1
00:00:02,210 --> 00:00:04,880
Now that we talked so much about relations,

2
00:00:04,940 --> 00:00:11,960
let me show you one useful operation mongodb offers for merging related documents that

3
00:00:11,960 --> 00:00:14,620
you split up by using the reference approach,

4
00:00:14,630 --> 00:00:18,050
so where you got no embedded documents and that is the lookup

5
00:00:18,110 --> 00:00:22,390
operator. Now the lookup operator is used like this,

6
00:00:22,430 --> 00:00:25,700
this uses the aggregate method which we haven't learned about,

7
00:00:25,700 --> 00:00:30,180
we got a whole module about the aggregation framework and I don't want to talk too much about it

8
00:00:30,290 --> 00:00:39,320
but lookup is essentially a helpful tool that allows you to fetch two related documents merged together

9
00:00:39,380 --> 00:00:43,850
in one document. in one step instead of having to do two steps

10
00:00:43,850 --> 00:00:50,360
and this mitigates some of the disadvantages of splitting your documents across collections because

11
00:00:50,360 --> 00:00:53,440
now you can at least merge them in one go.

12
00:00:53,450 --> 00:00:55,340
Let me show you how it works.

13
00:00:55,490 --> 00:00:58,690
Here we got our book author example still

14
00:00:58,740 --> 00:01:04,870
and now let's say I want to fetch all my books with the author data included, to do that

15
00:01:04,880 --> 00:01:09,800
I can use my books, my books collection and use the aggregate method,

16
00:01:09,830 --> 00:01:12,670
again this uses the so-called aggregation framework

17
00:01:12,710 --> 00:01:15,510
and we'll have a look at this later in the course in detail.

18
00:01:15,890 --> 00:01:21,280
There you pass an array because you can define multiple steps on how to aggregate your data,

19
00:01:21,320 --> 00:01:23,350
for now there is only one step I'm interested,

20
00:01:23,360 --> 00:01:29,470
a step is simply a document you pass to the array that is the lookup step, $lookup. That

21
00:01:29,480 --> 00:01:32,590
step is configured by passing a document as a value

22
00:01:32,720 --> 00:01:37,740
and here you need to define four things, I simply brought this into a new page to make it easier to read,

23
00:01:37,790 --> 00:01:40,370
so four things need to be defined here.

24
00:01:40,400 --> 00:01:42,230
The first one is the from thing

25
00:01:42,360 --> 00:01:49,480
and the question here is from which other collection do you want to relate documents?

26
00:01:49,490 --> 00:01:53,970
So here we're running aggregate on the books collection,

27
00:01:54,050 --> 00:02:00,660
so we want to pull in other documents from the authors collection,

28
00:02:00,680 --> 00:02:08,000
so that is the name of the collection where your related documents live in. You then define the local field,

29
00:02:08,040 --> 00:02:14,200
so in the collection we're running this on, so the books collection, where can the references to the other

30
00:02:14,310 --> 00:02:23,280
collection be found in? And for books, we stored that in the authors field,

31
00:02:26,850 --> 00:02:31,190
authors simply is an array of objectIds.

32
00:02:31,210 --> 00:02:36,660
Now we need to tell MongoDB ok I know where your keys are stored,

33
00:02:36,700 --> 00:02:40,540
which field are you relating to in your target collection,

34
00:02:40,540 --> 00:02:47,650
so in the author collection here, the authors collection and you define that with the foreign field key

35
00:02:47,850 --> 00:02:50,320
and there we know it's an _ID.

36
00:02:50,320 --> 00:02:56,700
Please note that all these values are enclosed in double quotation marks, authors, authors

37
00:02:56,830 --> 00:03:00,160
and here, my _ID field.

38
00:03:00,430 --> 00:03:07,570
So I know that in authors, in books I have objectId stored where the matching equivalence can be found

39
00:03:07,600 --> 00:03:11,680
in _ID in the authors collection.

40
00:03:11,810 --> 00:03:20,340
Last but not least, you give this an alias under which this will be merged and I'll name this creators, the

41
00:03:20,340 --> 00:03:21,290
name is up to you here.

42
00:03:22,530 --> 00:03:26,940
If you hit enter, let's pretty print this,

43
00:03:27,130 --> 00:03:33,500
we get back this result. What we see here is we get our book, actually all the books we have in there,

44
00:03:33,560 --> 00:03:37,390
only one in our case with the name with the authors and the objectIDs

45
00:03:37,400 --> 00:03:39,590
but then a new key was added,

46
00:03:39,680 --> 00:03:40,640
creators

47
00:03:40,820 --> 00:03:47,030
and that actually holds the author data which lived in a different collection and which was merged

48
00:03:47,060 --> 00:03:49,520
into a result by the aggregate

49
00:03:49,520 --> 00:03:51,660
and the lookup function.

50
00:03:51,710 --> 00:03:57,440
So this is very useful for merging data in one step and it allows you to get the best of both worlds,

51
00:03:57,710 --> 00:04:01,250
having it split up and still fetching it in one go.

52
00:04:01,250 --> 00:04:07,730
Now this still is not an excuse for always using references because obviously this costs more performance

53
00:04:07,850 --> 00:04:09,670
than having an embedded document,

54
00:04:09,680 --> 00:04:13,520
so if you can and if your application needs it, go for an embedded document.

55
00:04:13,610 --> 00:04:19,760
If you have to use references or if you want to use references, well then this lookup step in the aggregate

56
00:04:19,760 --> 00:04:23,450
method can at least help you get the data you need.

57
00:04:23,450 --> 00:04:28,490
Now it was just a first look at aggregate, as I mentioned we'll have a whole module where we see what

58
00:04:28,490 --> 00:04:30,340
else aggregate can do for us.

59
00:04:30,350 --> 00:04:30,890
Be aware of

60
00:04:30,890 --> 00:04:33,660
lookup for merging your relations though.