Mongodb 将文档与其嵌套数组及其嵌套数组合并
我正在尝试使用聚合框架创建一个查询,但无法获得所需的结果。 我有一个经销商集合,每个经销商有一个客户列表,每个客户有一个成员列表,结构如下:Mongodb 将文档与其嵌套数组及其嵌套数组合并,mongodb,aggregation-framework,Mongodb,Aggregation Framework,我正在尝试使用聚合框架创建一个查询,但无法获得所需的结果。 我有一个经销商集合,每个经销商有一个客户列表,每个客户有一个成员列表,结构如下: [ { "userID" : "xxx", "userType" : "RESELLER", "clients" : [ { "userID" : "xxx", "userType" : "CLIENT", "members" : [ {
[
{
"userID" : "xxx",
"userType" : "RESELLER",
"clients" : [
{
"userID" : "xxx",
"userType" : "CLIENT",
"members" : [
{
"userID" : "xxx",
"userType" : "MEMBER"
},
{
"userID" : "xxx",
"userType" : "MEMBER"
}
]
},
{
"userID" : "xxx",
"userType" : "CLIENT",
"members" : [
{
"userID" : "xxx",
"userType" : "MEMBER"
},
{
"userID" : "xxx",
"userType" : "MEMBER"
}
]
}
]
},
{
"userID" : "xxx",
"userType" : "RESELLER",
"clients" : [
{
"userID" : "xxx",
"userType" : "CLIENT",
"members" : [
{
"userID" : "xxx",
"userType" : "MEMBER"
},
{
"userID" : "xxx",
"userType" : "MEMBER"
}
]
},
{
"userID" : "xxx",
"userType" : "CLIENT",
"members" : [
{
"userID" : "xxx",
"userType" : "MEMBER"
},
{
"userID" : "xxx",
"userType" : "MEMBER"
}
]
}
]
}
]
我想得到的结果是:
[
{
"userID" : "xxx",
"userType" : "RESELLER"
},
{
"userID" : "xxx",
"userType" : "RESELLER"
},
{
"userID" : "xxx",
"userType" : "CLIENT"
},
{
"userID" : "xxx",
"userType" : "CLIENT"
},
{
"userID" : "xxx",
"userType" : "CLIENT"
},
{
"userID" : "xxx",
"userType" : "CLIENT"
},
{
"userID" : "xxx",
"userType" : "MEMBER"
},
{
"userID" : "xxx",
"userType" : "MEMBER"
},
{
"userID" : "xxx",
"userType" : "MEMBER"
},
{
"userID" : "xxx",
"userType" : "MEMBER"
},
{
"userID" : "xxx",
"userType" : "MEMBER"
}
]
我做了很多尝试,但我没有得到这个结果。
我所做的最接近的解决方案是:
db.resellers.aggregate([
{
$unwind: "$clients"
},
{
$project: {
_id : 0,
teamMembers : "$clients.members"
}
},
{
$unwind: "$members"
},
{
$project: {
_id : 0,
userID : "$members.userID",
type : "$members.type"
}
}
]).pretty()
此解决方案仅返回成员列表,因此我必须做些什么才能获得一个包含所有经销商、客户和成员的列表?您可以使用来扁平化数据结构,然后运行以获得每个文档的单个成员:
db.collection.aggregate([
{ "$project": {
"members": {
"$concatArrays": [
[{ "userID": "$userID", "userType": "$userType" }],
{ "$reduce": {
"input": "$clients",
"initialValue": [],
"in": {
"$concatArrays": [
"$$value",
[{ "userID": "$$this.userID", "userType": "$$this.userType" }],
"$$this.members"
]
}
}}
]
}
}},
{ "$unwind": "$members" },
{ "$replaceRoot": { "newRoot": "$members" }}
])
好吧,您可以在
$project
阶段完成这项工作
[
{
"$project": {
"members": {
"$reduce": {
"input": {
"$map": {
"input": "$clients",
"in": {
"$concatArrays": [
[
{
"userID": "$userID",
"userType": "$userType"
},
{
"userID": "$$this.userID",
"userType": "$$this.userType"
}
],
"$$this.members"
]
}
}
},
"initialValue": [
],
"in": {
"$concatArrays": [
"$$this",
"$$value"
]
}
}
}
}
}
]
谢谢styvane,您的解决方案有效,我只需使用replaceRoot函数将“$members”展开并使其成为新的根目录。使用此查询,您无需
$unwind
数组或使用$replaceRoot