我的文档结构如下:
_id: 5d090529397f79d2fefc57fb,
result: 23232,
temperatures: {
temp_limits: {
min: 10,
max: 31,
},
temp_cities:
[
{
city: 'citie_1',
city_temp: [
{ temp: 17, date: 2018-12-18T10:35:07.000Z}, // I want this
{ temp: 21, date: 2018-12-17T11:35:05.000Z},
{ temp: 23, date: 2018-12-17T14:36:07.000Z},
],
locked: false
},
{
city: 'citie_2',
city_temp: [
{ temp: 15, date: 2018-12-18T14:15:07.000Z}, // and this
{ temp: 22, date: 2018-12-17T11:33:02.000Z}
],
locked: false
}
]
}
我想要每个date
上的最大值temp
和他的city_temp
,但保持相同的结构。像这样:
_id: 5d090529397f79d2fefc57fb,
result: 23232,
temperatures: {
temp_limits: {
min: 10,
max: 31,
},
temp_cities:
[
{
city: 'citie_1',
city_temp: { temp: 17, date: 2018-12-18T10:35:07.000Z}
locked: false
},
{
city: 'citie_2',
city_temp: { temp: 15, date: 2018-12-18T14:15:07.000Z}
locked: false
}
]
}
我尝试了此操作,但结果分离或结构不同:
{$unwind: '$temperatures.temp_cities'},
{$unwind: '$temperatures.temp_cities.city_temp'},
{$sort: { '$temperatures.temp_cities.city_temp.date': -1 } },
{
$group: {
_id: {
_id: '$_id',
bookie: "$temperatures.temp_cities.city"
},
result: { $first: '$result' },
temperatures: {
$first: "$temperatures.temp_cities.city_temp"
}
}
}
答案 0 :(得分:1)
您可以对嵌套的$map使用$filter运算符。最长日期可以使用$max运算符获取:
db.collection.aggregate([
{
$addFields: {
"temperatures.temp_cities": {
$map: {
input: "$temperatures.temp_cities",
as: "tc",
in: {
locked: "$$tc.locked",
city: "$$tc.city",
city_temp: {
$let: {
vars: { maxDate: { $max: "$$tc.city_temp.date" } },
in: {
$arrayElemAt: [
{ $filter: { input: "$$tc.city_temp", cond: { $eq: [ "$$this.date", "$$maxDate" ] } } }, 0
]
}
}
}
}
}
}
}
}
])