Question

我想通过以下查询获得预期结果。我知道$where无法使用$match。

我应该如何比较$match阶段的字符串的每个字段？

查询

db.col_sample.aggregate([
    {$unwind: "$array"},
    {$lookup: {
        from: "col_population",
        localField: "array",
        foreignField: "foo.key",
        as: "join"
    }},
    {$match: {
        "join.foo.key2": "foo.key3"
    }}
])

预期结果

{
  "_id": (snip),
  "foo": {
    "key1": "0xBBB",
    "key2": "bbb",
    "key3": "ccc"
  },  
  "array": [
    "0xCCC",
    "0xDDD"
  ]
}

col_population

{ 
  "foo": {
    "key1": "0xBBB",
    "key2": "bbb",
    "key3": "ccc"
  },  
  "array": [
    "0xCCC",
    "0xDDD"
  ]
}
{ 
  "foo": {
    "key1": "0xCCC",
    "key2": "ccc",
    "key3": "ddd"
  },  
  "array": [
    "0xDDD",
    "0xEEE"
  ]
}

col_sample

{ 
  "foo": {
    "key1": "0xAAA",
    "key2": "aaa",
    "key3": "bbb"
  },  
  "array": [
    "0xBBB",
    "0xCCC"
  ]
}

Answer 1

您需要使用 $redact 管道而不是 $match ，如下所示：

db.col_sample.aggregate([
    {$unwind: "$array"},
    {$lookup: {
        from: "col_population",
        localField: "array",
        foreignField: "foo.key1",
        as: "join"
    }},
    {$unwind: "$join"},
    {$redact: {
        $cond: [
            { $eq: [ "$join.foo.key2", "$foo.key3" ] }, /// typo s/sjoin/join/
            "$$KEEP",
            "$$PRUNE"
        ]
    }}
])

比较$ match阶段的字段（字符串）

1 个答案: