如何在 mongo 展开后获取之前的值

Question

是否可以在 $unwind 之后访问以前的文档，然后 $group 在新的（已清理）object 中将它们一起访问？

使用这个 Mongo aggregate: mongoplayground.net/p/IsM-h9DnhjJ 我想得到页面内所有结果 stats 和 scores 的平均值，然后网站所有页面统计数据的平均值。

Collections

我有三个 collections 都是相关的。 Site > Page > Result（页面结果）

// sites
{
  _id: "si-site-a",
  teamId: "te-team-1",
  name: "Site a"
}

// pages
{
  _id: "pa-page-a",
  teamId: "te-team-1",
  siteId: "si-site-a",
  name: "Page A",
  status: "ok"
},
{
  _id: "pa-page-b",
  teamId: "te-team-1",
  siteId: "si-site-a",
  name: "Page B",
  status: "ok"
}

// results
{
  _id: "re-result-1",
  pageId: "pa-page-a",
  siteId: "si-site-a",
  results: {
    scores: {
      performance: 0.75,
      accessibility: 0.98,
      "best-practices": 0.93,
      seo: 1,
      pwa: 0.36,
      ecoscore: 0.94
    },
    stats: {
      transferSize: 876995,
      co2grams: 0.11,
      pageLoadTime: 1722.284
    }
  },
  region: "gb1",
  timestamp: ISODate("2022-02-05T16:38:35.454Z"),
  status: "ok"
}

期望的输出

  {
    _id: 'a',
    name: 'Site a',
    pages: [
      {
        _id: 'aa',
        name: 'Page a',
        status: 'ok',
        resultAverages: {
          ecoscore: 0.89,
          transferSize: 556081,
          co2grams: 0.19,
          pageLoadTime: 1011.7159999999999,
        },
      },
    ],
    pageAverages: {
          ecoscore: 0.89,
          transferSize: 556081,
          co2grams: 0.19,
          pageLoadTime: 1011.7159999999999,
    },
    comparedTo: {
          ecoscore: 0.89,
          transferSize: 556081,
          co2grams: 0.19,
          pageLoadTime: 1011.7159999999999,
    },
    percentageChange: {
          ecoscore: 0,
          transferSize: 0,
          co2grams: -0.19,
          pageLoadTime: 0.24,
    },

  },

Answer 1

您实际上在使用 sub-pipeline 的正确轨道上。您可以先在 sub-pipeline 中计算结果中的平均值。然后将它们存储在个人 result 中。在 sub-pipeline 之后，执行另一个 $avg 来计算之前的内部结果的平均值，以获得 pages.

的平均值

db.sites.aggregate([
  {
    $lookup: {
      from: "pages",
      localField: "_id",
      foreignField: "siteId",
      as: "pages",
      pipeline: [
        {
          $lookup: {
            from: "result",
            localField: "_id",
            foreignField: "pageId",
            as: "results",
            pipeline: [
              {
                $match: {
                  timestamp: {
                    $gte: ISODate("2022-01-01"),
                    $lte: ISODate("2022-12-01")
                  },
                  status: "ok"
                }
              }
            ]
          }
        },
        {
          $project: {
            _id: 1,
            name: 1,
            status: 1,
            resultAverages: {
              ecoscore: {
                $avg: "$results.results.stats.ecosore"
              },
              transferSize: {
                $avg: "$results.results.stats.transferSize"
              },
              co2grams: {
                $avg: "$results.results.stats.co2grams"
              },
              pageLoadTime: {
                $avg: "$results.results.stats.pageLoadTime"
              }
            }
          }
        }
      ]
    }
  },
  {
    "$addFields": {
      pageAverages: {
        ecoscore: {
          $avg: "$pages.resultAverages.ecosore"
        },
        transferSize: {
          $avg: "$pages.resultAverages.transferSize"
        },
        co2grams: {
          $avg: "$pages.resultAverages.co2grams"
        },
        pageLoadTime: {
          $avg: "$pages.resultAverages.pageLoadTime"
        }
      }
    }
  }
])

这里是Mongo playground供您参考。