如何为文档流水线

Question

我收集了这样的文档：

{ "_id" : ObjectId("5b91035eca4f00000124e1d8"), "status" : 0, "date" : NumberLong(0), "players" : [ { "name" : "Valentin", "sets" : [ { "points" : [ { "score" : 0, "comment" : "" }, { "score" : 0, "comment" : "" }, { "score" : 0, "comment" : "" }, { "score" : 1, "comment" : "sa" }, { "score" : 1, "comment" : "" }, { "score" : 1, "comment" : "" }, { "score" : 1, "comment" : "" }, { "score" : 1, "comment" : "" }, { "score" : 1, "comment" : "" }, { "score" : 1, "comment" : "" }, { "score" : 2, "comment" : "" }, { "score" : 2, "comment" : "" }, { "score" : 2, "comment" : "" }, { "score" : 2, "comment" : "" }, { "score" : 3,"comment" : "sw" }, { "score" : 3, "comment" : "" }, { "score" : 3, "comment" : "" }, { "score" : 4, "comment" : "sa" }, { "score" : 5, "comment" : "bh" }, { "score" : 5, "comment" : "" }, { "score" : 5, "comment" : "" }, { "score" : 6, "comment" : "fh" }, { "score" : 6, "comment" : "" }, { "score" : 7, "comment" : "sw" }, { "score" : 7, "comment" : "" }, { "score" : 7, "comment" : "" }, { "score" : 8, "comment" : "" }, { "score" : 8, "comment" : "" }, { "score" : 9, "comment" : "bh" }, { "score" : 10, "comment" : "sw" }, { "score" : 11, "comment" : "sw" } ] } ] }, { "name" : "Zalupkin", "sets" : [ { "points" : [ { "score" : 0, "comment" : "" }, { "score" : 1, "comment" : "" }, { "score" : 2, "comment" : "" }, { "score" : 2, "comment" : "" }, { "score" : 3, "comment" : "fh"}, { "score" : 4, "comment" : "sw" }, { "score" : 5, "comment" : "" }, { "score" : 6, "comment" : "" }, { "score" : 7, "comment" : "" }, { "score" : 8, "comment" : "sa" }, { "score" : 8, "comment" : "" }, { "score" : 9, "comment" : "fh" }, { "score" : 10, "comment" : "sa" }, { "score" : 11, "comment" : "bh" }, {"score" : 11, "comment" : "" }, { "score" : 12, "comment" : "" }, { "score" : 13, "comment" : "" }, { "score" : 13, "comment" : "" }, { "score" : 13, "comment" : "" }, { "score" : 14, "comment" : "" }, { "score" : 15, "comment" : "sa" }, { "score" : 15, "comment" : "" }, { "score" : 16, "comment" : "sw" }, { "score" : 16, "comment" : "" }, { "score" : 17, "comment" : "" }, { "score" : 18, "comment" : "" }, { "score" : 18, "comment" : "" }, { "score" : 19, "comment" : "sa" }, { "score" : 19, "comment" : "" }, { "score" : 19, "comment" : "" }, { "score" : 19, "comment" : "" } ] } ] } ] }

我尝试制作聚合管道来计算鬃毛评论“sw”有多少玩家名称为“Valentin”。

它也可以在许多文档中，我需要计算整个 cmets。

结果必须是这样的： { "_id" : null, "count" : 0 }

感谢您的帮助。

Answer 1

也许这行得通：

db.getCollection('test2').aggregate([
// Target only the documents which contains player Valentin

{ $match: {"players.name" : "Valentin"}},

//Unwind the array players so we can remove the part from the other players
{ $unwind: "$players"},

//Get the players whose name is Valentin of the unwind set
{ $match: {"players.name" : "Valentin"}},

//Reduce the deep of the array so we can count them well
{ $project: {"points": "$players.sets.points"}},
{ $unwind: "$points"},
{ $unwind: "$points"},

//Query of comment match
{ $match: {"points.comment" : "sw" }},

//Calculate count
{ $group: {
    _id: null,
    count: { $sum: 1 }
    }}
])

Answer 2

您可以尝试以下聚合。

初始$filter 过滤匹配的玩家和 下一个$filter 过滤comment = 'sw' 的点，然后$size 计算每个点数组中的匹配。使用$map 执行所有集合的逻辑，使用$sum 计算所有集合的所有匹配项

$group 计算所有文档中的匹配项。

db.colname.aggregate([
  {"$match":{"players.name":"Valentin"}},
  {"$project":{
    "count":{
      "$sum":{
        "$map":{
          "input":{
            "$let":{
              "vars":{
                "mplayer":{
                  "$filter":{"input":"$players","as":"player","cond":{"$eq":["$$player.name","Valentin"]}
                 }
                }
              },
              "in":{"$arrayElemAt":["$$mplayer.sets",0]}
            }
          },
          "as":"set",
          "in":{
            "$size":{
              "$filter":{"input":"$$set.points","cond":{"$eq":["$$this.comment","sw"]}}
            }
          }
        }
      }
    }
  }},
  {"$group":{"_id":null,"count":{"$sum":"$count"}}}
])