我对Java比较陌生,经常发现需要对值进行Map<Key,Value>排序。

由于这些值不是唯一的,我发现自己将keySet转换为一个数组,并使用自定义比较器通过数组排序对该数组进行排序,该比较器根据与该键关联的值进行排序。

有没有更简单的方法?


当前回答

发布我的答案版本

List<Map.Entry<String, Integer>> list = new ArrayList<>(map.entrySet());
    Collections.sort(list, (obj1, obj2) -> obj2.getValue().compareTo(obj1.getValue()));
    Map<String, Integer> resultMap = new LinkedHashMap<>();
    list.forEach(arg0 -> {
        resultMap.put(arg0.getKey(), arg0.getValue());
    });
    System.out.println(resultMap);

其他回答

当你有两个相等的项目时,投票给最多的答案不起作用。TreeMap保留相等的值。

示例:未排序地图

key/value: D/67.3
key/value: A/99.5
key/value: B/67.4
key/value: C/67.5
key/value: E/99.5

后果

key/value: A/99.5
key/value: C/67.5
key/value: B/67.4
key/value: D/67.3

所以省略了E!!

对我来说,它可以很好地调整比较器,如果它等于,则不返回0,而是返回-1。

在示例中:

类ValueComparator实现Comparator{地图库;公共ValueComparator(地图库){this.base=基数;}public int compare(对象a,对象b){如果((双)base.get(a)<(双)base.get(b)){返回1;}否则如果((双)base.get(a)==(双)base.get(b)){返回-1;}其他{返回-1;}}}

现在它返回:

未排序地图:

key/value: D/67.3
key/value: A/99.5
key/value: B/67.4
key/value: C/67.5
key/value: E/99.5

结果:

key/value: A/99.5
key/value: E/99.5
key/value: C/67.5
key/value: B/67.4
key/value: D/67.3

作为对《外国人》的回应(2011年11月22日):我将此解决方案用于整数Id和名称的映射,但想法是相同的,因此上面的代码可能不正确(我将在测试中编写并给您正确的代码),这是基于上面解决方案的map排序代码:

package nl.iamit.util;

import java.util.Comparator;
import java.util.Map;

public class Comparators {


    public static class MapIntegerStringComparator implements Comparator {

        Map<Integer, String> base;

        public MapIntegerStringComparator(Map<Integer, String> base) {
            this.base = base;
        }

        public int compare(Object a, Object b) {

            int compare = ((String) base.get(a))
                    .compareTo((String) base.get(b));
            if (compare == 0) {
                return -1;
            }
            return compare;
        }
    }


}

这是测试类(我刚刚测试了它,这适用于Integer,StringMap:

package test.nl.iamit.util;

import java.util.HashMap;
import java.util.TreeMap;
import nl.iamit.util.Comparators;
import org.junit.Test;
import static org.junit.Assert.assertArrayEquals;

public class TestComparators {


    @Test
    public void testMapIntegerStringComparator(){
        HashMap<Integer, String> unSoretedMap = new HashMap<Integer, String>();
        Comparators.MapIntegerStringComparator bvc = new Comparators.MapIntegerStringComparator(
                unSoretedMap);
        TreeMap<Integer, String> sorted_map = new TreeMap<Integer, String>(bvc);
        //the testdata:
        unSoretedMap.put(new Integer(1), "E");
        unSoretedMap.put(new Integer(2), "A");
        unSoretedMap.put(new Integer(3), "E");
        unSoretedMap.put(new Integer(4), "B");
        unSoretedMap.put(new Integer(5), "F");

        sorted_map.putAll(unSoretedMap);

        Object[] targetKeys={new Integer(2),new Integer(4),new Integer(3),new Integer(1),new Integer(5) };
        Object[] currecntKeys=sorted_map.keySet().toArray();

        assertArrayEquals(targetKeys,currecntKeys);
    }
}

以下是地图比较器的代码:

public static class MapStringDoubleComparator implements Comparator {

    Map<String, Double> base;

    public MapStringDoubleComparator(Map<String, Double> base) {
        this.base = base;
    }

    //note if you want decending in stead of ascending, turn around 1 and -1
    public int compare(Object a, Object b) {
        if ((Double) base.get(a) == (Double) base.get(b)) {
            return 0;
        } else if((Double) base.get(a) < (Double) base.get(b)) {
            return -1;
        }else{
            return 1;
        }
    }
}

这是一个测试用例:

@Test
public void testMapStringDoubleComparator(){
    HashMap<String, Double> unSoretedMap = new HashMap<String, Double>();
    Comparators.MapStringDoubleComparator bvc = new Comparators.MapStringDoubleComparator(
            unSoretedMap);
    TreeMap<String, Double> sorted_map = new TreeMap<String, Double>(bvc);
    //the testdata:
    unSoretedMap.put("D",new Double(67.3));
    unSoretedMap.put("A",new Double(99.5));
    unSoretedMap.put("B",new Double(67.4));
    unSoretedMap.put("C",new Double(67.5));
    unSoretedMap.put("E",new Double(99.5));

    sorted_map.putAll(unSoretedMap);

    Object[] targetKeys={"D","B","C","E","A"};
    Object[] currecntKeys=sorted_map.keySet().toArray();

    assertArrayEquals(targetKeys,currecntKeys);
}

当然,你可以让它更通用,但我只需要一个案例(地图)

迟到。

随着Java-8的出现,我们可以以非常简单/简洁的方式使用流进行数据操作。您可以使用流按值对映射条目进行排序,并创建一个LinkedHashMap,以保留插入顺序迭代。

Eg:

LinkedHashMap sortedByValueMap = map.entrySet().stream()
                .sorted(comparing(Entry<Key,Value>::getValue).thenComparing(Entry::getKey))     //first sorting by Value, then sorting by Key(entries with same value)
                .collect(LinkedHashMap::new,(map,entry) -> map.put(entry.getKey(),entry.getValue()),LinkedHashMap::putAll);

对于反向排序,请替换:

comparing(Entry<Key,Value>::getValue).thenComparing(Entry::getKey)

with

comparing(Entry<Key,Value>::getValue).thenComparing(Entry::getKey).reversed()
    Map<String, Integer> map = new HashMap<>();
    map.put("b", 2);
    map.put("a", 1);
    map.put("d", 4);
    map.put("c", 3);
    
    // ----- Using Java 7 -------------------
    List<Map.Entry<String, Integer>> entries = new ArrayList<>(map.entrySet());
    Collections.sort(entries, (o1, o2) -> o1.getValue().compareTo(o2.getValue()));
    System.out.println(entries); // [a=1, b=2, c=3, d=4]


    // ----- Using Java 8 Stream API --------
   map.entrySet().stream().sorted(Map.Entry.comparingByValue()).forEach(System.out::println); // {a=1, b=2, c=3, d=4}

    

这可以用java8非常容易地实现

public static LinkedHashMap<Integer, String> sortByValue(HashMap<Integer, String> map) {

        List<Map.Entry<Integer, String>> list = new ArrayList<>(map.entrySet());
        list.sort(Map.Entry.comparingByValue());
        LinkedHashMap<Integer, String> sortedMap = new LinkedHashMap<>();
        list.forEach(e -> sortedMap.put(e.getKey(), e.getValue()));
        return sortedMap;
    }

如果倾向于使用一个Map数据结构,该结构可以按值进行固有排序,而不必触发任何排序方法或显式传递给实用程序,则以下解决方案可能适用:

(1) org.rools.chance.core.util.ValueSortedMap(JBoss项目)在内部维护两个映射,一个用于查找,另一个用于维护排序值。与之前添加的答案非常相似,但可能是抽象和封装部分(包括复制机制)使其更安全地从外部使用。

(2) http://techblog.molindo.at/2008/11/java-map-sorted-by-value.html避免维护两个映射,而是依赖/扩展Apache Common的LinkedMap。(博客作者注:这里的所有代码都在公共领域):

// required to access LinkEntry.before and LinkEntry.after
package org.apache.commons.collections.map;

// SNIP: imports

/**
* map implementation based on LinkedMap that maintains a sorted list of
* values for iteration
*/
public class ValueSortedHashMap extends LinkedMap {
    private final boolean _asc;

    // don't use super()!
    public ValueSortedHashMap(final boolean asc) {
        super(DEFAULT_CAPACITY);
        _asc = asc;
    }

    // SNIP: some more constructors with initial capacity and the like

    protected void addEntry(final HashEntry entry, final int hashIndex) {
        final LinkEntry link = (LinkEntry) entry;
        insertSorted(link);
        data[hashIndex] = entry;
    }

    protected void updateEntry(final HashEntry entry, final Object newValue) {
        entry.setValue(newValue);
        final LinkEntry link = (LinkEntry) entry;
        link.before.after = link.after;
        link.after.before = link.before;
        link.after = link.before = null;
        insertSorted(link);
    }

    private void insertSorted(final LinkEntry link) {
        LinkEntry cur = header;
        // iterate whole list, could (should?) be replaced with quicksearch
        // start at end to optimize speed for in-order insertions
        while ((cur = cur.before) != header & amp; & amp; !insertAfter(cur, link)) {}
        link.after = cur.after;
        link.before = cur;
        cur.after.before = link;
        cur.after = link;
    }

    protected boolean insertAfter(final LinkEntry cur, final LinkEntry link) {
        if (_asc) {
            return ((Comparable) cur.getValue())
            .compareTo((V) link.getValue()) & lt; = 0;
        } else {
            return ((Comparable) cur.getValue())
            .compareTo((V) link.getValue()) & gt; = 0;
        }
    }

    public boolean isAscending() {
        return _asc;
    }
}

(3) 编写一个自定义映射或从LinkedHashMap扩展,该映射仅在枚举期间根据需要进行排序(例如,values()、keyset()、entryset())。内部实现/行为是从使用该类的实现/行为中抽象出来的,但在该类的客户端看来,当请求枚举时,值总是被排序的。如果所有的put操作都在枚举之前完成,这个类希望排序只发生一次。排序方法采用了前面对这个问题的一些回答。

public class SortByValueMap<K, V> implements Map<K, V> {

    private boolean isSortingNeeded = false;

    private final Map<K, V> map = new LinkedHashMap<>();

    @Override
    public V put(K key, V value) {
        isSortingNeeded = true;
        return map.put(key, value);
    }

    @Override
    public void putAll(Map<? extends K, ? extends V> map) {
        isSortingNeeded = true;
        map.putAll(map);
    }

    @Override
    public Set<K> keySet() {
        sort();
        return map.keySet();
    }

    @Override
    public Set<Entry<K, V>> entrySet() {
        sort();
        return map.entrySet();
    }

    @Override
    public Collection<V> values() {
        sort();
        return map.values();
    }

    private void sort() {
        if (!isSortingNeeded) {
            return;
        }

        List<Entry<K, V>> list = new ArrayList<>(size());

        for (Iterator<Map.Entry<K, V>> it = map.entrySet().iterator(); it.hasNext();) {
            Map.Entry<K, V> entry = it.next();
            list.add(entry);
            it.remove();
        }

        Collections.sort(list);

        for (Entry<K, V> entry : list) {
            map.put(entry.getKey(), entry.getValue());
        }

        isSortingNeeded = false;
    }

    @Override
    public String toString() {
        sort();
        return map.toString();
    }
}

(4) Guava提供了ImmutableMap.Builder.orderEntriesByValue(Comparator valueComparator),尽管生成的映射是不可变的:

将此生成器配置为根据指定的比较器。排序顺序是稳定的,也就是说,如果两个条目的值作为等价项进行比较,首先插入的条目将是第一个按照构建映射的迭代顺序。