我有这个 JSON 输入：
{
"entityType": "person",
"id": 24285,
"properties": {
"firstName": "Zia",
"lastName": "Rus",
"email": "[email protected]",
"phoneNumber": null,
"requestingUserGuid": "Cugxxasod Daspohs, Uyowiye",
"generalReportReference": 244,
"employment": [
{
"employerName": "Avature",
"startDate": "01-Jan-2003",
"endDate": "04-Nov-2008",
"jobTitle": ""
},
{
"employerName": "Avature",
"startDate": "06-Jul-2012",
"endDate": "",
"jobTitle": ""
}
],
"supervisor": "Supervisor",
"education": [
{
"institutionName": "Universitate",
"major": "gead",
"degree": "gased",
"startDate": "01-Feb-2022",
"endDate": "01-Oct-2024"
},
{
"institutionName": "College",
"major": "brup",
"degree": "brup",
"startDate": "05-Apr-2016",
"endDate": ""
}
],
"externalIdentifier": 24285,
"clientGuid": "6253d1c2-b02f-4513-8348-89db9b8ba449",
"productGuid": "e2f507b4-025c-439b-9ddb-833f9e537e60",
"applicantGuid": ""
}
}
到目前为止，我已经成功地根据需要进行了整个数据转换，但我缺少将数据字段格式化为“YYYY-MM-DD”格式的方法。 我见过格式化数据字段的方法，但大多数都是通过默认值或覆盖数据。 这些解决方案的问题在于，正如您在输入 JSON 中看到的那样，我可以在“employment”和“education”数组中包含一个或多个对象，并且可以包含“EndDate”，也可以不包含“EndDate”。因此，使用“默认数据”解决方案对我不起作用。
这是我目前用来转换数据的 JOLT：
[
{
"operation": "shift",
"spec": {
"properties": {
"firstName": "firstName",
"lastName": "lastName",
"email": "email",
"phoneNumber": "phoneNumber",
"requestingUserGuid": "requestingUserGuid",
"generalReportReference": "generalReportReference",
"clientGuid": "clientGuid",
"productId": "productId",
"applicantGuid": "applicantGuid",
"externalIdentifier": "externalIdentifier",
"employment": {
"*": {
"employerName|endDate": {
"": null,
"*": {
"@1": "&4[&3].&2"
}
},
"*": {
"*": {
"@1": "&4[&3].&2"
}
}
}
},
"education": {
"*": {
"institutionName|registrarPhone|endDate": {
"": null,
"*": {
"@1": "&4[&3].&2"
}
},
"*": {
"*": {
"@1": "&4[&3].&2"
}
}
}
}
},
"education": {
"*": {
"~institutionName": "N/A",
"~registrarPhone": "(111) 111-1111"
}
},
"employment": {
"*": {
"~employerName": "N/A"
}
}
}
},
{
"operation": "modify-default-beta",
"spec": {
"employment": {
"*": {
"employerName": "N/A",
"firstNameUsed": "@(4,firstName)",
"lastNameUsed": "@(4,lastName)"
}
},
"education": {
"*": {
"institutionName": "N/A",
"registrarPhone": "(111) 111-1111"
}
}
}
},
{
"operation": "modify-overwrite-beta",
"spec": {
"employment": {
"*": {
"supervisor": "Supervisor"
}
}
}
}
]
我看不到如何将“日期”字段转换为所需的格式。 有人可以帮我解决这个问题吗？ 这可以通过 JOLT 实现吗？或者我应该采取不同的方法并尝试在 NiFi 中使用另一个处理器？
提前致谢！
您可以使用以下转换规范处理 JoltTransformJSON
处理器中的所有date 排列，其中 modify 规范的 split 函数将所有 date 值转换为 arrays 以在其中使用shift，其中月份缩写与其后面的数字表示相匹配，然后再次通过 modify 规范的 join 函数组合日期数组的组成部分，例如
[
{
"operation": "modify-overwrite-beta",
"spec": {
"properties": {
"*": {
"*": {
"*Date": "=split('-',@(1,&))"
}
}
}
}
},
{
"operation": "shift",
"spec": {
"*": "&", //the elements other than "properties"
"properties": {
"*": "&1.&", //the elements other than "employment" and "education"
"emp*|edu*": {
"*": {
"*": "&3.&2[&1].&", //the elements other than "...Dates", reduce the levels by 1 w.r.t. the line 2 levels down
"*Date": {
"2": "&4.&3[&2].&1", //[&2] represents the indexes of the "employment" or "education" in order to provide array manner back
"1": { //Months part
"Jan": { "#01": "&6.&5[&4].&3" }, //leaf nodes are 2 levels deeper than the other Date arrays' nodes
"Feb": { "#02": "&6.&5[&4].&3" },
"Mar": { "#03": "&6.&5[&4].&3" },
"Apr": { "#04": "&6.&5[&4].&3" },
"May": { "#05": "&6.&5[&4].&3" },
"Jun": { "#06": "&6.&5[&4].&3" },
"Jul": { "#07": "&6.&5[&4].&3" },
"Aug": { "#08": "&6.&5[&4].&3" },
"Sep": { "#09": "&6.&5[&4].&3" },
"Oct": { "#10": "&6.&5[&4].&3" },
"Nov": { "#11": "&6.&5[&4].&3" },
"Dec": { "#12": "&6.&5[&4].&3" }
},
"0": "&4.&3[&2].&1"
}
}
}
}
}
},
{
"operation": "modify-overwrite-beta",
"spec": {
"properties": {
"*": {
"*": {
"*Date": "=join('-',@(1,&))"
}
}
}
}
}
]