MongoDb 聚合数据操作 - 对象到数组 [英] MongoDb Aggregation Data manipulation - Objects to Arrays

查看：24
本文介绍了MongoDb 聚合数据操作 - 对象到数组的处理方法，对大家解决问题具有一定的参考价值，需要的朋友们下面随着小编来一起学习吧！

问题描述

我有以下示例数据集

[{_id":{$oid":60f83d3cd66842301905aa77"},id":527438，名称":CryptoPunk #4050"，资产合同":{名称":CryptoPunks"，地址":0xb47e3cd837ddf8e4c57f05d70ab865de6e193bbb"},特性":[{trait_type":类型"，值":男"，显示类型":空，max_value":空，trait_count":6039，订单":空},{trait_type":附件"，价值":莫霍克"，显示类型":空，max_value":空，trait_count":441，订单":空},{trait_type":附件"，价值":耳环"，显示类型":空，max_value":空，trait_count":2459，订单":空},{trait_type":附件"，价值":皱眉"，显示类型":空，max_value":空，trait_count":261，订单":空}],token_id":4050"，永久链接":https://opensea.io/assets/0xb47e3cd837ddf8e4c57f05d70ab865de6e193bbb/4050"，background_color":空，image_url":https://lh3.googleusercontent.com/sO18rDQYhC5yIcj12RVsv31pbbsZo_2muQQbTJMQHn47EKGhnirs8mxzohm58HAZ7taBoe4pU6x1qntlExkquot;TtJimage_preview_url":https://lh3.googleusercontent.com/sO18rDQYhC5yIcj12RVsv31pbbsZo_2muQQbTJMQHn47EKGhnirs8mxzohm58HAZ7taBoe4pU6x1qntlJExk"Tanimation_url":空，vault_contract":0x269616d549d7e8eaa82dfb17028d0b212d11232a"},{_id":{$oid":60f83d3cbc3f0161da2141f7"},id":17736625，姓名":OJ辛普森"，资产合同":{名称":哈希掩码"，地址":0xc2c747e0f7004f9e8817db2ca4997657a7746928"},特性":[{trait_type":字符"，值":男"，显示类型":空，max_value":空，trait_count":8659，订单":空},{trait_type":面具"，价值":涂鸦"，显示类型":空，max_value":空，trait_count":2187，订单":空},{trait_type":眼睛颜色"，价值":黑暗"，显示类型":空，max_value":空，trait_count":7419，订单":空},{trait_type":物品"，价值":无项目"，显示类型":空，max_value":空，trait_count":14533，订单":空},{trait_type":肤色"，价值":黑暗"，显示类型":空，max_value":空，trait_count":3784，订单":空},{trait_type":令牌ID"，值":3535，display_type":数字"，max_value":空，trait_count":0，订单":空},{trait_type":背景"，价值":涂鸦"，显示类型":空，max_value":空，trait_count":5538，订单":空}],token_id":3535"，永久链接":https://opensea.io/assets/0xc2c747e0f7004f9e8817db2ca4997657a7746928/3535"，background_color":空，"image_url":"https://lh3.googleusercontent.com/NZQu7CNjgJ_1uhbUVwEb-14rZPJmPCaqaXy0qnUpgm5Qll0BvmmF7tPMjBhFH6ZZp_qzOPxHi0NFmRkOjDcLuBuQ0"image_preview_url":"https://lh3.googleusercontent.com/NZQu7CNjgJ_1uhbUVwEb-14rZPJmPCaqaXy0qnUpgm5Qll0BvmmF7tPMjBhFH6ZZp_qzOPxHi0NFmRk0sBvmmF7tPMjBhFH6ZZp_qzOPxHi0NFmRk8sB0DcOxLuBJ_1uhbUVwEb-14rZPJmPanimation_url":空，vault_contract":0xc7a8b45e184138114e6085c82936a8db93dd156a"}]

我想更新为

[{_id":{$oid":60f83d3cd66842301905aa77"},id":527438，名称":CryptoPunk #4050"，资产合同":{名称":CryptoPunks"，地址":0xb47e3cd837ddf8e4c57f05d70ab865de6e193bbb"},特质":{类型":男性"，配件":[莫霍克"、耳环"、皱眉"]、},token_id":4050"，永久链接":https://opensea.io/assets/0xb47e3cd837ddf8e4c57f05d70ab865de6e193bbb/4050"，background_color":空，image_url":https://lh3.googleusercontent.com/sO18rDQYhC5yIcj12RVsv31pbbsZo_2muQQbTJMQHn47EKGhnirs8mxzohm58HAZ7taBoe4pU6x1qntlExkquot;TtJimage_preview_url":https://lh3.googleusercontent.com/sO18rDQYhC5yIcj12RVsv31pbbsZo_2muQQbTJMQHn47EKGhnirs8mxzohm58HAZ7taBoe4pU6x1qntlJExk"Tanimation_url":空，vault_contract":0x269616d549d7e8eaa82dfb17028d0b212d11232a"},{_id":{$oid":60f83d3cbc3f0161da2141f7"},id":17736625，姓名":OJ辛普森"，资产合同":{名称":哈希掩码"，地址":0xc2c747e0f7004f9e8817db2ca4997657a7746928"},特性":{人物":男"，面具":涂鸦"，eye_color":深色"，项目":无项目"，skin_color":深色"，token_id":3535，背景":涂鸦"，},token_id":3535"，永久链接":https://opensea.io/assets/0xc2c747e0f7004f9e8817db2ca4997657a7746928/3535"，background_color":空，"image_url":"https://lh3.googleusercontent.com/NZQu7CNjgJ_1uhbUVwEb-14rZPJmPCaqaXy0qnUpgm5Qll0BvmmF7tPMjBhFH6ZZp_qzOPxHi0NFmRkOjDcLuBuQ0"image_preview_url":"https://lh3.googleusercontent.com/NZQu7CNjgJ_1uhbUVwEb-14rZPJmPCaqaXy0qnUpgm5Qll0BvmmF7tPMjBhFH6ZZp_qzOPxHi0NFmRk0sBvmmF7tPMjBhFH6ZZp_qzOPxHi0NFmRk8sB0DcOxLuBJ_1uhbUVwEb-14rZPJmPanimation_url":空，vault_contract":0xc7a8b45e184138114e6085c82936a8db93dd156a"}]

背后的逻辑是

  • 查看 Traits 数组对象
  • 获取 trait_type 值并使用小写名称(空格下划线)创建一个新键
  • 将新键的值设置为value"
    • 的值

所以，

 "trait_type": "type",值":男"，//变成类型":男性"

  • 如果有多个相同特征类型的实例，则创建一个值数组.

所以，

{trait_type":附件"，价值":莫霍克"，显示类型":空，max_value":空，trait_count":441，订单":空},{trait_type":附件"，价值":耳环"，显示类型":空，max_value":空，trait_count":2459，订单":空},//变成配饰":[莫霍克"、耳环"]

解决方案

查询

  • 即使管道不允许我们使用这里使用的组查找等阶段，它也是一个聚合更新.(您可以使用 $out 并在之后替换集合或 $merge 来替换文档(类似于更新))

  • 第一张地图

    • 对于每个特征(特征的文档成员)，它都将其放入数组
      [[trait_type":type"] [value":Male"] [display_type":null] ...]
    • 减少该数组以仅从它们构建 1 个文档
      <代码>{类型"类型"，值":"Male"}(小写和_"也是如此)

  • 现在的特点是一样的

    特性":[{类型":类型"，价值":男性"},{类型":附件"，价值":莫霍克"},{类型":附件"，价值":耳环"},{类型":附件"，价值":皱眉"}]

  • 使用虚拟集合 [{}] 查找(我们这样做是为了在该数组中创建一个组)它就像一个技巧，允许我们在 1 个文档中使用阶段运算符

    • 查找管道按类型展开和分组

    特性":[{值":[莫霍克"，耳环"，皱眉"],类型":附件"；},{值":[男"],类型":类型"}]

    • 然后它是一个替换根来获取类型的值，使其成为字段名和值作为值 (如果 size=1 删除数组)

  • 查找后我们有

    特质":[{附件":[莫霍克"，耳环"，皱眉"]},{类型":男性"}]

  • 所以我们要做的就是减少特征并合并对象(无论如何，键都是唯一的，因为我们按它们分组)

  • 我们得到了预期的输出(至少我认为没问题)

此处测试代码

db.collection.aggregate([{$set":{特性":{$地图":{输入":$traits"，作为":t"，在":{$减少":{输入":{$地图":{输入":{$objectToArray":$$t"},作为":m"，在":[$$m.k"，$$m.v"]}},初始值":{}，在":{$let":{变量":{type_value":$$value"，ta":$$this"},在":{$let":{变量":{键":{$arrayElemAt":[$$ta"，0]},价值":{$arrayElemAt":[$$ta"，1]}},在":{$开关":{分支机构":[{案例":{$eq":[$$key",价值"]},然后":{$mergeObjects":[$$type_value"，{价值":$$价值"}]}},{案例":{$eq":[$$key",trait_type"]},然后":{$mergeObjects":[$$type_value"，{类型":{$replaceAll":{输入":{$toLower":$$value"},发现":"",替换":_"}}}]}}],默认":$$type_value"；}}}}}}}}}}}},{$查找":{来自":假人"，让":{特质":$特质"；},管道":[{$set":{特质":$$特质"}},{$解开":{路径":$traits"}},{$replaceRoot":{newRoot":$traits"}},{$组":{_id":$type"，值":{$push":$value"}}},{$set":{类型":$_id"}},{$项目":{_id":0}},{$replaceRoot":{新根":{$cond":[{$eq":[{$size":$values"},1]},{$arrayToObject":{$let":{变量":{对":[[$type",{$arrayElemAt":[$values"，0]}]]},输入":$$pair"}}},{$arrayToObject":{$let":{变量":{对":[[$type",$values"]]},输入":$$pair"}}}]}}}],作为":特性"；}},{$set":{特性":{$mergeObjects":$traits"}}}])

I have the following example set of data

[{
  "_id": {
    "$oid": "60f83d3cd66842301905aa77"
  },
  "id": 527438,
  "name": "CryptoPunk #4050",
  "asset_contract": {
    "name": "CryptoPunks",
    "address": "0xb47e3cd837ddf8e4c57f05d70ab865de6e193bbb"
  },
  "traits": [
    {
      "trait_type": "type",
      "value": "Male",
      "display_type": null,
      "max_value": null,
      "trait_count": 6039,
      "order": null
    },
    {
      "trait_type": "accessory",
      "value": "Mohawk",
      "display_type": null,
      "max_value": null,
      "trait_count": 441,
      "order": null
    },
    {
      "trait_type": "accessory",
      "value": "Earring",
      "display_type": null,
      "max_value": null,
      "trait_count": 2459,
      "order": null
    },
    {
      "trait_type": "accessory",
      "value": "Frown",
      "display_type": null,
      "max_value": null,
      "trait_count": 261,
      "order": null
    }
  ],
  "token_id": "4050",
  "permalink": "https://opensea.io/assets/0xb47e3cd837ddf8e4c57f05d70ab865de6e193bbb/4050",
  "background_color": null,
  "image_url": "https://lh3.googleusercontent.com/sO18rDQYhC5yIcj12RVsv31pbbsZo_2muQQbTJMQHn47EKGhnirs8mxzohm58HAZ7taBoe4pU6x1qntlExk_TtJ-",
  "image_preview_url": "https://lh3.googleusercontent.com/sO18rDQYhC5yIcj12RVsv31pbbsZo_2muQQbTJMQHn47EKGhnirs8mxzohm58HAZ7taBoe4pU6x1qntlExk_TtJ-=s250",
  "animation_url": null,
  "vault_contract": "0x269616d549d7e8eaa82dfb17028d0b212d11232a"
},{
  "_id": { "$oid": "60f83d3cbc3f0161da2141f7" },
  "id": 17736625,
  "name": "OJ Simpson",
  "asset_contract": {
    "name": "Hashmasks",
    "address": "0xc2c747e0f7004f9e8817db2ca4997657a7746928"
  },
  "traits": [
    {
      "trait_type": "Character",
      "value": "Male",
      "display_type": null,
      "max_value": null,
      "trait_count": 8659,
      "order": null
    },
    {
      "trait_type": "Mask",
      "value": "Doodle",
      "display_type": null,
      "max_value": null,
      "trait_count": 2187,
      "order": null
    },
    {
      "trait_type": "Eye Color",
      "value": "Dark",
      "display_type": null,
      "max_value": null,
      "trait_count": 7419,
      "order": null
    },
    {
      "trait_type": "Item",
      "value": "No Item",
      "display_type": null,
      "max_value": null,
      "trait_count": 14533,
      "order": null
    },
    {
      "trait_type": "Skin Color",
      "value": "Dark",
      "display_type": null,
      "max_value": null,
      "trait_count": 3784,
      "order": null
    },
    {
      "trait_type": "Token ID",
      "value": 3535,
      "display_type": "number",
      "max_value": null,
      "trait_count": 0,
      "order": null
    },
    {
      "trait_type": "Background",
      "value": "Doodle",
      "display_type": null,
      "max_value": null,
      "trait_count": 5538,
      "order": null
    }
  ],
  "token_id": "3535",
  "permalink": "https://opensea.io/assets/0xc2c747e0f7004f9e8817db2ca4997657a7746928/3535",
  "background_color": null,
  "image_url": "https://lh3.googleusercontent.com/NZQu7CNjgJ_1uhbUVwEb-14rZPJmPCaqaXy0qnUpgm5Qll0BvmmF7tPMjBhFH6ZZp_qzOPxHi0NFmRkOjHoBQ0BODcWI8NlyBXLu",
  "image_preview_url": "https://lh3.googleusercontent.com/NZQu7CNjgJ_1uhbUVwEb-14rZPJmPCaqaXy0qnUpgm5Qll0BvmmF7tPMjBhFH6ZZp_qzOPxHi0NFmRkOjHoBQ0BODcWI8NlyBXLu=s250",
  "animation_url": null,
  "vault_contract": "0xc7a8b45e184138114e6085c82936a8db93dd156a"
}]

which I would like to be updated to

[{
    "_id": {
      "$oid": "60f83d3cd66842301905aa77"
    },
    "id": 527438,
    "name": "CryptoPunk #4050",
    "asset_contract": {
      "name": "CryptoPunks",
      "address": "0xb47e3cd837ddf8e4c57f05d70ab865de6e193bbb"
    },
    "traits":
      {
        "type": "Male",
        "accessory": ["Mohawk", "Earing", "Frown"], 
      },
    "token_id": "4050",
    "permalink": "https://opensea.io/assets/0xb47e3cd837ddf8e4c57f05d70ab865de6e193bbb/4050",
    "background_color": null,
    "image_url": "https://lh3.googleusercontent.com/sO18rDQYhC5yIcj12RVsv31pbbsZo_2muQQbTJMQHn47EKGhnirs8mxzohm58HAZ7taBoe4pU6x1qntlExk_TtJ-",
    "image_preview_url": "https://lh3.googleusercontent.com/sO18rDQYhC5yIcj12RVsv31pbbsZo_2muQQbTJMQHn47EKGhnirs8mxzohm58HAZ7taBoe4pU6x1qntlExk_TtJ-=s250",
    "animation_url": null,
    "vault_contract": "0x269616d549d7e8eaa82dfb17028d0b212d11232a"
  },{
  "_id": { "$oid": "60f83d3cbc3f0161da2141f7" },
  "id": 17736625,
  "name": "OJ Simpson",
  "asset_contract": {
    "name": "Hashmasks",
    "address": "0xc2c747e0f7004f9e8817db2ca4997657a7746928"
  },
  "traits": {
      "character": "Male",
      "mask": "Doodle",
      "eye_color": "Dark",
      "item": "No Item",
      "skin_color": "Dark",
      "token_id": 3535,
      "background": "Doodle",
    },
  "token_id": "3535",
  "permalink": "https://opensea.io/assets/0xc2c747e0f7004f9e8817db2ca4997657a7746928/3535",
  "background_color": null,
  "image_url": "https://lh3.googleusercontent.com/NZQu7CNjgJ_1uhbUVwEb-14rZPJmPCaqaXy0qnUpgm5Qll0BvmmF7tPMjBhFH6ZZp_qzOPxHi0NFmRkOjHoBQ0BODcWI8NlyBXLu",
  "image_preview_url": "https://lh3.googleusercontent.com/NZQu7CNjgJ_1uhbUVwEb-14rZPJmPCaqaXy0qnUpgm5Qll0BvmmF7tPMjBhFH6ZZp_qzOPxHi0NFmRkOjHoBQ0BODcWI8NlyBXLu=s250",
  "animation_url": null,
  "vault_contract": "0xc7a8b45e184138114e6085c82936a8db93dd156a"
}]

The logic behind it would be

  • Look at the Traits array objects
  • get the trait_type value and create a new key using the lower case name (underscores for spaces)
  • Set the value of the new key to be the value of "value"

So,

      "trait_type": "type",
      "value": "Male",

//becomes

            "type": "Male"

  • Where there are multiple instances of the same trait type, create an array of values.

So,

{
      "trait_type": "accessory",
      "value": "Mohawk",
      "display_type": null,
      "max_value": null,
      "trait_count": 441,
      "order": null
    },
    {
      "trait_type": "accessory",
      "value": "Earring",
      "display_type": null,
      "max_value": null,
      "trait_count": 2459,
      "order": null
    },

// becomes

            "accessory": ["Mohawk", "Earring"]

解决方案

Query

  • its an aggregation update even if pipeline doesn't allow us to use stages like group lookup etc that here is used. (you can use $out and replace collection after or $merge to replace documents(similar to update))

  • first map

    • for each trait(document member of traits), it makes it into array
      [["trait_type": "type"] ["value": "Male"] ["display_type": null] ...]
    • reduce on that array to contruct from them 1 document only
      {"type" "type","value" :"Male"} (does also that lowercase and "_")

  • Now traits its like

    "traits": [
  {
    "type": "type",
    "value": "Male"
  },
  {
    "type": "accessory",
    "value": "Mohawk"
  },
  {
    "type": "accessory",
    "value": "Earring"
  },
  {
    "type": "accessory",
    "value": "Frown"
  }
]

  • lookup with the dummy collection [{}] (we do that to make a group inside that array) its like a trick that allows us to use stage operators inside 1 document

    • lookup pipeline unwinds and groups by type

    "traits": [
  {
    "values": [
      "Mohawk",
      "Earring",
      "Frown"
    ],
    "type": "accessory"
  },
  {
    "values": [
      "Male"
    ],
    "type": "type"
  }
]

    • then its a replace root to do take the value of type, make it the field-name and the values as value (if size=1 removes the array)

  • After lookup we have

    "traits": [
  {
    "accessory": [
      "Mohawk",
      "Earring",
      "Frown"
    ]
  },
  {
    "type": "Male"
  }
]

  • so all we have to do is to reduce that traits and merge the objects (keys are unique anyways because we grouped by them)

  • and we get the expected output (at least i think its ok)

Test code here

db.collection.aggregate([
  {
    "$set": {
      "traits": {
        "$map": {
          "input": "$traits",
          "as": "t",
          "in": {
            "$reduce": {
              "input": {
                "$map": {
                  "input": {
                    "$objectToArray": "$$t"
                  },
                  "as": "m",
                  "in": [
                    "$$m.k",
                    "$$m.v"
                  ]
                }
              },
              "initialValue": {},
              "in": {
                "$let": {
                  "vars": {
                    "type_value": "$$value",
                    "ta": "$$this"
                  },
                  "in": {
                    "$let": {
                      "vars": {
                        "key": {
                          "$arrayElemAt": [
                            "$$ta",
                            0
                          ]
                        },
                        "value": {
                          "$arrayElemAt": [
                            "$$ta",
                            1
                          ]
                        }
                      },
                      "in": {
                        "$switch": {
                          "branches": [
                            {
                              "case": {
                                "$eq": [
                                  "$$key",
                                  "value"
                                ]
                              },
                              "then": {
                                "$mergeObjects": [
                                  "$$type_value",
                                  {
                                    "value": "$$value"
                                  }
                                ]
                              }
                            },
                            {
                              "case": {
                                "$eq": [
                                  "$$key",
                                  "trait_type"
                                ]
                              },
                              "then": {
                                "$mergeObjects": [
                                  "$$type_value",
                                  {
                                    "type": {
                                      "$replaceAll": {
                                        "input": {
                                          "$toLower": "$$value"
                                        },
                                        "find": " ",
                                        "replacement": "_"
                                      }
                                    }
                                  }
                                ]
                              }
                            }
                          ],
                          "default": "$$type_value"
                        }
                      }
                    }
                  }
                }
              }
            }
          }
        }
      }
    }
  },
  {
    "$lookup": {
      "from": "dummy",
      "let": {
        "traits": "$traits"
      },
      "pipeline": [
        {
          "$set": {
            "traits": "$$traits"
          }
        },
        {
          "$unwind": {
            "path": "$traits"
          }
        },
        {
          "$replaceRoot": {
            "newRoot": "$traits"
          }
        },
        {
          "$group": {
            "_id": "$type",
            "values": {
              "$push": "$value"
            }
          }
        },
        {
          "$set": {
            "type": "$_id"
          }
        },
        {
          "$project": {
            "_id": 0
          }
        },
        {
          "$replaceRoot": {
            "newRoot": {
              "$cond": [
                {
                  "$eq": [
                    {
                      "$size": "$values"
                    },
                    1
                  ]
                },
                {
                  "$arrayToObject": {
                    "$let": {
                      "vars": {
                        "pair": [
                          [
                            "$type",
                            {
                              "$arrayElemAt": [
                                "$values",
                                0
                              ]
                            }
                          ]
                        ]
                      },
                      "in": "$$pair"
                    }
                  }
                },
                {
                  "$arrayToObject": {
                    "$let": {
                      "vars": {
                        "pair": [
                          [
                            "$type",
                            "$values"
                          ]
                        ]
                      },
                      "in": "$$pair"
                    }
                  }
                }
              ]
            }
          }
        }
      ],
      "as": "traits"
    }
  },
  {
    "$set": {
      "traits": {
        "$mergeObjects": "$traits"
      }
    }
  }
])

这篇关于MongoDb 聚合数据操作 - 对象到数组的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！

查看全文
相关文章
其他开发最新文章
热门教程
热门工具
登录 关闭
扫码关注1秒登录
发送“验证码”获取 | 15天全站免登陆