WEBVTT 00:00.120 --> 00:00.680 Hi everyone. 00:00.680 --> 00:01.240 Welcome back. 00:01.240 --> 00:06.160 In this lecture we are going to be learning how to build our custom gatherers. 00:06.160 --> 00:12.720 So far we have explored all the different gatherers that are part of the API itself, which is basically 00:12.720 --> 00:14.040 part of the language itself. 00:14.040 --> 00:17.240 Now we are going to build something on our own for that purpose. 00:17.240 --> 00:22.360 I have a specific function called Demonstrate simple custom gatherer. 00:22.600 --> 00:24.000 So let's go to that function. 00:24.000 --> 00:26.480 So in this function we have this implementation. 00:26.480 --> 00:30.360 Currently the implementation here is that we are filtering the movies. 00:30.520 --> 00:36.640 And then we are transforming the movie to be in this format which is we have the star emoji over here. 00:36.640 --> 00:39.800 And then we are formatting using the title release. 00:39.840 --> 00:44.120 Here we are basically displaying the data in this particular format. 00:44.120 --> 00:46.560 And then we are printing the summary over here. 00:46.560 --> 00:52.400 So we are filtering the movies that are greater than 8.5 in rating and then printing the summary. 00:52.440 --> 00:55.280 Let's quickly execute this and then show you what the result is. 00:55.520 --> 01:01.640 And then we will see how we can implement the same logic using a custom gatherer implementation. 01:01.640 --> 01:01.710 Rotation. 01:02.990 --> 01:05.590 So this is how the result is going to be. 01:05.590 --> 01:08.350 So what we have is a traditional approach. 01:08.510 --> 01:13.110 Now we are going to be implementing the same logic using a custom gatherer. 01:13.110 --> 01:16.070 But as I mentioned before you have the prompt here. 01:16.070 --> 01:20.510 If you want to use prompt, please go ahead and use the prompt to write the code for you. 01:20.510 --> 01:26.270 But in this case, I'm going to be coding this by hand so that you are able to grasp the concept as 01:26.270 --> 01:26.910 we code. 01:26.990 --> 01:31.270 Okay, so in this case I'm going to call gatherer dot off. 01:31.590 --> 01:33.430 So you have multiple options off and off. 01:33.430 --> 01:36.070 Sequential off is what we are going to use for now. 01:36.230 --> 01:39.910 Off sequential is something which we will look at in the next lecture. 01:39.910 --> 01:41.950 So user off over here. 01:42.350 --> 01:43.470 What do we need actually. 01:43.470 --> 01:48.150 So if you go and look at the implementation of off so you have many different implementations. 01:48.150 --> 01:52.910 You have the integrator as like the first two options and then you have the supplier. 01:52.910 --> 01:54.630 But our use case is very simple. 01:54.630 --> 01:57.310 We don't want to maintain any state or anything like that. 01:57.350 --> 02:03.550 What we need is a transformer that's going to be taking the data, apply the filtering logic which is 02:03.550 --> 02:07.310 over here, and the map logic, and then push the value downstream. 02:07.310 --> 02:12.430 So what I'm going to do, I'm going to call something called gatherer dot integrator. 02:12.470 --> 02:14.030 So you can call an integrator. 02:14.030 --> 02:16.630 So integrator is the one which does the work for us. 02:16.630 --> 02:22.190 That's the one which is going to be taking care of applying the filtering logic and also the transforming 02:22.190 --> 02:22.630 logic. 02:22.630 --> 02:24.590 And then you have multiple options over here. 02:24.590 --> 02:30.110 You have the off greedy and off off greedy is something which is going to consume all the elements. 02:30.110 --> 02:30.910 That's part of this. 02:31.110 --> 02:33.790 So in our case we have a collection of movies, right. 02:33.790 --> 02:35.590 We want to consume all the elements. 02:35.590 --> 02:41.430 So in those kind of scenarios you would use off greedy off is something which we'll use in the integrator. 02:41.430 --> 02:45.870 If you want to apply some conditional logic which is not needed for our use case. 02:46.110 --> 02:46.550 Okay. 02:46.590 --> 02:48.470 And then I'm going to be adding a lambda. 02:48.710 --> 02:52.390 So this lambda is going to be holding a state. 02:52.390 --> 02:55.350 And then the movie and then the downstream. 02:55.350 --> 03:01.310 So downstream is how you basically push the value downstream in our code logic okay. 03:01.350 --> 03:03.430 So now the code logic is this one. 03:03.430 --> 03:10.060 If the movie Movieid rating is greater than 8.5, then take those movies and apply this filtering logic. 03:10.140 --> 03:13.380 So in here we need to basically give it a type actually. 03:13.380 --> 03:17.540 Otherwise it will complain about a movie is basically of type object. 03:17.540 --> 03:19.540 It cannot read this rating value. 03:19.540 --> 03:21.180 So in this case what we are going to do. 03:21.420 --> 03:23.700 So we are going to be calling gatherer. 03:23.740 --> 03:26.100 We need to define the types and everything over here. 03:26.140 --> 03:26.580 Okay. 03:26.660 --> 03:31.060 Then I'm going to give it a name called High Rated Movie. 03:31.100 --> 03:32.500 Summary okay. 03:32.540 --> 03:35.060 So this gatherer is going to be of type movie. 03:35.180 --> 03:39.620 And then the next step is going to be void because we are not going to maintain any state. 03:39.620 --> 03:41.180 The next one is going to be the output. 03:41.220 --> 03:42.540 What is the output we want. 03:42.660 --> 03:44.620 The output is going to be of type string. 03:44.660 --> 03:44.900 Right. 03:44.940 --> 03:47.460 So in this case let's provide the value as string okay. 03:47.500 --> 03:51.860 So as soon as I provided these three type now the compilation issue is gone. 03:51.860 --> 03:53.300 So the movie is a input. 03:53.300 --> 03:58.660 So if you go and take a look at the gatherer implementation here, it has a input type and output type 03:58.780 --> 04:05.460 and the actual state which is the one that you might need if you want to manage some State. 04:05.460 --> 04:08.220 We will look at it in the next lecture, but for now, let's keep it simple. 04:08.420 --> 04:11.580 So the input is movie and the output is string. 04:11.820 --> 04:12.260 Okay. 04:12.500 --> 04:17.700 And another thing that we need to do here is that after the downstream dot push is over, I think we 04:17.700 --> 04:19.420 have one another extra colon. 04:19.460 --> 04:22.460 Oh one thing we need to do is we need to return the value equal to true. 04:22.580 --> 04:26.380 So this means we are going to be continuing the processing. 04:26.380 --> 04:28.620 So that's what that one is meaning actually. 04:28.620 --> 04:30.620 And then add a semicolon over here. 04:30.620 --> 04:32.580 So this is our custom gatherer. 04:32.620 --> 04:36.660 This implementation is very similar to adding the filter and map. 04:36.780 --> 04:42.580 I do understand that this might look a little complex compared to this filter and map, which is very 04:42.580 --> 04:44.940 intuitive, and we are very much used to it. 04:45.100 --> 04:51.060 I'm just showing an option of how you can leverage gatherer for custom implementation, but if you think 04:51.100 --> 04:56.180 a traditional way of doing it using the filter and map and easy, please don't hesitate to use it. 04:56.180 --> 04:59.300 There is no need for you to use gatherer. 04:59.540 --> 05:05.620 If you think the logic is simple with filter and map intermediate operations, that is something I want 05:05.660 --> 05:06.490 to call it out. 05:06.490 --> 05:08.410 So now we have this gatherer created. 05:08.410 --> 05:14.250 Now what we are going to do here is that we are going to be using the movies dot stream, and then we 05:14.290 --> 05:19.970 are going to be passing the gatherer, which is this gatherer, and we are going to be printing the 05:19.970 --> 05:21.130 result in the console. 05:21.170 --> 05:21.650 Okay. 05:21.810 --> 05:25.850 So now let's go ahead and execute the program and see what the result is. 05:25.890 --> 05:29.130 So it's going to be rerunning the program one more time. 05:29.130 --> 05:31.650 You will see the same result being printed over here. 05:31.690 --> 05:32.170 There you go. 05:32.170 --> 05:33.770 So here is a traditional approach. 05:33.970 --> 05:37.210 And here is the actual gatherer approach. 05:37.210 --> 05:40.610 So now let's say if you don't push this what happens. 05:40.650 --> 05:40.810 Right. 05:40.810 --> 05:41.930 I'm not pushing anything. 05:41.930 --> 05:43.490 Ideally there should not be any data. 05:43.530 --> 05:43.970 Right. 05:43.970 --> 05:45.930 Let's go ahead and run this program one more time. 05:45.930 --> 05:47.690 You see there are no result. 05:47.690 --> 05:55.690 So downstream dot push is one of the code which is responsible for actually push the gathered result 05:55.690 --> 05:58.050 to the for each terminal operation. 05:58.090 --> 05:58.330 Okay. 05:58.370 --> 05:59.530 So this is really key. 05:59.650 --> 06:05.490 So so far the reason why we have been seeing the code working as expected for all the different gatherers 06:05.490 --> 06:06.450 that we have used. 06:06.570 --> 06:10.210 If you go ahead and take a look at the windows fixed logic. 06:10.250 --> 06:13.930 You will have a gatherer down the line that takes care of pushing the data. 06:13.970 --> 06:14.730 You can see that right? 06:14.770 --> 06:18.650 So here also we use the integrator dot off greedy. 06:18.690 --> 06:21.210 That's exactly what we are doing in our code also. 06:21.290 --> 06:26.410 So in our code also what we used is gatherer dot integrator dot off greedy. 06:26.690 --> 06:32.930 So if you don't use downstream dot push then no data will be pushed to the final result. 06:32.970 --> 06:36.930 Let me execute this program one more time and then show you the difference. 06:37.130 --> 06:37.530 There you go. 06:37.530 --> 06:38.650 Now we see the result. 06:38.690 --> 06:44.690 I hope you all were able to successfully implement a custom gatherer based on your traditional way of 06:44.730 --> 06:46.530 doing the filtering and mapping. 06:46.530 --> 06:51.010 So here we achieved both filtering and mapping within the gatherer itself. 06:51.050 --> 06:57.530 Again there is no hard rule of you need to use gatherer replacing all the filter map with gatherer. 06:57.530 --> 07:00.970 So gatherer is something which is handy for specific use cases. 07:00.970 --> 07:06.250 If you think if your use case can benefit from it, please go ahead and start using gatherer. 07:06.290 --> 07:07.890 But this marks the end of this lecture. 07:07.890 --> 07:08.890 Thank you for watching.