我试图从Mongo中获取“组”的每个最新条目。
您可以将数据像一组Wiki条目一样成像:
要更详细:
itemId
itemId
在数据集中可以出现1到n次orderHash
在这里您可以看到少量数据:
{
"_id": ObjectId("5f69b963b8705200282d6174"),
"itemId": "xszy",
"orderHash": "383u",
"title": "A",
"content": "ABC"
},
{
"_id": ObjectId("5f69b963b8705200282d6175"),
"itemId": "e92q",
"orderHash": "5j12",
"title": "K",
"content": "KLMN"
},
{
"_id": ObjectId("5f69b963b8705200282d6178"),
"itemId": "xszy",
"orderHash": "p578",
"title": "A",
"content": "ABD"
},
{
"_id": ObjectId("5f69b963b8705200282d6180"),
"itemId": "mtewy",
"orderHash": "383u",
"title": "L",
"content": "CASE"
},
{
"_id": ObjectId("5f69b963b8705200282d6189"),
"itemId": "mtewy",
"orderHash": "5j12",
"title": "L1",
"content": "CASE"
}
我使用以下聚合管道来获得所需的结果:
db.getCollection('wiki').aggregate([
{
// create a sortable field for mongo
$set: {
"sortField": {
"$indexOfArray": [
[ "p578", "5j12", "383u" ], "$orderHash"
]
}
}
},
{
// sort by created sort field
"$sort": {
"sortField": 1
}
},
{
// group items by itemId and "save" them sorted in an array
$group: {
_id: "$itemId",
data: {
$push: "$$ROOT"
}
}
},
{
// get first entry of each "group array" to obtain
// the latest version of each entry
$project: {
resp: { $arrayElemAt: ['$data', 0] }
}
}
])
到目前为止,这是可行的。但是将数据存储在额外的数组中感觉很不对劲,我不确定性能(尤其是对于具有大量编辑的较大数据集)。另一个缺陷是,我不能简单地获取文档数组,而是将数据嵌套在文档属性resp
中(不适用于例如猫鼬)。
我的问题是否还有更多的蒙古解决方案?
解决方案:
感谢乔(请参阅评论)。他为我指出了解决方案。
db.getCollection('wiki').aggregate([
{
$set: {
"sortField": {
"$indexOfArray": [
[ "p578", "5j12", "383u" ], "$orderHash"
]
}
}
},
{
"$sort": {
"sortField": 1
}
},
{
$group: {
_id: "$itemId",
resp: {
$first: "$$ROOT"
}
}
},
{
$replaceRoot: {
newRoot: "$resp"
}
}
])