Java 中深度嵌套的 hashmap

Question

I have used this example to Accessing Deeply nested HashMaps in Java build the data structure to store node names and properties.

Here is the updated code:

class NestedMap {

private final HashMap<String, NestedMap> child;

private Map<String, Object> value = new HashMap<>();

public NestedMap() {
    child = new HashMap<>();
    setValue(null);
}

public boolean hasChild(String k) {
    return this.child.containsKey(k);
}

public NestedMap getChild(String k) {
    return this.child.get(k);
}

public void makeChild(String k) {
    this.child.put(k, new NestedMap());
}

public Map<String, Object> getValue() {
    return value;
}

public void setValue(Map<String, Object> value) {
    this.value = value;
}

}

And my usage example:

    class NestedMapIllustration {
        public static void main(String[] args) {

        NestedMap m = new NestedMap();

        m.makeChild("de");
        m.getChild("de").makeChild("content");
        m.getChild("de").getChild("content").makeChild("00");
        m.getChild("de").getChild("content").makeChild("0");
        m.getChild("de").getChild("content").makeChild("1");
        m.getChild("de").getChild("content").makeChild("01");
        m.getChild("de").getChild("content").getChild("01").makeChild("fieldsets");
        m.getChild("de").getChild("content").getChild("01").getChild("fieldsets").makeChild("0");
        m.getChild("de").getChild("content").getChild("01").getChild("fieldsets").getChild("0").makeChild("fields");
        m.getChild("de").getChild("content").getChild("01").getChild("fieldsets").getChild("0").getChild("fields").makeChild("0");
        Map<String, Object> properties = new HashMap<>();
        properties.put("key", "value");
        properties.put("key2", "value");
        m.getChild("de").getChild("content").getChild("01").getChild("fieldsets").getChild("0").getChild("fields").setValue(properties);
}

Instead of creating a new object each value I would like to always create a new HashMap where I can store the node properties.

I receive my data structure by visiting nodes in the JCR datastore and extracting their values and properties. This is how my resulting data structure should look in the output yaml file:

How can I do that more efficiently?

Answer 1

You've gone out of your way to let you use any key, but you're using string keys, even though one of the keys is "01" which suggests it's a number instead.

Can I conclude from this that keys are always strings?

In that case, why not define a separator, say, the slash, and use a plain old TreeMap<String, V> ? Then you can do:

m.put("de/content/01/fieldsets/0/fields", properties);

If you want everything in the de/content/01 'tree', you can do:

m.subMap("de/content/01/", "de/content/010");

The above will give you a map containing every child of de/content/01. The 0 at the end of the 010 there is 'magic': Zero is the next character, after slash, in the ascii table.

If you want any given key to map to any number of values, you can use:

TreeMap<String, List<V>> map = new TreeMap<>();

to put things in:

map.computeIfAbsent(key, k -> new ArrayList<>()).add(elem);

and to get things out:

for (V value : map.getOrDefault(key, List.of())) {
    // works even if key isn't in there (loops 0 times then)
}

Answer 2

Solution to the problem using recursion

public HashMap<String,Object> nestedMap(Node node) {

           HashMap<String, Object> map = new LinkedHashMap<>();

           PropertyIterator pi;
            try {
                pi = node.getProperties();
                //Get properties for the root node
                   while(pi.hasNext())
                    {
                       Property p = pi.nextProperty();
                       String name = p.getName();
                       String val = p.getString();

                       map.put(name,val);

                    }//end of while for properties of root node
            } catch (RepositoryException e) {
                // TODO Auto-generated catch block
                e.printStackTrace();
            }


           Iterable<Node> children;
            try {
                children = NodeUtil.getNodes(node);
                for (Node child : children) {

                    if (!child.getPrimaryNodeType().getName().contains("mgnl:page")) {  
                      map.put (child.getName(), nestedMap(child));

                   }//end of checking if PrimaryNodeType is of type mgnl:page
                }

            } catch (RepositoryException e) {
                // TODO Auto-generated catch block
                e.printStackTrace();
            }

            return map;
        }