Display nearby books in Explore Nearby Screen

[Mohitmadhav]

Description

• Write code so that we can find all the users in a 5 km radius.
• We should be able to do it by calculating distance via a lat long field of users.
• We should then be able to fetch all books that belong to all the users within a 5 km radius.
• Similarly, we need to do for 5-10 km, and 10-20 km (The dummy UI is shown below)

Screenshots

Steps to Reproduce

Please provide detailed steps for reproducing the issue.

1. Firebase files: lib/services/database_service.dart
2. Explore Nearby Screen: lib/screens/explore_nearby.dart
3. UserData is created whenever the User gets signed in. (See lib/services/auth.dart)
4. So the book should contain all the book and owner info ( See Revamp Book and User Collections  #92 )
5. Add the code in both the files using Firebase code and Firebase ML (if required)
6. Do not remove the Empty Page Widget, just keep it for the case when no books are available.

Additional data

If there's something related to ML, then you can always discuss it with @ankitoscar here.
Share relevant app screenshots of your before committing any changes.

[kshitij5]

Hi, I would like to try solve this issue!

[Mohitmadhav]

Cool @kshitij5, and make sure to follow linting and the README guidelines.

[Mohitmadhav]

@ankitoscar you can help out @kshitij5 if required.

[kshitij5]

@Mohitmadhav yes, I might need help! As of now, what I am doing is fetching all the other user's details, calculating their distances from the current user, and displaying books listed by them as per the requirement! But, I think this will to too costly and not scalable also! So, is there any way to reduce the reads, or is it okay for now?

[Mohitmadhav]

We just need the users within 10 km radius

[kshitij5]

Hi, @Mohitmadhav. I tried implementing this feature but I just now figured out that redundant data is also getting added, every time the screen changes! I tried with Set but that doesn't seem to work. Is there any workaround?

[Mohitmadhav]

What redundant data exactly?
@ankitoscar @HARSHBHUDOLIA can you also join?

[HARSHBHUDOLIA]

For now I also can think of only that way which you did, one way to improve the algorithm is to store which city the user is in addition to its location, this can be helpful as we then only have to calculate distance of those in same city, we can also customise it to add blocks/sector in cities.

[HARSHBHUDOLIA]

The Geolocation package/api gives a detailed address based on location, so if we can store all the different values like city/state/pin number/street in user in different fields then we can do a better search, we can start with Pin Number, which is generally associated with nearest post office to limit the calculations.

I hope what I am saying is understandable