Snack3 之 Jsonpath使用

一、 Snack3 和 JSONPath 介绍

Snack3 是一个支持JSONPath的JSON框架。JSONPath是一个很强大的功能，也可以在Java框架中当作对象查询语言（OQL）来使用。

<dependency>
  <groupId>org.noear</groupId>
  <artifactId>snack3</artifactId>
  <version>3.1.5.10</version>
</dependency>

Snack3 借签了 Javascript 所有变量由 var 申明，及 Xml dom 一切都是 Node 的设计。其下一切数据都以ONode表示，ONode也即 One node 之意，代表任何类型，也可以转换为任何类型。

• 强调文档树的操控和构建能力
• 做为中间媒体，方便不同格式互转
• 高性能Json path查询（兼容性和性能很赞）
• 支持序列化、反序列化

二、接口

public class ONode{
    //...
    /**
     * Json path select
     *
     * @param jpath json path express
     * @param useStandard use standard mode(default: false)
     * @param cacheJpath cache json path parsing results
     */
    public ONode select(String jpath,  boolean useStandard, boolean cacheJpath) {
        return JsonPath.eval(this, jpath, useStandard, cacheJpath);
    }

    public ONode select(String jpath,  boolean useStandard) {
        return select(jpath, useStandard, true);
    }

    public ONode select(String jpath) {
        return select(jpath, false);
    }
    //...  
}

默认使用缓存JSONPath解析对象，可提供几倍性能效果。

三、支持语法

• 字符串使用单引号，例：[‘name‘]
• 过滤操作用空隔号隔开，例：[?(@.type == 1)]

像这两种写法的语义是差不多：

$.store.book[0].title //建议使用这种

$['store']['book'][0]['title']

四、语法示例说明

五、接口使用示例

示例1:

读取对象的属性

Entity entity = new Entity(123, new Object());
ONode n = ONode.load(entity);

assert n.select("$.id").getInt() == 123;
assert n.select("$.*").count() == 2;

public static class Entity {
   public int id;
   public String name;
   public Object value;
   public Entity(int id, Object value) { this.id = id; this.value = value; }
   public Entity(String name) { this.name = name; }
}

示例2

读取集合多个元素的某个属性

List<Entity> entities = new ArrayList<Entity>();
entities.add(new Entity("wenshao"));
entities.add(new Entity("ljw2083"));
ONode n = ONode.load(entities);

List<String> names = n.select("$.name").toObject(List.class); 
assert names.size() == 2;

示例3

返回集合中多个元素

List<Entity> entities = new ArrayList<Entity>();
entities.add(new Entity("wenshao"));
entities.add(new Entity("ljw2083"));
entities.add(new Entity("Yako"));
ONode n = ONode.load(entities);

List<Entity> result = n.select("$[1,2]").toObject((new ArrayList<Entity>() {}).getClass());
assert result.size() == 2;

示例4

按范围返回集合的子集

List<Entity> entities = new ArrayList<Entity>();
entities.add(new Entity("wenshao"));
entities.add(new Entity("ljw2083"));
entities.add(new Entity("Yako"));
ONode n = ONode.load(entities);

List<Entity> result = n.select("$[0:2]").toObject((new ArrayList<Entity>(){}).getClass());
assert result.size() == 2;

示例5

通过条件过滤，返回集合的子集

List<Entity> entities = new ArrayList<Entity>();
entities.add(new Entity(1001, "ljw2083"));
entities.add(new Entity(1002, "wenshao"));
entities.add(new Entity(1003, "yakolee"));
entities.add(new Entity(1004, null));
ONode n = ONode.load(entities);

ONode rst = n.select("$[?($.id in [1001,1002])]");
assert rst.count() == 2;

示例6

根据属性值过滤条件判断是否返回对象，修改对象，数组属性添加元素

Entity entity = new Entity(1001, "ljw2083");
ONode n = ONode.load(entity);

assert n.select("$[?(id == 1001)]").isObject();
assert n.select("$[?(id == 1002)]").isNull();

n.select("$").set("id",123456);
assert n.get("id").getInt() == 123456;

n.get("value").add(1).add(2).add(3);
assert n.get("value").count() == 3;

示例7

Map root = Collections.singletonMap("company",
        Collections.singletonMap("departs",
                Arrays.asList(
                        Collections.singletonMap("id",
                                1001),
                        Collections.singletonMap("id",
                                1002),
                        Collections.singletonMap("id", 1003)
                )
        ));

ONode n = ONode.load(root);

List<Object> ids = n.select("$..id").toObject(List.class);
assertEquals(3l, ids.size());
assertEquals(1001l, ids.get(0));
assertEquals(1002l, ids.get(1));
assertEquals(1003l, ids.get(2));

具体用例测试请看下面:

String jsonStr = "{\n" +
        "    \"store\": {\n" +
        "        \"bicycle\": {\n" +
        "            \"color\": \"red\",\n" +
        "            \"price\": 19.95\n" +
        "        },\n" +
        "        \"book\": [\n" +
        "            {\n" +
        "                \"author\": \"刘慈欣\",\n" +
        "                \"price\": 8.95,\n" +
        "                \"category\": \"科幻\",\n" +
        "                \"title\": \"三体\"\n" +
        "            },\n" +
        "            {\n" +
        "                \"author\": \"itguang\",\n" +
        "                \"price\": 12.99,\n" +
        "                \"category\": \"编程语言\",\n" +
        "                \"title\": \"go语言实战\"\n" +
        "            }\n" +
        "        ]\n" +
        "    }\n" +
        "}";

ONode o = ONode.load(jsonStr);

//得到所有的书
ONode books = o.select("$.store.book");
System.out.println("books=::" + books);

//得到所有的书名
ONode titles = o.select("$.store.book.title");
System.out.println("titles=::" + titles);

//第一本书title
ONode title = o.select("$.store.book[0].title");
System.out.println("title=::" + title);

//price大于10元的book
ONode list = o.select("$.store.book[?(price > 10)]");
System.out.println("price大于10元的book=::" + list);

//price大于10元的title
ONode list2 = o.select("$.store.book[?(price > 10)].title");
System.out.println("price大于10元的title=::" + list2);

//category(类别)为科幻的book
ONode list3 = o.select("$.store.book[?(category == '科幻')]");
System.out.println("category(类别)为科幻的book=::" + list3);


//bicycle的所有属性值
ONode values = o.select("$.store.bicycle.*");
System.out.println("bicycle的所有属性值=::" + values);


//bicycle的color和price属性值
ONode read = o.select("$.store.bicycle['color','price']");
System.out.println("bicycle的color和price属性值=::" + read);

打印结果

books=::[{"author":"刘慈欣","price":8.95,"category":"科幻","title":"三体"},{"author":"itguang","price":12.99,"category":"编程语言","title":"go语言实战"}]
titles=::["三体","go语言实战"]
title=::"三体"
price大于10元的book=::[{"author":"itguang","price":12.99,"category":"编程语言","title":"go语言实战"}]
price大于10元的title=::["go语言实战"]
category(类别)为科幻的book=::[{"author":"刘慈欣","price":8.95,"category":"科幻","title":"三体"}]
bicycle的所有属性值=::["red",19.95]
bicycle的color和price属性值=::["red",19.95]