如果專案需求是從某些複雜的json裡面取值進行計算,用jsonpath+IK(ik-expression)來處理十分方便,jsonpath用來取json裡面的值然後用IK自帶的函式進行計算,如果是特殊的計算那就自定義IK方法搞定,配置化很方便.

下面簡單介紹下jsonpath的使用方法,主要測試都在JsonPathDemo類裡面:

下面是一個簡單的java專案demo:

注意: 其中他的max,min,avg,stddev函式只能類似於如下處理:

//正確寫法 但是感覺很雞肋
context.read("$.avg($.result.records[0].loan_type,$.result.records[1].loan_type,$.result.records[2].loan_type)");

不能傳入list 感覺比較雞肋,如果傳入list 他會報錯(如下錯誤寫法):

//這樣會報錯
Object maxV = context.read("$.max($.result.records[*].loan_type)");
//這樣也會報錯
Object maxV = context.read("$.result.records[*].loan_type.max()");
//如果json檔案中是這樣:"loan_type":"2",也會報錯,"loan_type":2 這樣才被認為是數字

報錯資訊都一樣,如下:

Exception in thread "main" com.jayway.jsonpath.JsonPathException: Aggregation function attempted to calculate value using empty array

JsonPathDemo是一個測試demo:

public class JsonPathDemo {
 public static void main(String[] args) {
 String json = FileUtils.readFileByLines("demo.json");
 ReadContext context = JsonPath.parse(json);
 //1 返回所有name
 List<String> names = context.read("$.result.records[*].name");
 //["張三","李四","王五"]
 System.out.println(names);
 //2 返回所有陣列的值
 List<Map<String,String>> objs = context.read("$.result.records[*]");
 //[{"name":"張三","pid":"500234199212121212","mobile":"18623456789","applied_at":"3","confirmed_at":"5","confirm_type":"overdue","loan_type":"1","test":"mytest","all":"2"},{"name":"李四","pid":"500234199299999999","mobile":"13098765432","applied_at":"1","confirmed_at":"","loan_type":"3","all":"3"},{"name":"王五","pid":"50023415464654659","mobile":"1706454894","applied_at":"-1","loan_type":"3"}]
 System.out.println(objs);
 //3 返回第一個的name
 String name0 = context.read("$.result.records[0].name");
 //張三
 System.out.println(name0);
 //4 返回下標為0 和 2 的陣列值
 List<String> name0and2 = context.read("$.result.records[0,2].name");
 //["張三","王五"]
 System.out.println(name0and2);
 //5 返回下標為0 到 下標為1的 的陣列值 這裡[0:2] 表示包含0 但是 不包含2
 List<String> name0to2 = context.read("$.result.records[0:2].name");
 //["張三","李四"]
 System.out.println(name0to2);
 //6 返回陣列的最後兩個值
 List<String> lastTwoName = context.read("$.result.records[-2:].name");
 //["李四","王五"]
 System.out.println(lastTwoName);
 //7 返回下標為1之後的所有陣列值 包含下標為1的
 List<String> nameFromOne = context.read("$.result.records[1:].name");
 //["李四","王五"]
 System.out.println(nameFromOne);
 //8 返回下標為3之前的所有陣列值 不包含下標為3的
 List<String> nameEndTwo = context.read("$.result.records[:3].name");
 //["張三","王五"]
 System.out.println(nameEndTwo);
 //9 返回applied_at大於等於2的值
 List<Map<String,String>> records = context.read("$.result.records[?(@.applied_at >= '2')]");
 //[{"name":"張三","all":"2"}]
 System.out.println(records);
 //10 返回name等於李四的值
 List<Map<String,String>> records0 = context.read("$.result.records[?(@.name == '李四')]");
 //[{"name":"李四","loan_type":"3"}]
 System.out.println(records0);
 //11 返回有test屬性的陣列
 List<Map<String,String>> records1 = context.read("$.result.records[?(@.test)]");
 //[{"name":"張三","all":"2"}]
 System.out.println(records1);
 //12 返回有test屬性的陣列
 List<String> list = context.read("$..all");
 //["1","4","2","3"]
 System.out.println(list);
 //12 以當前json的某個值為條件查詢 這裡ok為1 取出records陣列中applied_at等於1的陣列
 List<String> ok = context.read("$.result.records[?(@.applied_at == $['ok'])]");
 //["1","3"]
 System.out.println(ok);
 //13 正則匹配
 List<String> regexName = context.read("$.result.records[?(@.pid =~ /.*999/i)]");
 //[{"name":"李四","all":"3"}]
 System.out.println(regexName);
 //14 多條件
 List<String> mobile = context.read("$.result.records[?(@.all == '2' || @.name == '李四' )].mobile");
 //["18623456789","13098765432"]
 System.out.println(mobile);
 //14 查詢陣列長度
 Integer length01 = context.read("$.result.records.length()");
 //3
 System.out.println(length01);
 //15 查詢list裡面每個物件長度
 List<Integer> length02 = context.read("$.result.records[?(@.all == '2' || @.name == '李四' )].length()");
 //[9,8]
 System.out.println(length02);
 //16 最大值
 Object maxV = context.read("$.max($.result.records[0].loan_type,$.result.records[2].loan_type)");
 //3.0
 System.out.println(maxV);
 //17 最小值
 Object minV = context.read("$.min($.result.records[0].loan_type,$.result.records[2].loan_type)");
 //1.0
 System.out.println(minV);
 //18 平均值
 double avgV = context.read("$.avg($.result.records[0].loan_type,$.result.records[2].loan_type)");
 //2.3333333333333335
 System.out.println(avgV);
 //19 標準差
 double stddevV = context.read("$.stddev($.result.records[0].loan_type,$.result.records[2].loan_type)");
 //0.9428090415820636
 System.out.println(stddevV);
 //20 讀取一個不存在的
 String haha = context.read("$.result.haha");
 //丟擲異常
 //Exception in thread "main" com.jayway.jsonpath.PathNotFoundException: No results for path: $['result']['haha']
 //at com.jayway.jsonpath.internal.path.EvaluationContextImpl.getValue(EvaluationContextImpl.java:133)
 //at com.jayway.jsonpath.JsonPath.read(JsonPath.java:187)
 //at com.jayway.jsonpath.internal.JsonContext.read(JsonContext.java:102)
 //at com.jayway.jsonpath.internal.JsonContext.read(JsonContext.java:89)
 //at cn.lijie.jsonpath.JsonPathDemo.main(JsonPathDemo.java:58)
 //at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
 //at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
 //at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
 //at java.lang.reflect.Method.invoke(Method.java:498)
 //at com.intellij.rt.execution.application.AppMain.main(AppMain.java:147)
 System.out.println(haha);
 }
}

pom檔案引入:

 <dependency>
  <groupId>com.jayway.jsonpath</groupId>
  <artifactId>json-path</artifactId>
  <version>2.3.0</version>
 </dependency>

其中demo.json是一個測試json:

{
 "action": "/interface.service/xxx/queryBlackUserData","all": "1","result": {
 "count": 2,"tenant_count": 2,"records": [
 {
 "name": "張三","pid": "500234199212121212","mobile": "18623456789","applied_at": "3","confirmed_at": "5","confirm_type": "overdue","loan_type": 1,"test": "mytest","all": "2"
 },{
 "name": "李四","pid": "500234199299999999","mobile": "13098765432","applied_at": "1","confirmed_at": "","loan_type": 3,"all": "3"
 },{
 "name": "王五","pid": "50023415464654659","mobile": "1706454894","applied_at": "-1","loan_type": 3
 }
 ],"all": "4"
 },"code": 200,"subtime": "1480495123550","status": "success","ok": 3
}

FileUtils類是用於讀取xx.json檔案為字串的json:

public class FileUtils {
 /**
 * 以行為單位讀取檔案，常用於讀面向行的格式化檔案
 */
 public static String readFileByLines(String fileName) {
 File file = new File(fileName);
 BufferedReader reader = null;
 String str = "";
 try {
  InputStream is = FileUtils.class.getClassLoader().getResourceAsStream(fileName);
  reader = new BufferedReader(new InputStreamReader(is));
  String tempString = null;
  int line = 1;
  // 一次讀入一行，直到讀入null為檔案結束
  while ((tempString = reader.readLine()) != null) {
  // 顯示行號
  str += tempString;
  }
  reader.close();
 } catch (IOException e) {
  e.printStackTrace();
 } finally {
  if (reader != null) {
  try {
   reader.close();
  } catch (IOException e1) {
  }
  }
 }
 return str;
 }
}

補充：json介面測試的利器jsonpath

在測試REST介面的時候，經常要解析JSON，那麼可以使用開源jsonpath進行，其中看網上看到相關的說法不錯的使用場景為：

1、介面關聯

也稱為關聯引數。在應用業務介面中，完成一個業務功能時，有時候一個介面可能不滿足業務的整個流程邏輯，需要多個介面配合使用，簡單的案例如：B介面的成功呼叫依賴於A介面，需要在A介面的響應資料（response)中拿到需要的欄位，在呼叫B介面的時候，傳遞給B介面作為B介面請求引數，拿到後續響應的響應資料。

介面關聯通常可以使用正則表示式去提取需要的資料，但對於json這種簡潔、清晰層次結構、輕量級的資料互動格式，使用正則未免有點殺雞用牛刀的感覺(是的，因為我不擅長寫正則表示式)，我們需要更加簡單、直接的提取json資料的方式。

2、資料驗證

這裡的資料驗證指的是對響應結果進行資料的校驗

介面自動化測試中，對於簡單的響應結果（json），可以直接和期望結果進行比對，判斷是否完全相等即可。

如 json {"status":1,"msg":"登入成功"}

3、對於格式較複雜

尤其部分資料存在不確定性、會根據實際情況變化的響應結果，簡單的判斷是否完全相等（斷言）通常會失敗。

如：

json {"status":1,"code":"10001","data":[{"id":1,"investId":"1","createTime":"2018-04-27 12:24:01","terms":"1","unfinishedInterest":"1.0","unfinishedPrincipal":"0","repaymentDate":"2018-05-27 12:24:01","actualRepaymentDate":null,"status":"0"},{"id":2,"terms":"2","repaymentDate":"2018-06-27 12:24:01",{"id":3,"terms":"3","unfinishedPrincipal":"100.00","repaymentDate":"2018-07-27 12:24:01","status":"0"}],"msg":"獲取資訊成功"}

上面的json結構嵌套了很多資訊，完整的匹配幾乎不可能成功。比如其中的createTime資訊，根據執行介面測試用例的時間每次都不一樣。同時這個時間是響應結果中較為次要的資訊，在進行介面自動化測試時，是可以選擇被忽略的。

4、我們需要某種簡單的方法

能夠從json中提取出我們真正關注的資訊（通常也被稱為關鍵資訊）。

如提取出status的值為1，data陣列中每個物件的investId都為1，data中第三個物件的unfinishedPrincipal值為100.00，只要這三個關鍵資訊校驗通過，我們就認為響應結果沒有問題。

JSONPATH有點像XPATH了，語法規則小結下：

這裡有個表格，說明JSONPath語法元素和對應XPath元素的對比。

下面是一個簡單的json資料結構代表一個書店（原始的xml檔案是）

{ "store": {
 "book": [ 
 { "category": "reference","author": "Nigel Rees","title": "Sayings of the Century","price": 8.95
 },{ "category": "fiction","author": "Evelyn Waugh","title": "Sword of Honour","price": 12.99
 },"author": "Herman Melville","title": "Moby Dick","isbn": "0-553-21311-3","price": 8.99
 },"author": "J. R. R. Tolkien","title": "The Lord of the Rings","isbn": "0-395-19395-8","price": 22.99
 }
 ],"bicycle": {
 "color": "red","price": 19.95
 }
 }
}

比如在單元測試MOCK中，就可以這樣使用：

@RunWith(SpringRunner.class)
@SpringBootTest
@AutoConfigureMockMvc
@ActiveProfiles("test")
public class BookControllerTest { 
 @Autowired
 private MockMvc mockMvc; 
 @MockBean
 private BookRepository mockRepository; 
 /*
 {
  "timestamp":"2019-03-05T09:34:13.280+0000","status":400,"errors":["Author is not allowed.","Please provide a price","Please provide a author"]
 }
 */
 //article : jsonpath in array
 @Test
 public void save_emptyAuthor_emptyPrice_400() throws Exception {
 
 String bookInJson = "{\"name\":\"ABC\"}"; 
 mockMvc.perform(post("/books")
  .content(bookInJson)
  .header(HttpHeaders.CONTENT_TYPE,MediaType.APPLICATION_JSON))
  .andDo(print())
  .andExpect(status().isBadRequest())
  .andExpect(jsonPath("$.timestamp",is(notNullValue())))
  .andExpect(jsonPath("$.status",is(400)))
  .andExpect(jsonPath("$.errors").isArray())
  .andExpect(jsonPath("$.errors",hasSize(3)))
  .andExpect(jsonPath("$.errors",hasItem("Author is not allowed.")))
  .andExpect(jsonPath("$.errors",hasItem("Please provide a author")))
  .andExpect(jsonPath("$.errors",hasItem("Please provide a price")));
 
 verify(mockRepository,times(0)).save(any(Book.class));
 
 }
}

以上為個人經驗，希望能給大家一個參考，也希望大家多多支援我們。如有錯誤或未考慮完全的地方，望不吝賜教。