BD and mock data - IvanAlvarezEnri/dragon-squad-api GitHub Wiki

DataBase: We access to our location database in order to do a predictive search.

We have simulated a registry of 500 rooms spread over 100 locations around Madrid. We have done this creating random latitudes and longitudes near to Madrid and using reverse Geocode to stock the street name. Our objective with that is to achieve the following points:

• Stock ONLY the locations registered by the users.
• By limiting it to the area of Madrid, we can implement an endpoint that returns the rooms to us in locations near the given location, calculating a range with the bounds. If we covered more territory, it would be less efficient to test.

Why we did it like that? Pros and cons

• The OpenCage API does not support autocompletion nor partial matching. Tested by us and confirmed by contacting with them. We tried to call his api in order to get a predictable response, but we found a Bug with the country code filter and they said later that they are not good at predictable search.
• Pros: It allows a goos scalability and gives us more freedom: we can be more precise at searching, we can sort by countof rooms and show it to the users.
• Cons: Our server wil have more work to do.