Question

我对同一集合有多个查询：

第一个查询：

db.getCollection('messagelogs').find({'intents.intent':'General_Positive_Feedback'},{'currentStep':1,'_id':0})

第一个查询结果

{
"currentStep" : [ 
    "flowkuec8ta1o"
]
}

我将第二个查询的第一个查询结果'flowkuec8ta1o'用于查询（查询1结果中的'currentStep'用作以下第二个查询中键'previousStep'的值）：< / p>

第二个查询：

db.getCollection('messagelogs').find({'previousStep':'flow1pemwl7ws'},{'userMessage':1})

第二个查询结果：

{
"_id" : ObjectId("5cec2dd29b806c4a00f91f18"),
"userMessage" : "ILikedIt#1"
 }

如何将两个查询合并为一个？
请不要在this问题中标记该问题。我尝试了该问题的解决方案，但并没有为我提供理想的结果。有人可以帮我吗？

附加示例文档1（从中获取“ currentStep”的位置）：

{
"_id" : ObjectId("5cec2dc69b806c4a00f91f16"),
"currentStep" : [ 
    "flowkuec8ta1o"
],
"previousStep" : [ 
    "conditioncplwf7pw1", 
    "condition00yokr6jv", 
    "conditionzq9koi6i3"
],
"userMessage" : "Liked It :)",
"intents" : [ 
    {
        "_id" : ObjectId("5cec2dc69b806c4a00f91f17"),
        "intent" : "General_Positive_Feedback",
        "score" : "0.9774518966674806"
    }
]
}

附加示例文档2（将从查询1中获取的“ currentStep”值用作第二个查询中的“ previousStep”）

{
"_id" : ObjectId("5cec2dd29b806c4a00f91f18"),
"currentStep" : [],
"previousStep" : [ 
    "flowkuec8ta1o"
],
"userMessage" : "ILikedIt#1",
    "intents" : [ 
    {
        "_id" : ObjectId("5cec2dd29b806c4a00f91f19"),
        "intent" : "Feedback",
        "score" : "1"
    }
]
}

示例输出：

{
"_id" : ObjectId("5cec2dd29b806c4a00f91f18"),
"userMessage" : "ILikedIt#1"
}

根据@Fanparks请求添加示例文档：
以下两个文档与第一个查询匹配（以获取“ currentStep”）
文档1：

{
"currentStep": [
    "flowkuec8ta1o"
],
"previousStep": [
    "conditioncplwf7pw1",
    "condition00yokr6jv",
    "conditionzq9koi6i3"
],
"userMessage": "Liked It :)",
"intents": [{
    "_id": ObjectId("5cec2dc69b806c4a00f91f17"),
    "intent": "General_Positive_Feedback",
    "score": "0.9774518966674806"
}]
}

文档2：

{
"currentStep": [
    "flowkuec8ta1o"
],
"previousStep": [
    "conditioncplwf7pw1",
    "condition00yokr6jv",
    "conditionzq9koi6i3"
],
"userMessage": "Just Okay, Could Be better",
"intents": [{
    "_id": ObjectId("5cec2f1a9b806c4a00f91f41"),
    "intent": "General_Positive_Feedback",
    "score": "1"
}]
}

以下文档与第二个查询匹配（“ previousStep”与从第一个查询获取的“ currentStep”匹配）
文档1：

{
"currentStep": [],
"previousStep": [
    "flowkuec8ta1o"
],
"userMessage": "ILikedIt#1",
"intents": [{
    "_id": ObjectId("5cec2dd29b806c4a00f91f19"),
    "intent": "Feedback",
    "score": "1"
}]
}

文档2：

{
"currentStep": [],
"previousStep": [
    "flowkuec8ta1o"
],
"userMessage": "JustOkayCouldBeBetter#1",
"intents": [{
    "_id": ObjectId("5cec2f2b9b806c4a00f91f43"),
    "intent": "Feedback",
    "score": "1"
}]
}

文档3：

{
"currentStep": [],
"previousStep": [
    "flowkuec8ta1o"
],
"userMessage": "I'mGivingAPositiveFeedback",
"intents": [{
    "_id": ObjectId("5ced84e2fdf046078c85d9cb"),
    "intent": "Feedback",
    "score": "1"
}]
}

文档4：

{
"currentStep": [],
"previousStep": [
    "flowkuec8ta1o"
],
"userMessage": "I'm giving a positive feedback for the second time!",
"intents": [{
    "_id": ObjectId("5cede37dfdf046078c85d9e0"),
    "intent": "Feedback",
    "score": "1"
}]
}

所需的输出：

/* 1 */
{
"_id" : ObjectId("5cec2dd29b806c4a00f91f18"),
"userMessage" : "ILikedIt#1"
}

/* 2 */
{
"_id" : ObjectId("5cec2f2b9b806c4a00f91f42"),
"userMessage" : "JustOkayCouldBeBetter#1"
}

/* 3 */
{
"_id" : ObjectId("5ced84e2fdf046078c85d9ca"),
"userMessage" : "I'mGivingAPositiveFeedback"
}

/* 4 */
{
"_id" : ObjectId("5cede37dfdf046078c85d9df"),
"userMessage" : "I'm giving a positive feedback for the second time!"
}

**实际输出：**

/* 1 */
{
"_id" : ObjectId("5cec2dd29b806c4a00f91f18"),
"userMessage" : "ILikedIt#1"
}

/* 2 */
{
"_id" : ObjectId("5cec2f2b9b806c4a00f91f42"),
"userMessage" : "JustOkayCouldBeBetter#1"
}

/* 3 */
{
"_id" : ObjectId("5ced84e2fdf046078c85d9ca"),
"userMessage" : "I'mGivingAPositiveFeedback"
}

/* 4 */
{
"_id" : ObjectId("5cede37dfdf046078c85d9df"),
"userMessage" : "I'm giving a positive feedback for the second time!"
}

/* 5 */
{
"_id" : ObjectId("5cec2dd29b806c4a00f91f18"),
"userMessage" : "ILikedIt#1"
}

/* 6 */
{
"_id" : ObjectId("5cec2f2b9b806c4a00f91f42"),
"userMessage" : "JustOkayCouldBeBetter#1"
}

/* 7 */
{
"_id" : ObjectId("5ced84e2fdf046078c85d9ca"),
"userMessage" : "I'mGivingAPositiveFeedback"
}

/* 8 */
{
"_id" : ObjectId("5cede37dfdf046078c85d9df"),
"userMessage" : "I'm giving a positive feedback for the second time!"
}

Answer 1

您可以使用以下聚合

db.getCollection('messagelogs').aggregate([
  { "$match": { "intents.intent": "General_Positive_Feedback" }},
  { "$lookup": {
      "from": "messagelogs",
      "let": { "currentStep": "$currentStep" },
      "pipeline": [
        { "$match": { "$expr": { "$eq": ["$previousStep", "$$currentStep"] }}},
        { "$project": { "userMessage": 1 }}
      ],
      "as": "previousStep"
    }},
  { "$unwind": "$previousStep" },
  { "$replaceRoot": { "newRoot": "$previousStep" }},
  { "$group": {
    "_id": "$userMessage",
    "id": { "$first": "$_id" }
  }},
  { "$project": {
    "_id": "$id",
    "userMessage": "$_id"
  }}
])

MongoPlayground

Mongodb-Query db使用另一个查询的结果

1 个答案: