JIYIK CN >

This article will explore the number of ways to get the last N records in MongoDB, where N is a positive number greater than zero. We will see how to retrieve the number of documents with and without sorting.

Get the last N records in MongoDB

We must have a collection with documents to proceed with the actual example. So, let’s go ahead and create a collection and insert some documents into it using the following queries.

Sample code:

> use get_n_records;
> db.createCollection('stock');
> db.stock.insertMany([
    {"item": "abc", "quantity": 12},
    {"item": "def", "quantity": 234},
    {"item": "ghi", "quantity": 45},
    {"item": "jkl", "quantity": 345},
    {"item": "mno", "quantity": 243},
    {"item": "pqr", "quantity": 345},
    {"item": "stu", "quantity": 56},
    {"item": "vwx", "quantity": 575},
    {"item": "yzz", "quantity": 398}
]);
> db.stock.find();

Output:

{ "_id" : ObjectId("629ac651b7e637e64d8e75fb"), "item" : "abc", "quantity" : 12 }
{ "_id" : ObjectId("629ac651b7e637e64d8e75fc"), "item" : "def", "quantity" : 234 }
{ "_id" : ObjectId("629ac651b7e637e64d8e75fd"), "item" : "ghi", "quantity" : 45 }
{ "_id" : ObjectId("629ac651b7e637e64d8e75fe"), "item" : "jkl", "quantity" : 345 }
{ "_id" : ObjectId("629ac651b7e637e64d8e75ff"), "item" : "mno", "quantity" : 243 }
{ "_id" : ObjectId("629ac651b7e637e64d8e7600"), "item" : "pqr", "quantity" : 345 }
{ "_id" : ObjectId("629ac651b7e637e64d8e7601"), "item" : "stu", "quantity" : 56 }
{ "_id" : ObjectId("629ac651b7e637e64d8e7602"), "item" : "vwx", "quantity" : 575 }
{ "_id" : ObjectId("629ac651b7e637e64d8e7603"), "item" : "yzz", "quantity" : 398 }

Now, we have our collection ready and populated. The key thing to consider is whether we want to get the last N records from the most recently inserted document to the least recently inserted document or vice versa.

Let us study each of the scenarios given below.

Get the last N records from the most recently inserted document to the most recently inserted document and vice versa

Before we proceed, let’s review these two terms to understand them easily:

Use cursor.sort()method to get last N records from most recently inserted document to least recently inserted document and vice versa

Sample code (LIFO):

> db.stock.find().sort({_id:-1}).limit(3);

Output:

{ "_id" : ObjectId("629ac651b7e637e64d8e7603"), "item" : "yzz", "quantity" : 398 }
{ "_id" : ObjectId("629ac651b7e637e64d8e7602"), "item" : "vwx", "quantity" : 575 }
{ "_id" : ObjectId("629ac651b7e637e64d8e7601"), "item" : "stu", "quantity" : 56 }

Sample code (FIFO):

> db.stock.find().sort({_id:1}).limit(3);

Output:

{ "_id" : ObjectId("629ac651b7e637e64d8e75fb"), "item" : "abc", "quantity" : 12 }
{ "_id" : ObjectId("629ac651b7e637e64d8e75fc"), "item" : "def", "quantity" : 234 }
{ "_id" : ObjectId("629ac651b7e637e64d8e75fd"), "item" : "ghi", "quantity" : 45 }

Here, we use sort()the method to sort the documents from the most recently inserted to the most recent inserted order (descending). So, before fetching any document (record) from the database, we apply sort()the method to the cursor.

sort()The field value pairs of the method are {field: value}, we can perform sorting on up to 32 keys. When using MongoDB, documents are not stored in the collection in the specified order.

That is why when we perform a sort on a specific field with duplicate values, documents with these values ​​may be returned in any order. In order to get a consistent sort, we must include at least one field that contains unique values.

The easiest way to ensure this is sort()to use _idthe field in the method.

limit(N)Used to get a specified number of documents. If we are interested in getting all the documents in sorted order, we can omit this function.

Alternatively, we can also use the following query without specific fields.

Sample code:

> db.stock.find().sort({$natural:1}).limit(3);

Output:

{ "_id" : ObjectId("629ac651b7e637e64d8e75fb"), "item" : "abc", "quantity" : 12 }
{ "_id" : ObjectId("629ac651b7e637e64d8e75fc"), "item" : "def", "quantity" : 234 }
{ "_id" : ObjectId("629ac651b7e637e64d8e75fd"), "item" : "ghi", "quantity" : 45 }

If the values ​​of are and $naturalrespectively , they work like FIFO and LIFO. If we have MongoDB 4.4 or higher, we can use it.1-1

Use the $sortand $limit(aggregate) stage to get the last N records from the most recently inserted document to the most recently inserted document and vice versa

Sample code (LIFO):

> db.stock.aggregate([
     { $sort : { _id : -1 } },
     { $limit : 4 }
]);

Output:

{ "_id" : ObjectId("629ac651b7e637e64d8e7603"), "item" : "yzz", "quantity" : 398 }
{ "_id" : ObjectId("629ac651b7e637e64d8e7602"), "item" : "vwx", "quantity" : 575 }
{ "_id" : ObjectId("629ac651b7e637e64d8e7601"), "item" : "stu", "quantity" : 56 }
{ "_id" : ObjectId("629ac651b7e637e64d8e7600"), "item" : "pqr", "quantity" : 345 }

Sample code (FIFO):

> db.stock.aggregate([
     { $sort : { _id : 1 } },
     { $limit : 4 }
]);

Output:

{ "_id" : ObjectId("629ac651b7e637e64d8e75fb"), "item" : "abc", "quantity" : 12 }
{ "_id" : ObjectId("629ac651b7e637e64d8e75fc"), "item" : "def", "quantity" : 234 }
{ "_id" : ObjectId("629ac651b7e637e64d8e75fd"), "item" : "ghi", "quantity" : 45 }
{ "_id" : ObjectId("629ac651b7e637e64d8e75fe"), "item" : "jkl", "quantity" : 345 }

$sortThe aggregation stage is similar to cursor.sort()the method. It is used to sort all the documents and return them to the pipeline in sorted order.

It requires a document with field names and the corresponding sort order. Remember that 1and -1are used for ascending and descending order.

As with cursor.sort()the method, we can use $sortthe aggregate stage to sort on up to 32 keys, with at least one field (containing unique values) to achieve a consistent sort order.

If we need to sort on multiple fields, the sort order is evaluated from left to right. $limittakes a number showing how many documents need to be printed on the screen; furthermore, it has a 64-bit integer limit.

Use the skip()and count()methods together to get the last N records in MongoDB

This approach is appealing because, given the project requirements, we are not sorting the data, but rather getting a segment of the document from the beginning or the end.

Sample code (get the first 3 records without sorting the documents):

> db.stock.find().limit(3);

Output:

{ "_id" : ObjectId("629ac651b7e637e64d8e75fb"), "item" : "abc", "quantity" : 12 }
{ "_id" : ObjectId("629ac651b7e637e64d8e75fc"), "item" : "def", "quantity" : 234 }
{ "_id" : ObjectId("629ac651b7e637e64d8e75fd"), "item" : "ghi", "quantity" : 45 }

Sample code (get the last 3 records without sorting the documents):

> db.stock.find().skip(db.stock.count() - 3)

Output:

{ "_id" : ObjectId("629ac651b7e637e64d8e7601"), "item" : "stu", "quantity" : 56 }
{ "_id" : ObjectId("629ac651b7e637e64d8e7602"), "item" : "vwx", "quantity" : 575 }
{ "_id" : ObjectId("629ac651b7e637e64d8e7603"), "item" : "yzz", "quantity" : 398 }

In this example, we use db.stock.count()to count the total number of documents, subtract 3 from it and use skip()the method to skip all documents.

Remember, we subtracted the number of documents that were not skipped (3 in this case). We will end up with those 3 documents.