Removing Documents: MongoDB

Lets learn how to remove documents from the collection using remove() and drop() methods.

remove drop mongodb

test database, names collection

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
> db.names.find().pretty()
{
        "_id" : ObjectId("53c6392a2eea8062e084cb57"),
        "Company" : "Google",
        "Product" : "Nexus",
        "No" : 1
}
{
        "_id" : ObjectId("53c639392eea8062e084cb58"),
        "Company" : "Apple",
        "Product" : "Mac",
        "No" : 2
}
{
        "_id" : ObjectId("53c63b26b003603dfdcf8c52"),
        "Company" : "Xiaomi",
        "Product" : "Mi3",
        "No" : 3
}
{
        "_id" : ObjectId("53c63bd1b003603dfdcf8c53"),
        "Product" : "Smart Watch",
        "No" : 4,
        "Company" : "Sony"
}

We have 4 documents in “names” collection.

remove() method, with simple condition

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
> db.names.remove({"No": 4});
WriteResult({ "nRemoved" : 1 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c6392a2eea8062e084cb57"),
        "Company" : "Google",
        "Product" : "Nexus",
        "No" : 1,
        "IT" : "true"
}
{
        "_id" : ObjectId("53c639392eea8062e084cb58"),
        "Company" : "Apple",
        "Product" : "Mac",
        "No" : 2,
        "IT" : "true"
}
{
        "_id" : ObjectId("53c63b26b003603dfdcf8c52"),
        "Company" : "Xiaomi",
        "Product" : "Mi3",
        "No" : 3,
        "IT" : "true"
}

Here the document with “No: 4” was removed from the collection.

Removing Documents: MongoDB


You need to a flashplayer enabled browser to view this YouTube video

YouTube Link: http://www.youtube.com/watch?v=so0_5Tdb_bQ [Watch the Video In Full Screen.]



remove() method, with comparison condition

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
> db.names.remove({"No": {$gt: 2}});
WriteResult({ "nRemoved" : 1 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c6392a2eea8062e084cb57"),
        "Company" : "Google",
        "Product" : "Nexus",
        "No" : 1,
        "IT" : "true"
}
{
        "_id" : ObjectId("53c639392eea8062e084cb58"),
        "Company" : "Apple",
        "Product" : "Mac",
        "No" : 2,
        "IT" : "true"
}

Here whatever documents which has “No” greater than 2 got removed.

remove() method to remove all documents from the collection

1
2
3
4
5
6
7
8
9
10
11
> db.names.remove({});
WriteResult({ "nRemoved" : 2 })
 
> db.names.find().pretty()
 
> show collections
names
system.indexes
 
> db.system.indexes.find().pretty()
{ "v" : 1, "key" : { "_id" : 1 }, "name" : "_id_", "ns" : "test.names" }

If we pass empty argument to remove() method, it matches with all the documents present in the collection, hence removes all the documents one-by-one.

But it doesn’t remove the contents/documents/index present in “system.indexes” collection.

drop() method

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
> db.names.find()
 
> db.names.insert({"Company": "Apple", "Product": "iPhone", "No": 1});
WriteResult({ "nInserted" : 1 })
 
> db.names.insert({"Company": "Google", "Product": "Nexus", "No": 2});
WriteResult({ "nInserted" : 1 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c6c5d95879b1ff1f0b8356"),
        "Company" : "Apple",
        "Product" : "iPhone",
        "No" : 1
}
{
        "_id" : ObjectId("53c6392a2eea8062e084cb57"),
        "Company" : "Google",
        "Product" : "Nexus",
        "No" : 1,
        "IT" : "true"
}
 
> db.names.drop();
true
 
> db.names.find().pretty()
 
> db.system.indexes.find().pretty()

Since “names” collection was empty, we inserted 2 documents into it. Now we applied drop() method on the collection, which drops all the document present in the collections at once. It also removes the document/index/content present inside “system.indexes” collection.

Note: If you want to remove/drop all the documents present inside the collection, make use of drop() method, as it removes all the documents at once, its more efficient than remove({}) method which removes documents one by one. Use remove() method, when you want to remove one or a set of documents from the collection.

Multi-Update: MongoDB

Lets learn how to update all the documents present in a collection using update() method, using the option multi: true

multi update mongodb

test database, names collection

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
> db.names.find().pretty()
{
        "_id" : ObjectId("53c6392a2eea8062e084cb57"),
        "Company" : "Google",
        "Product" : "Nexus",
        "No" : 1
}
{
        "_id" : ObjectId("53c639392eea8062e084cb58"),
        "Company" : "Apple",
        "Product" : "Mac",
        "No" : 2
}
{
        "_id" : ObjectId("53c63b26b003603dfdcf8c52"),
        "Company" : "Xiaomi",
        "Product" : "Mi3",
        "No" : 3
}
{
        "_id" : ObjectId("53c63bd1b003603dfdcf8c53"),
        "Product" : "Smart Watch",
        "No" : 4,
        "Company" : "Sony"
}

We have 4 documents in “names” collection.

find() method

1
2
3
4
5
> db.names.find({}, {"Company": 1, "_id": 0}).pretty()
{ "Company" : "Google" }
{ "Company" : "Apple" }
{ "Company" : "Xiaomi" }
{ "Company" : "Sony" }

If we leave the first argument of find() method empty, it matches with all the documents of the collection. Hence it fetched all the Company names from the documents.

Multi-Update: MongoDB


You need to a flashplayer enabled browser to view this YouTube video

YouTube Link: http://www.youtube.com/watch?v=yuVBs4hcYOQ [Watch the Video In Full Screen.]



update() method

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
> db.names.update({}, {$set: {"IT": "true"}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c6392a2eea8062e084cb57"),
        "Company" : "Google",
        "Product" : "Nexus",
        "No" : 1,
        "IT" : "true"
}
{
        "_id" : ObjectId("53c639392eea8062e084cb58"),
        "Company" : "Apple",
        "Product" : "Mac",
        "No" : 2
}
{
        "_id" : ObjectId("53c63b26b003603dfdcf8c52"),
        "Company" : "Xiaomi",
        "Product" : "Mi3",
        "No" : 3
}
{
        "_id" : ObjectId("53c63bd1b003603dfdcf8c53"),
        "Product" : "Smart Watch",
        "No" : 4,
        "Company" : "Sony"
}

But in case of update() method, if the first argument is left empty, it randomly matches to only 1 document in the collection.

Related Read: Update with SET Operator: MongoDB

unset the IT field

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
> db.names.update({"IT": {$exists: true}}, {$unset: {"IT": "true"}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c6392a2eea8062e084cb57"),
        "Company" : "Google",
        "Product" : "Nexus",
        "No" : 1
}
{
        "_id" : ObjectId("53c639392eea8062e084cb58"),
        "Company" : "Apple",
        "Product" : "Mac",
        "No" : 2
}
{
        "_id" : ObjectId("53c63b26b003603dfdcf8c52"),
        "Company" : "Xiaomi",
        "Product" : "Mi3",
        "No" : 3
}
{
        "_id" : ObjectId("53c63bd1b003603dfdcf8c53"),
        "Product" : "Smart Watch",
        "No" : 4,
        "Company" : "Sony"
}

Here we select the document which has a field called “IT” and remove it from that document.

Related Read:
$exists, $type, $regex operators: MongoDB
Update with UNSET Operator: MongoDB

multi-update true

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
> db.names.update({}, {$set: {"IT": "true"}}, {multi: true});
WriteResult({ "nMatched" : 4, "nUpserted" : 0, "nModified" : 4 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c6392a2eea8062e084cb57"),
        "Company" : "Google",
        "Product" : "Nexus",
        "No" : 1,
        "IT" : "true"
}
{
        "_id" : ObjectId("53c639392eea8062e084cb58"),
        "Company" : "Apple",
        "Product" : "Mac",
        "No" : 2,
        "IT" : "true"
}
{
        "_id" : ObjectId("53c63b26b003603dfdcf8c52"),
        "Company" : "Xiaomi",
        "Product" : "Mi3",
        "No" : 3,
        "IT" : "true"
}
{
        "_id" : ObjectId("53c63bd1b003603dfdcf8c53"),
        "Product" : "Smart Watch",
        "No" : 4,
        "Company" : "Sony",
        "IT" : "true"
}

Here we have 3 arguments for update() method. First one is intentionally left empty. In second argument, we specify the changes needed to the documents. In third argument we specify the option, multi: true – which tells mongo engine to match with all the documents in the collection.

Update with upsert: MongoDB

Lets learn to use upsert option with update() method.

upsert basically inserts document into the collection, if there is no matching document found in the collection. If there is a match found, then the document simply gets updated – in which case upsert option will not have any effect on the selected document.

update with upsert mongodb

test database, names collection

1
2
3
4
5
6
7
8
9
10
11
12
13
> db.names.find().pretty()
{
        "_id" : ObjectId("53c6392a2eea8062e084cb57"),
        "Company" : "Google",
        "Product" : "Nexus",
        "No" : 1
}
{
        "_id" : ObjectId("53c639392eea8062e084cb58"),
        "Company" : "Apple",
        "Product" : "iPhone",
        "No" : 2
}

There are 2 documents inside “names” collection, with _id, Company, Product and No as its fields.

Update with upsert: MongoDB


You need to a flashplayer enabled browser to view this YouTube video

YouTube Link: http://www.youtube.com/watch?v=ZQjNHjVv_W8 [Watch the Video In Full Screen.]



update() method, with $set operator

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
> db.names.update({"Company": "Apple"}, {$set: {"Product": "Mac"}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c6392a2eea8062e084cb57"),
        "Company" : "Google",
        "Product" : "Nexus",
        "No" : 1
}
{
        "_id" : ObjectId("53c639392eea8062e084cb58"),
        "Company" : "Apple",
        "Product" : "Mac",
        "No" : 2
}

We select second document using “Company” – “Apple“. Here, the “Product” field value gets updated from iPhone to Mac.

Related Read: Update with SET Operator: MongoDB

No Match Found for Updation

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
> db.names.update({"Company": "Xiaomi"}, {$set: {"Product": "Mi3", "No": 3}});
WriteResult({ "nMatched" : 0, "nUpserted" : 0, "nModified" : 0 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c6392a2eea8062e084cb57"),
        "Company" : "Google",
        "Product" : "Nexus",
        "No" : 1
}
{
        "_id" : ObjectId("53c639392eea8062e084cb58"),
        "Company" : "Apple",
        "Product" : "Mac",
        "No" : 2
}

Since, there is no matching “Company” called Xiaomi in “names” collection, there will be no matching document, hence no modification of document.

updation with upsert option

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
> db.names.update({"Company": "Xiaomi"}, 
                  {$set: {"Product": "Mi3", "No": 3}},
                  {upsert: true});
WriteResult({
        "nMatched" : 0,
        "nUpserted" : 1,
        "nModified" : 0,
        "_id" : ObjectId("53c63b26b003603dfdcf8c52")
})
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c6392a2eea8062e084cb57"),
        "Company" : "Google",
        "Product" : "Nexus",
        "No" : 1
}
{
        "_id" : ObjectId("53c639392eea8062e084cb58"),
        "Company" : "Apple",
        "Product" : "Mac",
        "No" : 2
}
{
        "_id" : ObjectId("53c63b26b003603dfdcf8c52"),
        "Company" : "Xiaomi",
        "Product" : "Mi3",
        "No" : 3
}

Observe the 3rd argument of update() method – upsert: true. In this case, since there is no matching document found inside the “names” collection, for “Company” Xiaomi, it inserts the document with all the fields and values specified in the update command. i.e., Company: Xiaomi, Product: Mi3, No: 3

unintentional modification of data with upsert

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
> db.names.update({"No": {$gt: 3}}, 
                  {$set: {"Product": "Glass", "No": 4}}, 
                  {upsert: true});
WriteResult({
        "nMatched" : 0,
        "nUpserted" : 1,
        "nModified" : 0,
        "_id" : ObjectId("53c63bd1b003603dfdcf8c53")
})
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c6392a2eea8062e084cb57"),
        "Company" : "Google",
        "Product" : "Nexus",
        "No" : 1
}
{
        "_id" : ObjectId("53c639392eea8062e084cb58"),
        "Company" : "Apple",
        "Product" : "Mac",
        "No" : 2
}
{
        "_id" : ObjectId("53c63b26b003603dfdcf8c52"),
        "Company" : "Xiaomi",
        "Product" : "Mi3",
        "No" : 3
}
{
        "_id" : ObjectId("53c63bd1b003603dfdcf8c53"),
        "Product" : "Glass",
        "No" : 4
}

We have documents which has values 1, 2 and 3 for the field “No”. But in above command the condition is to select documents which has “No” value above 3 – which results in 0 documents being selected. But since we have “upsert: true” as the third argument of update() method, the “field: value” – “Product: Glass” and “No: 4” gets inserted into the “names” collection. So we need to take proper care of our commands and conditions – so that unintentional insertion of data or document can be avoided.

upsert effect on existing data/document

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
> db.names.update({"No": 4}, 
                  {$set: {"Company": "Sony", "Product": "Smart Watch"}}, 
                  {upsert: true});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c6392a2eea8062e084cb57"),
        "Company" : "Google",
        "Product" : "Nexus",
        "No" : 1
}
{
        "_id" : ObjectId("53c639392eea8062e084cb58"),
        "Company" : "Apple",
        "Product" : "Mac",
        "No" : 2
}
{
        "_id" : ObjectId("53c63b26b003603dfdcf8c52"),
        "Company" : "Xiaomi",
        "Product" : "Mi3",
        "No" : 3
}
{
        "_id" : ObjectId("53c63bd1b003603dfdcf8c53"),
        "Product" : "Smart Watch",
        "No" : 4,
        "Company" : "Sony"
}

In this case, there is a document inside the collection which has “No”: 4. So the upsert option will not have any effect on the selected data/document – it simply gets updated with the field values specified in the second argument of update() method. i.e., the document with “No”: 4 is selected and the “Company”: “Sony”, “Product”: “Smart Watch” gets added to the existing document.

Working With Arrays: MongoDB

Lets learn some of the methods and operators to work with arrays in MongoDB.

In this video tutorial, we’ll be looking at:
update()
$set
$push
$pop
$pushAll
$pull
$pullAll
$addToSet

update set push pop pushAll pull pullAll addToSet operators mongodb

test database, names collection

1
2
3
4
5
6
7
8
9
MongoDB shell version: 2.6.1
connecting to: test
> db.names.find()
 
> db.names.insert({"_id": 1, "a": [1, 2, 3, 4]});
WriteResult({ "nInserted" : 1 })
 
> db.names.find()
{ "_id" : 1, "a" : [ 1, 2, 3, 4 ] }

We insert a document into “names” collection. We’ll be working on the array field present in the document.

Working With Arrays: MongoDB


You need to a flashplayer enabled browser to view this YouTube video

YouTube Link: http://www.youtube.com/watch?v=xc5TiSqvnPQ [Watch the Video In Full Screen.]



update() method

1
2
3
4
5
> db.names.update({"_id": 1}, {"a": [1, 2, 3, 4, 5]});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find()
{ "_id" : 1, "a" : [ 1, 2, 3, 4, 5 ] }

We can update the array by simply using update() method. But here, we need to remember all the elements of the array as well as the new element to be inserted into the array. Thus, this method is somewhat tedious.

$set operator

1
2
3
4
5
> db.names.update({"_id": 1}, {$set: {"a.5": 6}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find()
{ "_id" : 1, "a" : [ 1, 2, 3, 4, 5, 6 ] }

We could insert an element into the array by making use of $set operator. Here we need to know the position where the new element needs to be inserted. In mongoDB, array index starts from zero.

$push operator

1
2
3
4
5
> db.names.update({"_id": 1}, {$push: {"a": 7}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find()
{ "_id" : 1, "a" : [ 1, 2, 3, 4, 5, 6, 7 ] }

using $push operator we can insert an element to the right hand side of the array. Here we simply specify the key and the value/element to be inserted.

$pop: {a: 1}

1
2
3
4
5
> db.names.update({"_id": 1}, {$pop: {"a": 1}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find()
{ "_id" : 1, "a" : [ 1, 2, 3, 4, 5, 6 ] }

$pop operator which has 1 as value to the key, removes the right most element from the array.

$pop: {a: -1}

1
2
3
4
5
> db.names.update({"_id": 1}, {$pop: {"a": -1}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find()
{ "_id" : 1, "a" : [ 2, 3, 4, 5, 6 ] }

$pop operator which has -1 as value to the key, removes the left most element from the array.

$pushAll operator

1
2
3
4
5
> db.names.update({"_id": 1}, {$pushAll: {"a": [7, 8]}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find()
{ "_id" : 1, "a" : [ 2, 3, 4, 5, 6, 7, 8 ] }

$pushAll operator inserts all the elements(array of elements) specified, to the right hand side of the existing array.

$pull operator

1
2
3
4
5
> db.names.update({"_id": 1}, {$pull: {"a": 3}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find()
{ "_id" : 1, "a" : [ 2, 4, 5, 6, 7, 8 ] }

$pull operator pulls or removes the specified element from the array, irrespective of its position.

$pullAll operator

1
2
3
4
5
> db.names.update({"_id": 1}, {$pullAll: {"a": [2, 7, 8]}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find()
{ "_id" : 1, "a" : [ 4, 5, 6 ] }

$pullAll operator pulls/removes all the elements(array of elements) specified from the array, irrespective of its position.

$addToSet operator

1
2
3
4
5
> db.names.update({"_id": 1}, {$addToSet: {"a": 3}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find()
{ "_id" : 1, "a" : [ 4, 5, 6, 3 ] }

$addToSet operator adds specified element to the array, if its not already present in the array. If the element is already present in the array, then it doesn’t add it once again.

1
2
3
4
> db.names.update({"_id": 1}, {$addToSet: {"a": 3}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 0 })
> db.names.find()
{ "_id" : 1, "a" : [ 4, 5, 6, 3 ] }

If the element is already present in the array, then $addToSet doesn’t add it once again.

Increment($inc) operator: MongoDB

Lets learn how to use $inc operator with update() method in MongoDB.

inc operator with update

test database, names collection

1
2
3
4
5
6
7
8
9
10
11
12
> db.names.find().pretty()
{
        "_id" : ObjectId("53c3fc2005e0ce2719d91bd2"),
        "Company" : "Google",
        "No" : 1
}
{
        "_id" : ObjectId("53c3fd3bb9ae26fa217b1e12"),
        "Company" : "Apple",
        "Product" : "iPhone",
        "No" : 2
}

Here we have 2 documents. And we’ll be working on first document to illustrate the working of $inc operator.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
> db.names.update({"Company": "Google"}, {$inc: {"No": 2}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c3fc2005e0ce2719d91bd2"),
        "Company" : "Google",
        "No" : 3
}
{
        "_id" : ObjectId("53c3fd3bb9ae26fa217b1e12"),
        "Company" : "Apple",
        "Product" : "iPhone",
        "No" : 2
}

Here “No” field will be incremented by 2. So the final value of “No” field is 3 (1+2).

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
> db.names.update({"Company": "Google"}, {$inc: {"Sl_no": 1}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c3fc2005e0ce2719d91bd2"),
        "Company" : "Google",
        "No" : 3,
        "Sl_no" : 1
}
{
        "_id" : ObjectId("53c3fd3bb9ae26fa217b1e12"),
        "Company" : "Apple",
        "Product" : "iPhone",
        "No" : 2
}

If we apply $inc operator on an non-existing field, it will be created with the increment value itself.

Increment($inc) operator: MongoDB


You need to a flashplayer enabled browser to view this YouTube video

YouTube Link: http://www.youtube.com/watch?v=LjJJOY4DLMs [Watch the Video In Full Screen.]



1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
> db.names.update({"Company": "Google"}, {$inc: {"No": 1}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
> db.names.find().pretty()
{
        "_id" : ObjectId("53c3fc2005e0ce2719d91bd2"),
        "Company" : "Google",
        "No" : 4,
        "Sl_no" : 1
}
{
        "_id" : ObjectId("53c3fd3bb9ae26fa217b1e12"),
        "Company" : "Apple",
        "Product" : "iPhone",
        "No" : 2
}

Here we increment the value of “No” once again by 1. So the final value of “No” field is 4 (3+1).

Update with UNSET Operator: MongoDB

Lets use $unset operator along with update() method, to update the documents.

update with set operator mongodb Update with UNSET Operator: MongoDB

test database, names collection

1
2
3
4
5
6
7
8
9
10
11
12
13
> db.names.find().pretty()
{
        "_id" : ObjectId("53c3fc2005e0ce2719d91bd2"),
        "Company" : "Google",
        "Product" : "Nexus",
        "No" : 1
}
{
        "_id" : ObjectId("53c3fd3bb9ae26fa217b1e12"),
        "Company" : "Apple",
        "Product" : "iPhone",
        "No" : 2
}

Here we have 2 documents with fields _id, Company, Product and No.

Update with UNSET Operator: MongoDB


You need to a flashplayer enabled browser to view this YouTube video

YouTube Link: http://www.youtube.com/watch?v=8UUdITiZ5ZE [Watch the Video In Full Screen.]



We could remove a field using update() method, by not specifying it in the second argument.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
> db.names.update({"No": 1}, {"Company": "Google", "No": 1});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c3fc2005e0ce2719d91bd2"),
        "Company" : "Google",
        "No" : 1
}
{
        "_id" : ObjectId("53c3fd3bb9ae26fa217b1e12"),
        "Company" : "Apple",
        "Product" : "iPhone",
        "No" : 2
}

Related Read: Update Method: MongoDB

This way, we could eliminate the field Product. But this is a tedious process – we need to remember all the fields inorder to achieve this. It gets difficult when we have many fields in our document. To solve this problem, we have $unset operator, where we only need to know the field name which we want to remove.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
> db.names.update({"No": 1}, {$unset: {"Product": 1}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c3fc2005e0ce2719d91bd2"),
        "Company" : "Google",
        "No" : 1
}
{
        "_id" : ObjectId("53c3fd3bb9ae26fa217b1e12"),
        "Company" : "Apple",
        "Product" : "iPhone",
        "No" : 2
}

$unset operator syntax is same to unset or remove array from a document.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
> db.names.update({"No": 1}, {$set: {"Product": ["LG", "Samsung", "HTC"]}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c3fc2005e0ce2719d91bd2"),
        "Company" : "Google",
        "No" : 1,
        "Product" : [
                "LG",
                "Samsung",
                "HTC"
        ]
}
{
        "_id" : ObjectId("53c3fd3bb9ae26fa217b1e12"),
        "Company" : "Apple",
        "Product" : "iPhone",
        "No" : 2
}
 
> db.names.update({"No": 1}, {$unset: {"Product": 1}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c3fc2005e0ce2719d91bd2"),
        "Company" : "Google",
        "No" : 1
}
{
        "_id" : ObjectId("53c3fd3bb9ae26fa217b1e12"),
        "Company" : "Apple",
        "Product" : "iPhone",
        "No" : 2
}

Related Read: Update with SET Operator: MongoDB

Update with SET Operator: MongoDB

Lets use $set operator along with update() method, to update the documents.

test database, names collection

1
2
3
> db.names.find()
{ "_id" : ObjectId("53be5d4604cc1cb0a7bfc3c0"), "name" : "Alia" }
{ "_id" : ObjectId("53be5d5204cc1cb0a7bfc3c1"), "name" : "Bebo" }

update-with-set-operator-mongodb

Related Read: Update Method: MongoDB

Lets update the first document using only update method.

1
2
3
4
5
6
7
8
9
10
> db.names.update({"name": "Alia"}, {"name": "Alia", "age": 25});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{ "_id" : ObjectId("53be5d5204cc1cb0a7bfc3c1"), "name" : "Bebo" }
{
        "_id" : ObjectId("53be5d4604cc1cb0a7bfc3c0"),
        "name" : "Alia",
        "age" : 25
}

Everything is ok in this case. But what if you forget to mention the name field in the second argument of update() method.

1
2
3
4
5
6
> db.names.update({"name": "Alia"}, {"age": 25});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{ "_id" : ObjectId("53be5d5204cc1cb0a7bfc3c1"), "name" : "Bebo" }
{ "_id" : ObjectId("53be5d4604cc1cb0a7bfc3c0"), "age" : 25 }

In this case, the name field gets erased. As we did not specify name field in the second parameter of update() method.

Using $set operator

1
2
3
4
5
6
7
8
9
10
> db.names.update({"name": "Bebo"}, {$set: {"age": 25}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{ "_id" : ObjectId("53be5d4604cc1cb0a7bfc3c0"), "age" : 25 }
{
        "_id" : ObjectId("53be5d5204cc1cb0a7bfc3c1"),
        "name" : "Bebo",
        "age" : 25
}

Here we make use of document with name as Bebo. Using $set operator we only specify the fields we want to add or update. Need not specify other existing fields in order to retain them.

Update with SET Operator: MongoDB


You need to a flashplayer enabled browser to view this YouTube video

YouTube Link: http://www.youtube.com/watch?v=nsDnZcstUV4 [Watch the Video In Full Screen.]



1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
> db.names.update({"name": "Bebo"}, {$set: {"age": 25, "salary": 25}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{ "_id" : ObjectId("53be5d4604cc1cb0a7bfc3c0"), "age" : 25 }
{
        "_id" : ObjectId("53be5d5204cc1cb0a7bfc3c1"),
        "name" : "Bebo",
        "age" : 25,
        "salary" : 25
}
 
 
> db.names.update({"name": "Bebo"}, {$set: {"age": 26, "salary": 25}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{ "_id" : ObjectId("53be5d4604cc1cb0a7bfc3c0"), "age" : 25 }
{
        "_id" : ObjectId("53be5d5204cc1cb0a7bfc3c1"),
        "name" : "Bebo",
        "age" : 26,
        "salary" : 25
}
 
 
> db.names.update({"name": "Bebo"}, {$set: {"salary": 30}});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{ "_id" : ObjectId("53be5d4604cc1cb0a7bfc3c0"), "age" : 25 }
{
        "_id" : ObjectId("53be5d5204cc1cb0a7bfc3c1"),
        "name" : "Bebo",
        "age" : 26,
        "salary" : 30
}

We can update existing field or add a field to the existing document using $set operator and need not to worry about the other fields in the document.

Update Method: MongoDB

Lets learn how to update MongoDB document using update() method.

test database, names collection

1
2
3
4
5
6
> db.names.find().pretty()
{
        "_id" : ObjectId("53c3c4b1a0eddb0a706e4f56"),
        "name" : "Satish",
        "age" : 27
}

Observe the document, with fields name and age. We’ll be illustrating update() method by working on this document.

update-method-mongodb

1
2
3
4
5
> db.names.update({"name": "Satish"}, {"age": 28});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{ "_id" : ObjectId("53c3c4b1a0eddb0a706e4f56"), "age" : 28 }

update() method takes at least 2 arguments. First argument being the condition(WHERE clause in sql) and the second argument being the fields to be updated. Observe that, whatever the fields we specify in the second argument are only retained(along with _id), all other fields will be erased.

1
2
3
4
5
6
7
8
9
> db.names.update({"age": 28}, {"name": "Satish", "age": 28});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c3c4b1a0eddb0a706e4f56"),
        "name" : "Satish",
        "age" : 28
}

Now we’ve updated with name as well as age field and it reflects in the document in the collection.

Update Method: MongoDB


You need to a flashplayer enabled browser to view this YouTube video

YouTube Link: http://www.youtube.com/watch?v=6LfvsoqRoP8 [Watch the Video In Full Screen.]



1
2
3
4
5
6
7
8
9
10
> db.names.update({"age": 28}, {"name": "Satish", "age": 28, "salary": 200000});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{
        "_id" : ObjectId("53c3c4b1a0eddb0a706e4f56"),
        "name" : "Satish",
        "age" : 28,
        "salary" : 200000
}

If we want to update/add a new field to the document, we must also specify all other fields we want to retain in the document.

1
2
3
4
5
> db.names.update({"age": 28}, {"salary": 300000});
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
 
> db.names.find().pretty()
{ "_id" : ObjectId("53c3c4b1a0eddb0a706e4f56"), "salary" : 300000 }

If we forget to specify other fields, then they will be erased(except _id field).

Count Method: MongoDB

Let’s learn to use count() method in MongoDB.

Count method outputs numeric value of the number of documents retrieved.

count-method-mongodb

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
> use test
switched to db test
> db.names.find().pretty()
{ "_id" : ObjectId("53be5d4604cc1cb0a7bfc3c0"), "name" : "Alia" }
{ "_id" : ObjectId("53be5d5204cc1cb0a7bfc3c1"), "name" : "Bebo" }
{ "_id" : ObjectId("53be5d5904cc1cb0a7bfc3c2"), "name" : "Chameli" }
{ "_id" : ObjectId("53be5d6104cc1cb0a7bfc3c3"), "name" : "Dev D" }
{ "_id" : ObjectId("53be5d6804cc1cb0a7bfc3c4"), "name" : "Emli" }
{ "_id" : ObjectId("53be5d8604cc1cb0a7bfc3c5"), "name" : "Farhan" }
{ "_id" : ObjectId("53be5d9204cc1cb0a7bfc3c6"), "name" : "Gangs" }
{ "_id" : ObjectId("53be5d9904cc1cb0a7bfc3c7"), "name" : "Hum" }
{ "_id" : ObjectId("53be5e3704cc1cb0a7bfc3c8"), "name" : 25 }
> db.names.count();
9
> db.names.count({"name": {$type: 2}});
8
> db.names.count({"name": {$type: 1}});
1
> db.names.count({"name": {$regex: "e"}});
3

We have 9 documents, out of which 8 documents have string values as name and 1 document has numeric value as it’s name.

Count Method: MongoDB


You need to a flashplayer enabled browser to view this YouTube video

YouTube Link: http://www.youtube.com/watch?v=1Cx0KT43xlw [Watch the Video In Full Screen.]



Related Read: $exists, $type, $regex operators: MongoDB

$type: 2, represents string value in BSON specification. And we have 8 documents with string value for the name field.

$type: 1, represents double value in BSON specification. And we have 1 document with numeric value for the name field.

$regex: “e”, we have 3 documents which has small letter e in the string value for name field.

Note: We can make use of count() method, when we want to show the number of friends or followers each member has. So MongoDB makes it easy by facilitating developers with count() method.

SELECT / LIST Records From Database Table: PHP & MySQL
In sql, we have SELECT count(*) FROM table_name;

Cursor Object: MongoDB

Lets have a deeper look into the MongoDB cursor object.

cursor-object-mongodb

Documents in our collection
test database, names collection.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
> use test
switched to db test
> show collections
names
system.indexes
> db.names.find();
{ "_id" : ObjectId("53be5d4604cc1cb0a7bfc3c0"), "name" : "Alia" }
{ "_id" : ObjectId("53be5d5204cc1cb0a7bfc3c1"), "name" : "Bebo" }
{ "_id" : ObjectId("53be5d5904cc1cb0a7bfc3c2"), "name" : "Chameli" }
{ "_id" : ObjectId("53be5d6104cc1cb0a7bfc3c3"), "name" : "Dev D" }
{ "_id" : ObjectId("53be5d6804cc1cb0a7bfc3c4"), "name" : "Emli" }
{ "_id" : ObjectId("53be5d8604cc1cb0a7bfc3c5"), "name" : "Farhan" }
{ "_id" : ObjectId("53be5d9204cc1cb0a7bfc3c6"), "name" : "Gangs" }
{ "_id" : ObjectId("53be5d9904cc1cb0a7bfc3c7"), "name" : "Hum" }
{ "_id" : ObjectId("53be5e3704cc1cb0a7bfc3c8"), "name" : 25 }

We have 8 documents with name in alphabetical order, and 1 document with name as 25. So in total we have 9 documents in our names collection.

Establishing Cursor

1
2
3
4
5
6
7
8
9
10
11
12
> var cur = db.names.find();
 
> cur
{ "_id" : ObjectId("53be5d4604cc1cb0a7bfc3c0"), "name" : "Alia" }
{ "_id" : ObjectId("53be5d5204cc1cb0a7bfc3c1"), "name" : "Bebo" }
{ "_id" : ObjectId("53be5d5904cc1cb0a7bfc3c2"), "name" : "Chameli" }
{ "_id" : ObjectId("53be5d6104cc1cb0a7bfc3c3"), "name" : "Dev D" }
{ "_id" : ObjectId("53be5d6804cc1cb0a7bfc3c4"), "name" : "Emli" }
{ "_id" : ObjectId("53be5d8604cc1cb0a7bfc3c5"), "name" : "Farhan" }
{ "_id" : ObjectId("53be5d9204cc1cb0a7bfc3c6"), "name" : "Gangs" }
{ "_id" : ObjectId("53be5d9904cc1cb0a7bfc3c7"), "name" : "Hum" }
{ "_id" : ObjectId("53be5e3704cc1cb0a7bfc3c8"), "name" : 25 }

Look at the contents of our cursor object cur.

hasNext() and next() methods on Cursor

1
2
3
4
5
6
7
8
9
> var cur = db.names.find();
> cur.hasNext();
true
> cur.next()
{ "_id" : ObjectId("53be5d4604cc1cb0a7bfc3c0"), "name" : "Alia" }
> cur.next()
{ "_id" : ObjectId("53be5d5204cc1cb0a7bfc3c1"), "name" : "Bebo" }
> cur.next()
{ "_id" : ObjectId("53be5d5904cc1cb0a7bfc3c2"), "name" : "Chameli" }

If there are any documents to iterate inside cursor object, then hasNext() will return true orelse it’ll return false. If hasNext() returns true, then we can iterate through the documents using next() method on the cursor object.

sort() method on Cursor Object

1
2
3
4
5
6
7
8
9
10
11
12
> var cur = db.names.find();
 
> cur.sort({"name": -1});
{ "_id" : ObjectId("53be5d9904cc1cb0a7bfc3c7"), "name" : "Hum" }
{ "_id" : ObjectId("53be5d9204cc1cb0a7bfc3c6"), "name" : "Gangs" }
{ "_id" : ObjectId("53be5d8604cc1cb0a7bfc3c5"), "name" : "Farhan" }
{ "_id" : ObjectId("53be5d6804cc1cb0a7bfc3c4"), "name" : "Emli" }
{ "_id" : ObjectId("53be5d6104cc1cb0a7bfc3c3"), "name" : "Dev D" }
{ "_id" : ObjectId("53be5d5904cc1cb0a7bfc3c2"), "name" : "Chameli" }
{ "_id" : ObjectId("53be5d5204cc1cb0a7bfc3c1"), "name" : "Bebo" }
{ "_id" : ObjectId("53be5d4604cc1cb0a7bfc3c0"), "name" : "Alia" }
{ "_id" : ObjectId("53be5e3704cc1cb0a7bfc3c8"), "name" : 25 }

We could modify the cursor object using methods like sort(), limit() and skip(). In above example, we are modifying cursor object using sort() method, and we are sorting it in reverse lexicographical order on the name field.

limit() method on Cursor Object

1
2
3
4
5
> var cur = db.names.find();
> cur.limit(3);
{ "_id" : ObjectId("53be5d4604cc1cb0a7bfc3c0"), "name" : "Alia" }
{ "_id" : ObjectId("53be5d5204cc1cb0a7bfc3c1"), "name" : "Bebo" }
{ "_id" : ObjectId("53be5d5904cc1cb0a7bfc3c2"), "name" : "Chameli" }

We could limit the output/result using limit() method.

Chaining method on Cursor Object

1
2
3
4
5
6
7
8
> var cur = db.names.find();
 
> cur.sort({"name": -1}).limit(5).skip(2);
{ "_id" : ObjectId("53be5d8604cc1cb0a7bfc3c5"), "name" : "Farhan" }
{ "_id" : ObjectId("53be5d6804cc1cb0a7bfc3c4"), "name" : "Emli" }
{ "_id" : ObjectId("53be5d6104cc1cb0a7bfc3c3"), "name" : "Dev D" }
{ "_id" : ObjectId("53be5d5904cc1cb0a7bfc3c2"), "name" : "Chameli" }
{ "_id" : ObjectId("53be5d5204cc1cb0a7bfc3c1"), "name" : "Bebo" }

Here we chain the methods sort(), limit() and skip(). We are sorting in reverse lexicographical order on the name field, then skipping the first 2 documents and then limiting the result/output to 5 documents.

The order in which these 3 methods are applied are: First sort, then skip and then limit.
Also note that, these methods modify cursor object at the server side and not on client site.

Cursor Object: MongoDB


You need to a flashplayer enabled browser to view this YouTube video

YouTube Link: http://www.youtube.com/watch?v=_LexQW0vSlQ [Watch the Video In Full Screen.]



explain() method on Cursor Object

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
> db.names.find().explain();
{
        "cursor" : "BasicCursor",
        "isMultiKey" : false,
        "n" : 9,
        "nscannedObjects" : 9,
        "nscanned" : 9,
        "nscannedObjectsAllPlans" : 9,
        "nscannedAllPlans" : 9,
        "scanAndOrder" : false,
        "indexOnly" : false,
        "nYields" : 0,
        "nChunkSkips" : 0,
        "millis" : 0,
        "server" : "PC:27017",
        "filterSet" : false
}

explain() method shows that find() returns a basic cursor. More on explain() method in coming videos.

Programmatic way of printing Cursor Object content

1
2
3
4
5
6
7
8
9
10
11
12
> var cur = db.names.find();
 
> while(cur.hasNext()) printjson(cur.next());
{ "_id" : ObjectId("53be5d4604cc1cb0a7bfc3c0"), "name" : "Alia" }
{ "_id" : ObjectId("53be5d5204cc1cb0a7bfc3c1"), "name" : "Bebo" }
{ "_id" : ObjectId("53be5d5904cc1cb0a7bfc3c2"), "name" : "Chameli" }
{ "_id" : ObjectId("53be5d6104cc1cb0a7bfc3c3"), "name" : "Dev D" }
{ "_id" : ObjectId("53be5d6804cc1cb0a7bfc3c4"), "name" : "Emli" }
{ "_id" : ObjectId("53be5d8604cc1cb0a7bfc3c5"), "name" : "Farhan" }
{ "_id" : ObjectId("53be5d9204cc1cb0a7bfc3c6"), "name" : "Gangs" }
{ "_id" : ObjectId("53be5d9904cc1cb0a7bfc3c7"), "name" : "Hum" }
{ "_id" : ObjectId("53be5e3704cc1cb0a7bfc3c8"), "name" : 25 }

While loop executes until cur.hasNext() returns true. Until cur.hasNext() is true, cur.next() keeps printing next document in the cursor cur.