fastjson确实很厉害(毕竟是阿里大牛写的),在逻辑解析的快捷上可以坐到很方便的搞定,这样直接就直接缩小的很多的工作量,并且代码结构很清晰。
基于以上摘要内容,我就闲着无聊比较了一下。代码如下
public void testFastJson(){
String jsonString = "[\n" +
"{\n" +
"orderId: 2964,\n" +
"code: \"2800001296442\",\n" +
"radiateCenterId: 1,\n" +
"userId: 7,\n" +
"totalMoney: 22400,\n" +
"status: 7,\n" +
"payStatus: 2,\n" +
"sendStatus: 2,\n" +
"payway: 1,\n" +
"gmtCreate: \"2015-10-28 18:28:42\",\n" +
"shopName: \"水果无忧\",\n" +
"itemsName: \"小黄瓜,无籽瓜\"\n" +
"},\n" +
"{\n" +
"orderId: 2940,\n" +
"code: \"2700001294044\",\n" +
"radiateCenterId: 1,\n" +
"userId: 7,\n" +
"totalMoney: 650,\n" +
"status: 7,\n" +
"payStatus: 2,\n" +
"sendStatus: 2,\n" +
"payway: 1,\n" +
"gmtCreate: \"2015-10-27 20:05:44\",\n" +
"shopName: \"水果无忧\",\n" +
"itemsName: \"两个装柠檬\"\n" +
"},\n" +
"{\n" +
"orderId: 2913,\n" +
"code: \"2600001291319\",\n" +
"radiateCenterId: 1,\n" +
"userId: 7,\n" +
"totalMoney: 72500,\n" +
"status: 7,\n" +
"payStatus: 2,\n" +
"sendStatus: 2,\n" +
"payway: 1,\n" +
"gmtCreate: \"2015-10-26 20:56:19\",\n" +
"shopName: \"水果无忧\",\n" +
"itemsName: \"韩国梨,纸箱白心柚,小黄瓜,野生苹果\"\n" +
"},\n" +
"{\n" +
"orderId: 2891,\n" +
"code: \"2500001289120\",\n" +
"radiateCenterId: 1,\n" +
"userId: 7,\n" +
"totalMoney: 9000,\n" +
"status: 7,\n" +
"payStatus: 2,\n" +
"sendStatus: 2,\n" +
"payway: 1,\n" +
"gmtCreate: \"2015-10-25 21:06:20\",\n" +
"shopName: \"水果无忧\",\n" +
"itemsName: \"哈密瓜(4个装)\"\n" +
"},\n" +
"{\n" +
"orderId: 2865,\n" +
"code: \"2400001286555\",\n" +
"radiateCenterId: 1,\n" +
"userId: 7,\n" +
"totalMoney: 81100,\n" +
"status: 7,\n" +
"payStatus: 2,\n" +
"sendStatus: 2,\n" +
"payway: 1,\n" +
"gmtCreate: \"2015-10-24 19:57:55\",\n" +
"shopName: \"水果无忧\",\n" +
"itemsName: \"西梅,哈密瓜(4个装),云南蜜桔,小黄瓜,木瓜,无籽瓜\"\n" +
"},\n" +
"{\n" +
"orderId: 2839,\n" +
"code: \"2300001283923\",\n" +
"radiateCenterId: 1,\n" +
"userId: 7,\n" +
"totalMoney: 69000,\n" +
"status: 7,\n" +
"payStatus: 2,\n" +
"sendStatus: 2,\n" +
"payway: 1,\n" +
"gmtCreate: \"2015-10-23 20:50:23\",\n" +
"shopName: \"水果无忧\",\n" +
"itemsName: \"哈密瓜(4个装),有机红心柚,凯特芒果\"\n" +
"},\n" +
"{\n" +
"orderId: 2820,\n" +
"code: \"2200001282039\",\n" +
"radiateCenterId: 1,\n" +
"userId: 7,\n" +
"totalMoney: 37500,\n" +
"status: 7,\n" +
"payStatus: 2,\n" +
"sendStatus: 2,\n" +
"payway: 1,\n" +
"gmtCreate: \"2015-10-22 20:43:39\",\n" +
"shopName: \"水果无忧\",\n" +
"itemsName: \"马奶提,哈密瓜(4个装),小黄瓜\"\n" +
"},\n" +
"{\n" +
"orderId: 2792,\n" +
"code: \"2100001279242\",\n" +
"radiateCenterId: 1,\n" +
"userId: 7,\n" +
"totalMoney: 65500,\n" +
"status: 7,\n" +
"payStatus: 2,\n" +
"sendStatus: 2,\n" +
"payway: 1,\n" +
"gmtCreate: \"2015-10-21 22:11:42\",\n" +
"shopName: \"水果无忧\",\n" +
"itemsName: \"菲律宾凤梨,辽宁葡萄,小黄瓜,野生苹果\"\n" +
"},\n" +
"{\n" +
"orderId: 2773,\n" +
"code: \"2000001277334\",\n" +
"radiateCenterId: 1,\n" +
"userId: 7,\n" +
"totalMoney: 68800,\n" +
"status: 7,\n" +
"payStatus: 2,\n" +
"sendStatus: 2,\n" +
"payway: 1,\n" +
"gmtCreate: \"2015-10-20 22:14:34\",\n" +
"shopName: \"水果无忧\",\n" +
"itemsName: \"菲律宾香蕉,无籽瓜,野生苹果\"\n" +
"},\n" +
"{\n" +
"orderId: 2737,\n" +
"code: \"1900001273702\",\n" +
"radiateCenterId: 1,\n" +
"userId: 7,\n" +
"totalMoney: 101000,\n" +
"status: 7,\n" +
"payStatus: 2,\n" +
"sendStatus: 2,\n" +
"payway: 1,\n" +
"gmtCreate: \"2015-10-19 21:35:02\",\n" +
"shopName: \"水果无忧\",\n" +
"itemsName: \"韩国梨,猕猴桃金果\"\n" +
"}\n" +
"]";
//这是待解析的json字符串。
long t1,t2;
t1 =System.currentTimeMillis();
List<MyOrderBean> list = JSONArray.parseArray(jsonString, MyOrderBean.class);
t2 =System.currentTimeMillis();
logError("fastjson = " + (t2 - t1));
}
以及自己的解析代码:
public class OrderHelper {
public static long t1,t2;
public static List<MyOrderBean> getOrderList(JSONObject object){
t1 =System.currentTimeMillis();
JSONArray array = JsonUtil.getJsonArray(object, "data");
List<MyOrderBean> list = CollectionUtil.newArrayList();
for (int i = 0; i < array.length(); i++) {
try {
JSONObject obj = JsonUtil.getJsonObject(array.get(i));
list.add(getOrder(obj));
} catch (JSONException e) {
e.printStackTrace();
}
}
t2 =System.currentTimeMillis();
LogUtil.logTagE("my useTime = " + (t2 - t1));
return list;
}
public static MyOrderBean getOrder(JSONObject obj){
MyOrderBean orders = new MyOrderBean();
orders.setId(JsonUtil.getInt(obj, "orderId", 0));
orders.setCode(JsonUtil.getString(obj, "code", ""));
orders.setGmtCreate(JsonUtil.getString(obj, "gmtCreate", ""));
orders.setItemName(JsonUtil.getString(obj, "itemsName", ""));
orders.setItemNum(JsonUtil.getInt(obj, "itemNum", 0));
orders.setPayWay(JsonUtil.getInt(obj, "payway", 0));
orders.setShopName(JsonUtil.getString(obj, "shopName", ""));
orders.setStatus(JsonUtil.getInt(obj, "status", 0));
orders.setShopId(JsonUtil.getInt(obj, "shopId", 0));
orders.setTotalMoney(PriceUtil.showPriceYang( JsonUtil.getDouble(obj, "totalMoney", 0.0d)));
orders.setPayStatus(JsonUtil.getInt(obj, "payStatus", 1));
orders.setSendStatus(JsonUtil.getInt(obj, "sendStatus", 1));
return orders;
}
}
运行结果如下:
结论如上、差距比较大。。。看到这个结果我就心安了,不说了我要去嘲讽我同学了。
我要声明,我没有讽刺fastjson,我必须承认fastjson是目前json解析界特别强大的类库,关于为什么可以移步到这个博客http://www.oschina.net/question/54100_30041
只是总结了一下我不用fastjson的原因如下:
每个项目,服务端的数据是多变的,根据不同的界面,服务端传回的数据是不一样的,这样就导致你整个实体里面可能有很多属性不是当前页面用得到的,fastjson的解析原理是通过反射找到所有get方法,于是再逐个比较解析。这可能是一个耗时长的原因,毕竟有很多无用的属性,fastjson是越来越快的,也就是说几十条几百条数据是看不出来结果的,他的性能优化真正是体现在大数据量的处理。本篇比较是不具有权威性和代表性的,毕竟数据量小,只是记录娱乐性的。所以希望大家结合使用情况选择使用,另外每次的开源库最好是使用最新版本的,比如fastjson1.1和1.2的解析速度差距也是很大的。
习惯自己写解析类,更加方便精确的针对每一个实体做解析,主要也是我自己写习惯了,闭着眼睛也知道怎么写,不过我这样有点小不方便,就是换一家公司之后,新的网络规范必须要重新适应,fastjson使用频率广泛我想也是有这点原因在里面。毕竟不用动脑直接一波流。
====================================分割线================================
好久没写博客了,毕业两年来,做合伙人这么长时间,经历了很多艰难的时期,挺过了好几个,什么都干过,运营,产品,市场地推,物流跟车,仓库分拣,累的不要不要的,技术成长却很少,看书时间减少,个人安排时间不足,上周末跟网易的同学聊天扯淡,真心羡慕一心写代码的生活。大企业不管有多么多的诟病沉珂,但是不可否认那里聚集了很多各自领域都很优秀的人才,圈子决定眼界,至少技术上可以帮助更加快速成长,这是我在预感项目失败公司清算的时候自己应该怎么选择的感悟。最近几个月是公司从未有过的艰难时期,这一次,可能坚持不下去了。毕竟太年轻,积累太少,也因为年轻,可以多尝试几条路。
很多年轻人(我自己也是)不喜欢大公司的环境,不习惯大公司的相处方式,整天吐槽臃肿不堪甚至浪费资源的公司结构和制度,网上一直在讨论的,去大公司还是去小公司,大神们有无数的结果和比较,我小小平凡人一枚就不在此浪费字数了。你选择什么样的工作,什么样的环境,什么样规模的平台,优劣都有, 其实在我看来都无所谓,别忘了自己还年轻,走错路可以回头,就算不想回头也可以继续往下走。所以只要肯走,没什么可怕的。
共勉。
文章转载自 开源中国社区[https://www.oschina.net]