如何以一种聪明的方式将Java.util.Map写入parcel?

时间:2011-11-24 09:20:43

标签: android bundle parcelable

我有一个字符串的通用映射(键,值),这个字段是我需要可以分配的Bean的一部分。 所以,我可以使用Parcel#writeMap方法。 API Doc说:

  

请改用writeBundle(Bundle)。将地图展平为包裹   在当前的dataPosition()中,如果需要,增加dataCapacity()。该   映射键必须是String对象。 Map值使用   writeValue(Object)并且必须遵循那里的规范。它是   强烈建议使用writeBundle(Bundle)代替此   方法,因为Bundle类提供了允许的类型安全API   你可以避免在编组时出现神秘的类型错误。

所以,我可以迭代我的Map中的每个Entry并将它放入Bundle中,但我仍然在寻找一种更聪明的方法。我缺少Android SDK中的任何方法吗?

目前我这样做:

final Bundle bundle = new Bundle();
final Iterator<Entry<String, String>> iter = links.entrySet().iterator();
while(iter.hasNext())
{
    final Entry<String, String>  entry =iter.next();
    bundle.putString(entry.getKey(), entry.getValue());
}
parcel.writeBundle(bundle);

7 个答案:

答案 0 :(得分:81)

我最后做的有点不同了。它遵循您期望处理Parcelables的模式,因此它应该是熟悉的。

public void writeToParcel(Parcel out, int flags){
  out.writeInt(map.size());
  for(Map.Entry<String,String> entry : map.entrySet()){
    out.writeString(entry.getKey());
    out.writeString(entry.getValue());
  }
}

private MyParcelable(Parcel in){
  //initialize your map before
  int size = in.readInt();
  for(int i = 0; i < size; i++){
    String key = in.readString();
    String value = in.readString();
    map.put(key,value);
  }
}

在我的应用程序中,地图中键的顺序很重要。我使用LinkedHashMap来保留排序,这样做可以保证密钥在从Parcel中提取后以相同的顺序出现。

答案 1 :(得分:25)

你可以尝试:

bundle.putSerializable(yourSerializableMap);

如果您选择的地图实现了可序列化(如HashMap),那么您可以轻松使用writeBundle

答案 2 :(得分:23)

如果地图的keyvalue都延伸Parcelable,那么您可以使用非常漂亮的Generics解决方案:

代码

// For writing to a Parcel
public <K extends Parcelable,V extends Parcelable> void writeParcelableMap(
        Parcel parcel, int flags, Map<K, V > map)
{
    parcel.writeInt(map.size());
    for(Map.Entry<K, V> e : map.entrySet()){
        parcel.writeParcelable(e.getKey(), flags);
        parcel.writeParcelable(e.getValue(), flags);
    }
}

// For reading from a Parcel
public <K extends Parcelable,V extends Parcelable> Map<K,V> readParcelableMap(
        Parcel parcel, Class<K> kClass, Class<V> vClass)
{
    int size = parcel.readInt();
    Map<K, V> map = new HashMap<K, V>(size);
    for(int i = 0; i < size; i++){
        map.put(kClass.cast(parcel.readParcelable(kClass.getClassLoader())),
                vClass.cast(parcel.readParcelable(vClass.getClassLoader())));
    }
    return map;
}

用法

// MyClass1 and MyClass2 must extend Parcelable
Map<MyClass1, MyClass2> map;

// Writing to a parcel
writeParcelableMap(parcel, flags, map);

// Reading from a parcel
map = readParcelableMap(parcel, MyClass1.class, MyClass2.class);

答案 3 :(得分:12)

好问题。除了putSerializable和writeMap之外,我所知道的API中没有任何方法。出于性能原因,序列化为not recommended,并且出于某种神秘的原因,不推荐使用writeMap(),正如您已经指出的那样。

我今天需要打包一个HashMap,所以我尝试编写一些实用方法,以推荐的方式将Map分配到Bundle和从Bundle中分配Map:

// Usage:

// read map into a HashMap<String,Foo>
links = readMap(parcel, Foo.class);

// another way that lets you use a different Map implementation
links = new SuperDooperMap<String, Foo>;
readMap(links, parcel, Foo.class);

// write map out
writeMap(links, parcel);

////////////////////////////////////////////////////////////////////
// Parcel methods

/**
 * Reads a Map from a Parcel that was stored using a String array and a Bundle.
 *
 * @param in   the Parcel to retrieve the map from
 * @param type the class used for the value objects in the map, equivalent to V.class before type erasure
 * @return     a map containing the items retrieved from the parcel
 */
public static <V extends Parcelable> Map<String,V> readMap(Parcel in, Class<? extends V> type) {

    Map<String,V> map = new HashMap<String,V>();
    if(in != null) {
        String[] keys = in.createStringArray();
        Bundle bundle = in.readBundle(type.getClassLoader());
        for(String key : keys)
            map.put(key, type.cast(bundle.getParcelable(key)));
    }
    return map;
}


/**
 * Reads into an existing Map from a Parcel that was stored using a String array and a Bundle.
 *
 * @param map  the Map<String,V> that will receive the items from the parcel
 * @param in   the Parcel to retrieve the map from
 * @param type the class used for the value objects in the map, equivalent to V.class before type erasure
 */
public static <V extends Parcelable> void readMap(Map<String,V> map, Parcel in, Class<V> type) {

    if(map != null) {
        map.clear();
        if(in != null) {
            String[] keys = in.createStringArray();
            Bundle bundle = in.readBundle(type.getClassLoader());
            for(String key : keys)
                map.put(key, type.cast(bundle.getParcelable(key)));
        }
    }
}


/**
 * Writes a Map to a Parcel using a String array and a Bundle.
 *
 * @param map the Map<String,V> to store in the parcel
 * @param out the Parcel to store the map in
 */
public static void writeMap(Map<String,? extends Parcelable> map, Parcel out) {

    if(map != null && map.size() > 0) {
        /*
        Set<String> keySet = map.keySet();
        Bundle b = new Bundle();
        for(String key : keySet)
            b.putParcelable(key, map.get(key));
        String[] array = keySet.toArray(new String[keySet.size()]);
        out.writeStringArray(array);
        out.writeBundle(b);
        /*/
        // alternative using an entrySet, keeping output data format the same
        // (if you don't need to preserve the data format, you might prefer to just write the key-value pairs directly to the parcel)
        Bundle bundle = new Bundle();
        for(Map.Entry<String, ? extends Parcelable> entry : map.entrySet()) {
            bundle.putParcelable(entry.getKey(), entry.getValue());
        }

        final Set<String> keySet = map.keySet();
        final String[] array = keySet.toArray(new String[keySet.size()]);
        out.writeStringArray(array);
        out.writeBundle(bundle);
        /**/
    }
    else {
        //String[] array = Collections.<String>emptySet().toArray(new String[0]);
        // you can use a static instance of String[0] here instead
        out.writeStringArray(new String[0]);
        out.writeBundle(Bundle.EMPTY);
    }
}

编辑:修改了writeMap以使用entrySet,同时保留与原始答案相同的数据格式(显示在切换注释的另一侧)。如果您不需要或不想保持读兼容性,那么在每次迭代中存储键值对可能更简单,如@bcorso和@Anthony Naddeo的答案。

答案 4 :(得分:1)

如果你的地图键是String,你可以使用它在javadocs中提到的Bundle:

/**
 * Please use {@link #writeBundle} instead.  Flattens a Map into the parcel
 * at the current dataPosition(),
 * growing dataCapacity() if needed.  The Map keys must be String objects.
 * The Map values are written using {@link #writeValue} and must follow
 * the specification there.
 *
 * <p>It is strongly recommended to use {@link #writeBundle} instead of
 * this method, since the Bundle class provides a type-safe API that
 * allows you to avoid mysterious type errors at the point of marshalling.
 */
public final void writeMap(Map val) {
    writeMapInternal((Map<String, Object>) val);
}

所以我写了下面的代码:

private void writeMapAsBundle(Parcel dest, Map<String, Serializable> map) {
    Bundle bundle = new Bundle();
    for (Map.Entry<String, Serializable> entry : map.entrySet()) {
        bundle.putSerializable(entry.getKey(), entry.getValue());
    }
    dest.writeBundle(bundle);
}

private void readMapFromBundle(Parcel in, Map<String, Serializable> map, ClassLoader keyClassLoader) {
    Bundle bundle = in.readBundle(keyClassLoader);
    for (String key : bundle.keySet()) {
        map.put(key, bundle.getSerializable(key));
    }
}

因此,您可以使用Parcelable而不是Serializable

答案 5 :(得分:0)

这是我的一些简单,但到目前为止,我在Kotlin实施。如果它不满足一个需求,它可以很容易地修改

但请不要忘记K,V如果与通常Parcelable等不同,则必须String, Int,...

parcel.writeMap(map)

parcel.readMap(map)

读取重叠

fun<K,V> Parcel.readMap(map: MutableMap<K,V>) : MutableMap<K,V>{

    val tempMap = LinkedHashMap<Any?,Any?>()
    readMap(tempMap, map.javaClass.classLoader)

    tempMap.forEach {
        map[it.key as K] = it.value as V
    }
    /* It populates and returns the map as well
       (useful for constructor parameters inits)*/
    return map
}

答案 6 :(得分:0)

这里提到的所有解决方案都是有效的,但没有一个解决方案具有足够的通用性。通常,您有包含字符串,整数,浮点数等值和/或键的映射。在这种情况下,您不能使用<... extended Parcelable>,并且我不想为任何其他键/值组合编写自定义方法。在这种情况下,您可以使用以下代码:

@FunctionalInterface
public interface ParcelWriter<T> {
    void writeToParcel(@NonNull final T value,
                       @NonNull final Parcel parcel, final int flags);
}

@FunctionalInterface
public interface ParcelReader<T> {
    T readFromParcel(@NonNull final Parcel parcel);
}

public static <K, V> void writeParcelableMap(
        @NonNull final Map<K, V> map,
        @NonNull final Parcel parcel,
        final int flags,
        @NonNull final ParcelWriter<Map.Entry<K, V>> parcelWriter) {
    parcel.writeInt(map.size());

    for (final Map.Entry<K, V> e : map.entrySet()) {
        parcelWriter.writeToParcel(e, parcel, flags);
    }
}

public static <K, V> Map<K, V> readParcelableMap(
        @NonNull final Parcel parcel,
        @NonNull final ParcelReader<Map.Entry<K, V>> parcelReader) {
    int size = parcel.readInt();
    final Map<K, V> map = new HashMap<>(size);

    for (int i = 0; i < size; i++) {
        final Map.Entry<K, V> value = parcelReader.readFromParcel(parcel);
        map.put(value.getKey(), value.getValue());
    }
    return map;
}

它比较冗长,但通用。这是写用法:

writeParcelableMap(map, dest, flags, (mapEntry, parcel, __) -> {
        parcel.write...; //key from mapEntry
        parcel.write...; //value from mapEntry
    });

并阅读:

map = readParcelableMap(in, parcel ->
    new AbstractMap.SimpleEntry<>(parcel.read... /*key*/, parcel.read... /*value*/)
);