问题
I have a HashMap for storing objects:
private Map<T, U> fields = Collections.synchronizedMap(new HashMap<T, U>());
but, when trying to check existence of a key, containsKey method returns false.equals and hashCode methods are implemented, but the key is not found.
When debugging a piece of code:
return fields.containsKey(bean) && fields.get(bean).isChecked();
I have:
bean.hashCode() = 1979946475
fields.keySet().iterator().next().hashCode() = 1979946475
bean.equals(fields.keySet().iterator().next())= true
fields.keySet().iterator().next().equals(bean) = true
but
fields.containsKey(bean) = false
What could cause such strange behavioure?
public class Address extends DtoImpl<Long, Long> implements Serializable{
<fields>
<getters and setters>
@Override
public int hashCode() {
final int prime = 31;
int result = 1;
result = prime * result + StringUtils.trimToEmpty(street).hashCode();
result = prime * result + StringUtils.trimToEmpty(town).hashCode();
result = prime * result + StringUtils.trimToEmpty(code).hashCode();
result = prime * result + ((country == null) ? 0 : country.hashCode());
return result;
}
@Override
public boolean equals(Object obj) {
if (this == obj)
return true;
if (obj == null)
return false;
if (getClass() != obj.getClass())
return false;
Address other = (Address) obj;
if (!StringUtils.trimToEmpty(street).equals(StringUtils.trimToEmpty(other.getStreet())))
return false;
if (!StringUtils.trimToEmpty(town).equals(StringUtils.trimToEmpty(other.getTown())))
return false;
if (!StringUtils.trimToEmpty(code).equals(StringUtils.trimToEmpty(other.getCode())))
return false;
if (country == null) {
if (other.country != null)
return false;
} else if (!country.equals(other.country))
return false;
return true;
}
}
回答1:
You shall not modify the key after having inserted it in the map.
Edit : I found the extract of javadoc in Map :
Note: great care must be exercised if mutable objects are used as map keys. The behavior of a map is not specified if the value of an object is changed in a manner that affects equals comparisons while the object is a key in the map.
Example with a simple wrapper class:
public static class MyWrapper {
private int i;
public MyWrapper(int i) {
this.i = i;
}
public void setI(int i) {
this.i = i;
}
@Override
public boolean equals(Object o) {
if (this == o) return true;
if (o == null || getClass() != o.getClass()) return false;
return i == ((MyWrapper) o).i;
}
@Override
public int hashCode() {
return i;
}
}
and the test :
public static void main(String[] args) throws Exception {
Map<MyWrapper, String> map = new HashMap<MyWrapper, String>();
MyWrapper wrapper = new MyWrapper(1);
map.put(wrapper, "hello");
System.out.println(map.containsKey(wrapper));
wrapper.setI(2);
System.out.println(map.containsKey(wrapper));
}
Output :
true
false
Note : If you dont override hashcode() then you will get true only
回答2:
As Arnaud Denoyelle points out, modifying a key can have this effect. The reason is that containsKey cares about the key's bucket in the hash map, while the iterator doesn't. If the first key in your map --disregarding buckets -- just happens to be the one you want, then you can get the behavior you're seeing. If there's only one entry in the map, this is of course guaranteed.
Imagine a simple, two-bucket map:
[0: empty] [1: yourKeyValue]
The iterator goes like this:
- iterate over all of the elements in bucket 0: there are none
- iterate over all the elements in bucket 1: just the one
yourKeyValue
The containsKey method, however, goes like this:
keyToFindhas ahashCode() == 0, so let me look in bucket 0 (and only there). Oh, it's empty -- returnfalse.
In fact, even if the key stays in the same bucket, you'll still have this problem! If you look at the implementation of HashMap, you'll see that each key-value pair is stored along with the key's hash code. When the map wants to check the stored key against an incoming one, it uses both this hashCode and the key's equals:
((k = e.key) == key || (key != null && key.equals(k))))
This is a nice optimization, since it means that keys with different hashCodes that happen to collide into the same bucket will be seen as non-equal very cheaply (just an int comparison). But it also means that changing the key -- which will not change the stored e.key field -- will break the map.
回答3:
Debugging the java source code I realized that the method containsKey checks two things on the searched key against every element in the key set: hashCode and equals; and it does it in that order.
It means that if obj1.hashCode() != obj2.hashCode(), it returns false (without evaluating obj1.equals(obj2). But, if obj1.hashCode() == obj2.hashCode(), then it returns obj1.equals(obj2)
You have to be sure that both methods -may be you have to override them- evaluate to true for your defined criteria.
回答4:
Here is SSCCE for your issue bellow. It works like a charm and it couldn't be else, because your hashCode and equals methods seem to be autogenerated by IDE and they look fine.
So, the keyword is when debugging. Debug itself can harm your data. For example somewhere in debug window you set expression which changes your fields object or bean object. After that your other expressions will give you unexpected result.
Try to add all this checks inside your method from where you got return statement and print out their results.
import org.apache.commons.lang.StringUtils;
import java.io.Serializable;
import java.util.Collections;
import java.util.HashMap;
import java.util.Map;
public class Q21600344 {
public static void main(String[] args) {
MapClass<Address, Checkable> mapClass = new MapClass<>();
mapClass.put(new Address("a", "b", "c", "d"), new Checkable() {
@Override
public boolean isChecked() {
return true;
}
});
System.out.println(mapClass.isChecked(new Address("a", "b", "c", "d")));
}
}
interface Checkable {
boolean isChecked();
}
class MapClass<T, U extends Checkable> {
private Map<T, U> fields = Collections.synchronizedMap(new HashMap<T, U>());
public boolean isChecked(T bean) {
return fields.containsKey(bean) && fields.get(bean).isChecked();
}
public void put(T t, U u) {
fields.put(t, u);
}
}
class Address implements Serializable {
private String street;
private String town;
private String code;
private String country;
Address(String street, String town, String code, String country) {
this.street = street;
this.town = town;
this.code = code;
this.country = country;
}
String getStreet() {
return street;
}
String getTown() {
return town;
}
String getCode() {
return code;
}
String getCountry() {
return country;
}
@Override
public int hashCode() {
final int prime = 31;
int result = 1;
result = prime * result + StringUtils.trimToEmpty(street).hashCode();
result = prime * result + StringUtils.trimToEmpty(town).hashCode();
result = prime * result + StringUtils.trimToEmpty(code).hashCode();
result = prime * result + ((country == null) ? 0 : country.hashCode());
return result;
}
@Override
public boolean equals(Object obj) {
if (this == obj)
return true;
if (obj == null)
return false;
if (getClass() != obj.getClass())
return false;
Address other = (Address) obj;
if (!StringUtils.trimToEmpty(street).equals(StringUtils.trimToEmpty(other.getStreet())))
return false;
if (!StringUtils.trimToEmpty(town).equals(StringUtils.trimToEmpty(other.getTown())))
return false;
if (!StringUtils.trimToEmpty(code).equals(StringUtils.trimToEmpty(other.getCode())))
return false;
if (country == null) {
if (other.country != null)
return false;
} else if (!country.equals(other.country))
return false;
return true;
}
}
来源：https://stackoverflow.com/questions/21600344/java-hashmap-containskey-returns-false-for-existing-object