019-12-12T02:44:04.013

0

Imagine a hash algorithm (and not necessarily a good one) as the following.

Take the last 4 digits our some assigned ID.
Imagine two ID's of 12341234 and 67281234.

The hashCodes are identical but the id's and hence what they identify may not be.

edited Dec 12 '19 at 02:44

answered Dec 11 '19 at 14:51

WJS

22,083
3
14
32

score 0 · Answer 4 · answered Dec 11 '19 at 14:51

It's a bit like:

If John and James are twins, then they must belong to the same mother

Which (I'm sure you agree) does not mean that

If John and James belong to the same mother, then they must be twins

Where "having the same mother" is having the same hash code, and "being twins" is being equal.

For the rest, I suggest you read through the first answer of this question What is the use of hashCode in Java?

if two objects hashcode is a same, why doesnot mean that o1.equals(o2)?

4 Answers4