有条件地将键添加到 MongoDB 数据库，其值基于其他条目

Question

我有一个 MongoDB 数据库，其中数据结构多年来发生了变化。数据示例如下：

{ entry_id: 111, entrant_company:"Magic", year:2021, credit_name:"Bob Jones", credit_role:"Designer"}
{ entry_id: 111, entrant_company:"Magic", year:2021, credit_name:"Allan Jeggs", credit_role:"Director"}
{ entry_id: 111, entrant_company:"Magic", year:2021, credit_name:"Sarah Jones", credit_role:"Artist"}
{ entry_id: 348, year:2022, credit_name:"Glitter", credit_role:"Advertising Agency"}
{ entry_id: 348, year:2022, credit_name:"Jim Jones", credit_role:"Designer"}
{ entry_id: 348, year:2022, credit_name:"Allan Yi", credit_role:"Publisher"}
{ entry_id: 555, year:2023, credit_name:"Sparkle", credit_role:"Advertising Agency"}
{ entry_id: 555, year:2023, credit_name:"Josh James", credit_role:"Designer"}
{ entry_id: 555, year:2023, credit_name:"Ellen Dodd", credit_role:"Publisher"}
...

对于 2021 年以上的年份，我缺少

entrant_company

键，而是在

credit_role

等于“广告代理机构”时提供此信息。

按

entry_id

对我的数据进行分组，如何添加名为

entrant_company

的键并将其设置为等于

credit_name

，其中

credit_role

是“广告代理机构”并且

year

大于 2021 ？

为了清楚起见，我的目标是让数据看起来像这样：

{ entry_id: 111, entrant_company:"Magic", year:2021, credit_name:"Bob Jones", credit_role:"Designer"}
{ entry_id: 111, entrant_company:"Magic", year:2021, credit_name:"Allan Jeggs", credit_role:"Director"}
{ entry_id: 111, entrant_company:"Magic", year:2021, credit_name:"Sarah Jones", credit_role:"Artist"}
{ entry_id: 348, entrant_company:"Glitter", year:2022, credit_name:"Glitter", credit_role:"Advertising Agency"}
{ entry_id: 348, entrant_company:"Glitter", year:2022, credit_name:"Jim Jones", credit_role:"Designer"}
{ entry_id: 348, entrant_company:"Glitter", year:2022, credit_name:"Allan Yi", credit_role:"Publisher"}
{ entry_id: 555, entrant_company:"Sparkle", year:2023, credit_name:"Sparkle", credit_role:"Advertising Agency"}
{ entry_id: 555, entrant_company:"Sparkle", year:2023, credit_name:"Josh James", credit_role:"Designer"}
{ entry_id: 555, entrant_company:"Sparkle", year:2023, credit_name:"Ellen Dodd", credit_role:"Publisher"}

我可以像这样更新单个条目，但我想为每个给定的

entrant_company

更新

entry_id

:

db.entries.updateMany(
    { year: { $gt: 2021 } },
    [{
        $set: {
            entrant_company: {
                $cond: {
                    if: { $eq: ["$credit_role", "Advertising Agency"] },
                    then: "$credit_name",
                    else: "$$REMOVE"
                }
            }
        }
    }])

Answer 1

您需要通过 entry_id 加入基于

credit_role: "Advertising Agency"

和

$lookup

的文档的

entries

集合。但是，使用

$lookup

聚合阶段无法在

.update()

操作中执行。

因此，您需要切换到

.aggregate()

查询和

$merge

阶段来替换文档。

db.entries.aggregate([
  {
    $match: {
      $expr: {
        $and: [
          {
            $gt: [
              "$year",
              2021
            ]
          },
          {
            $eq: [
              {
                $type: "$entrant_company"
              },
              "missing"
            ]
          }
        ]
      }
    }
  },
  {
    $lookup: {
      from: "entries",
      let: {
        entry_id: "$entry_id"
      },
      pipeline: [
        {
          $match: {
            $expr: {
              $and: [
                {
                  $eq: [
                    "$entry_id",
                    "$$entry_id"
                  ]
                },
                {
                  $eq: [
                    "$credit_role",
                    "Advertising Agency"
                  ]
                }
              ]
            }
          }
        }
      ],
      as: "joined_entrant_company"
    }
  },
  {
    $set: {
      entrant_company: {
        $first: "$joined_entrant_company.credit_name"
      }
    }
  },
  {
    $unset: "joined_entrant_company"
  },
  {
    $merge: {
      into: "entries",
      on: "_id",
      whenMatched: "replace"
    }
  }
])

演示@Mongo Playground