Saving an existing JSON file in Firebase using Swift? - json

Question

I have a JSON file, saved locally in my project. Now I'm curious about saving it in Firebase. I know that Firebase saves data in a NoSQL format (as a JSON). I've seen several tutorials about adding files to Firebase, (after some action, like tapping a button in an example iOS app). But is it possible to save a whole JSON file itself (not just chunks of it) in firebase, initially (not after tapping some button), and then load that file, because in my project users can't add files on their own? The app should load the data from JSON file.
I'm new to Firebase, so I have almost no experience using it. If you have any suggestions or know how to do this, I would appreciate your help.
EDITED
Well, to clarify what I am asking about. I have a project which contains a text file, which contains a JSON data. I added it to my project, initially, because I had no experience using Firebase, AWS etc. I just needed to see how the app reacts to data. In the future the data will grow significantly, so obviously I'm not going to save it inside of the project.
The problem is, I just want to save the JSON data in Firebase. And I don't know the right way to do that: by just importing the JSON or writing it from scratch (programmatically or in a console). Also, I would like to know whether it is generally a good idea or not, because this is the first time I'm using online databases, and many people say that Firebase is the best database to start from. I'm really new to all these stuff, that's why I'm asking questions like these.
EDITED_2
{
"list": {
"restaurants": [{
"id": "231311234342bn123",
"name": "Grand Royal",
"suggested": true,
"address": {
"longitude": "30.31527",
"latitude": "59.93688",
"formatted": ["Nevskiy prospekt 10, Saint Petersburg",
"Russian Federation"
]
},
"tags": ["fish",
"Italian",
"Friends"
],
"category": "$$$$",
"description": "Lorem ipsum dolor sit er elit lamet, consectetaur cillium adipisicing pecu, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum. Nam liber te conscient to factor tum poen legum odioque civiuda.",
"shortDescription": "an",
"collectionName": "Relaxing with friends",
"distance": "1324",
"country": "Russian Federation",
"Region": "North-West region",
"city": "Saint Petersburg"
},
{
"id": "123234344t5213",
"name": "N",
"image": "A",
"suggested": true,
"address": {
"longitude": "30.31821",
"latitude": "59.93605",
"formatted": ["Nevskiy prospekt 15, Saint Petersburg",
"Russian Federation"
]
},
"tags": ["fish",
"french",
"family"
],
"category": "$$$$",
"description": "Lorem ipsum dolor sit er elit lamet, consectetaur cillium adipisicing pecu, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum. Nam liber te conscient to factor tum poen legum odioque civiuda. ",
"shortDescription": "ab",
"collectionName": "Evening with a family",
"distance": "556",
"country": "Russian Federation",
"Region": "North-West region",
"city": "Saint Petersburg"
}]}}
Here is a snippet of my initial JSON data (it isn't completed, for example, at this moment there is only a string for an image). The data is used throughout the whole app (in a table view, separate views etc.). For example, I use it to populate the table view, use it in a mapView and in other UIViewController subclasses. Also, I need an ability to filter it and to search through it. Currently, I'm doing that by decoding the whole JSON and adding it to an array, and then, I filter that array. But I guess, that isn't the best choice when you have a large amount of data with images.
And one more thing, in Core Data it is possible to set batch requests, so it will initially fetch only some amount of data. It works fine with table views, because when you scroll them, Core Data automatically fetches new data to populate cells. Is it possible to do something like that with firebase, or should I read all data at once?
I've also read that Firebase doesn't require to save array-like data, but there are several ways to solve that issue. It is important to me, because I do need some part of the data to be in an array.

Answer 1

Hopefully this will provide some direction, pardon the length.
Right of the bat, there are usually better ways to store data in Firebase than arrays; arrays are not searchable, editable or addable and if you want to change something the entire array will need to be overridden. Please see Arrays Are Evil.
Secondly Denormalizing Data Is Normal in Firebase. What this means is that when data gets too 'deep' it's much more difficult to query for the data you want. Generally flatter is better - depending on what you want to get out of Firebase.
Third item is there are 100 different ways to work with the data in your question, I'll present two; one to demonstrate conceptually how to write the data and the second is probably a better option.
Here's how to write your data in a brute force kind of way
let addressDict = [
"longitude": "30.31527",
"latitude": "59.93688",
"formatted1": "Nevskiy prospekt 10, Saint Petersburg",
"formatted2": "Russian Federation"
]
let tagsDict = [
"fish": true,
"italian": true,
"friends": true
]
let restaurantDict = [
"name": "Grand Royal",
"suggested": true,
"address": addressDict,
"tags": tagsDict,
"category": "$$$$",
"description": "Lorem ipsum dolor si.",
"shortDescription": "an",
"collectionName": "Relaxing with friends",
"distance": "1324",
"country": "Russian Federation",
"Region": "North-West region",
"city": "Saint Petersburg"
] as [String : Any]
let restaurantsRef = self.ref.child("restaurants").childByAutoId()
restaurantsRef.setValue(restaurantDict)
Running this code will result in a JSON structure similar to your structure. Note that the restaurant key (KzEIkF6nxgWH1nA8B2D) was created automatically when the data is written so that's how the restaurant would be referred to in other structures, by its key.
"restaurants" : {
"-KzEIkF6nxgWH1nA8B2D" : {
"Region" : "North-West region",
"address" : {
"formatted1" : "Nevskiy prospekt 10, Saint Petersburg",
"formatted2" : "Russian Federation",
"latitude" : "59.93688",
"longitude" : "30.31527"
},
"category" : "$$$$",
"city" : "Saint Petersburg",
"collectionName" : "Relaxing with friends",
"country" : "Russian Federation",
"description" : "Lorem ipsum dolor si.",
"distance" : "1324",
"name" : "Grand Royal",
"shortDescription" : "an",
"suggested" : true,
"tags" : {
"fish" : true,
"friends" : true,
"italian" : true
}
}
},
To query for all $$$$ restaurants:
//print all restaurants with $$$$
let restRef = self.ref.child("restaurants")
let query = restRef.queryOrdered(byChild: "category").queryEqual(toValue: "$$$$")
query.observeSingleEvent(of: .value) { snapshot in
for child in snapshot.children {
let snap = child as! DataSnapshot
print(snap)
}
}
While we can query for all restaurants that have a category == "$$$$", the address and tags node are too deep (which is not actually accurate, see Deep Query, but we're going with it for this example as the concept is sound).
So a better structure is
"restaurants" : {
"-KzEIkF6nxgWH1nA8B2D" : {
"Region" : "North-West region",
"category" : "$$$$",
"city" : "Saint Petersburg",
"collectionName" : "Relaxing with friends",
"country" : "Russian Federation",
"description" : "Lorem ipsum dolor si.",
"distance" : "1324",
"name" : "Grand Royal",
"shortDescription" : "an",
"suggested" : true,
}
},
"addresses"
"-KzEIkF6nxgWH1nA8B2D"
"formatted1" : "Nevskiy prospekt 10, Saint Petersburg",
"formatted2" : "Russian Federation",
"latitude" : "59.93688",
"longitude" : "30.31527"
"tags"
"-KzEIkF6nxgWH1nA8B2D"
"fish" : true,
"friends" : true,
"italian" : true
and the code to create it:
let restDict = [
"name": "Jay's Eatery",
"category": "$$$$"
]
let addressDict = [
"formatted1": "Anytown"]
let tagsDict = [
"fish": true,
"friends": true,
"italian": true
]
let restRef = self.ref.child("restaurants")
let addressesRef = self.ref.child("addresses")
let tagsRef = self.ref.child("tags")
let thisRestaurantRef = restRef.childByAutoId()
thisRestaurantRef.setValue(restDict, withCompletionBlock: { (error, snapshot) in
if error == nil {
let thisRestaurantKey = snapshot.key
addressesRef.child(thisRestaurantKey).setValue(addressDict)
tagsRef.child(thisRestaurantKey).setValue(tagsDict)
}
})
Now you can easily query for restaurants with tags of fish, latitude starting with 59 or name of Grand Royal using the query from above.
As you can see, when data in stored in Firebase, there no encoding or decoding involved.
Yes, Firebase can batch load data to populate an array to be used as a tableView datasource
Best practice is to create class object that represents data in your code and to the user, and leverage Firebase on the back in to store that data.
class RestaurantClass {
key = "" //"-KzEIkF6nxgWH1nA8B2D"
name = "" //"Grand Royal
etc
func init(withSnapshot: DataSnapshot) {
//populate the class vars from the snapshot
}
Then the query for $$$$ restaurants would be something like this:
let restRef = self.ref.child("restaurants")
let query = restRef.queryOrdered(byChild: "category").queryEqual(toValue: "$$$$")
query.observeSingleEvent(of: .value) { snapshot in
for child in snapshot.children {
let snap = child as! DataSnapshot
let rest = RestaurantClass(withSnap: snap)
self.myTableViewDataSourceArray.append(rest)
}
self.myTableViewDataSourceArray.reloadData()
}
There's a lot more but this should get you going in the right direction.
The takeaway from this answer is to leverage Firebase to store raw data on the backend - it's blisteringly fast, expandable and maintainable given the data is structured properly. Use classes in code that are populated with that back end data to be used as a datasource and present data to the user.
One final note is if you are working with locations, you're going to want a tool to make that easier. Check out GeoFire for more information.
EDIT:
To address a comment for clarification.
You don't exactly create Firebase data structures in code like you would create properties for a class or struct and it's nothing like a empty SQL table with headers. It's more how you tell Firebase how to write that data. So for example:
ref.child("users").child("uid_0").setValue("Leroy")
tells Firebase that within the users node, write a key: value pair of a string: string of uid: Leroy like this:
users
uid_0: "Leroy"
and
let aDict = ["user_name": "Leroy"]
ref.child("users").child("uid_0").setValue(aDict)
will result in telling Firebase that within the users node to write a key: value pair of a string: dictionary of uid: aDict. Note that aDict is a key: value pair itself of string: string (e.g user_name: Leroy)
users
uid_0
user_name: "Leroy"
I like to think of Firebase is a big Dictionary of key: value pairs where keys are always strings (avoid arrays!) and values are of Any type which could be strings, ints, etc or other Dictionaries.