I am new to Swift and trying to write my first function that calls a closure that the user passes in. I am having trouble calling my function (which I named fun). I also was unable to find any examples of this online. I just want to call my closure (I am unsure what to pass to it?) and then make a decision based on the boolean result? This seems very easy yet I am not sure.
The goal of the method is to remove duplicates in an array based on the users specifications. In my case I may want to pass in a certain class and an array of it and then remove all classes that have the same name property (ie a name string that matches).
extension Array{
func removeDuplicates<T: Comparable>(fun: (elem: T, arr: [T]) -> Bool) -> [T]
{
var array = [T]()
for element in self
{
if fun(elem: T(), arr: [T])
{
println("hello")
}
}
return array
}
}
This is a slight generalization of Does there exist within Swift's API an easy way to remove duplicate elements from an array? and might be what you are
looking for:
extension Array {
func withoutDuplicates<U : Hashable>(attribute : T -> U) -> [T] {
var result : [T] = []
var seen : Set<U> = Set()
for elem in self {
let value = attribute(elem)
if !seen.contains(value) {
result.append(elem)
seen.insert(value)
}
}
return result
}
}
The attribute closure is applied to each array element, and
a Set is used to keep track of values that occurred already.
Therefore the value type U is required to be Hashable (which
is the case for strings).
Example:
struct Person : Printable {
let firstName : String
let lastName : String
var description: String {
return "\(firstName) \(lastName)"
}
}
let array = [
Person(firstName: "John", lastName: "Doe"),
Person(firstName: "Joe", lastName: "Miller"),
Person(firstName: "Jane", lastName: "Doe")
]
let result = array.withoutDuplicates( { $0.lastName } )
println(result)
// [John Doe, Joe Miller]
An alternative implementation is
func withoutDuplicates<U : Hashable>(attribute : T -> U) -> [T] {
var seen : [U : Bool] = [:]
return self.filter { seen.updateValue(true, forKey: attribute($0)) == nil }
}
which utilizes the fact that the updateValue() method of Dictionary
returns the previous value for the key, and in particular returns nil if the key was not set previously. This is also just a slight generalization of #rintaro's answer to iOS Swift: Filter array to unique items.
Related
Closed. This question needs debugging details. It is not currently accepting answers.
Edit the question to include desired behavior, a specific problem or error, and the shortest code necessary to reproduce the problem. This will help others answer the question.
Closed 2 years ago.
Improve this question
I have a computed property of type Parameters in my APIRouter
// MARK: - Parameters
private var parameters: Parameters? {
switch self {
case .searchForDoctors(let doctorsFilter):
var params: Parameters = ["main_category_id": doctorsFilter.0, "page": doctorsFilter.1, "specialty_id": doctorsFilter.2, "city_id": doctorsFilter.3, "region_id": doctorsFilter.4, "name": doctorsFilter.5, "company_id": doctorsFilter.6, "order_by": doctorsFilter.7]
return params
default:
return nil
}
}
some values in the Typealias called doctorsFilter are optional.
currently I have a warning asking me to provide default value for the optional values, and I don't want to provide default values , I want to check if the value exist to add it, otherwise i will not add the key and the value
how can I safely unwrap the optional values and add it to the parameters dictionary with out saying if let for all optional values?
example:
if let specialtyID = doctorsFilter.2 {
params["specialty_id"] = specialtyID
}
I don't want to unwrap it this way as I will check for all optional values and it will take more lines of code
EDIT:-
the DoctorsFilter type is documented, when I initialize an instance of type DoctorsFilter the autocomplete tells me which of them is what, I I've thought about making the DoctorsFilter class before but I'm looking for another way if any, maybe a built in reserved word can handle the whole situation! , I want to make it simple as much as possible.
making a function that handles the dictionary and returns it in DoctorsFilter class is an option. I'm thinking of adding this function to the APIRouter, is it fine to add it there? is it the rule of the APIRouter to handle the parameters ? or the APIRouter just interested in taking the parameters and will not handle it ?
There is no "one line" solution, but you can use KeyPaths to reduce the series of if let ... statements down to a loop.
Start by creating a struct for your filter rather than using a tuple.
To facilitate this, we define a protocol for Parameterable - This protocol requires a dictionary that maps parameter names (String) to the property (KeyPath) that holds that parameter name as well as a function to return the parameters dictionary.
protocol Parameterable {
var paramNames: [String:KeyPath<Self,String?>] {get}
func parameters() -> [String:Any]
}
Use an extension to create a default implementation of the parameters() function, as the code will be the same for all Parameterables. It iterates over the dictionary entries and uses the associated KeyPath to access the relevant property and put it in the output dictionary. If a given property is nil then it simply isn't added to the output dictionary, because that is how dictionaries work. No need to explicitly check.
(If you import Alamofire then you can use the typedef Parameters where I have used [String:Any])
extension Parameterable {
func parameters() -> [String:Any] {
var parameters = [String:Any]()
for (paramName,keypath) in self.paramNames {
parameters[paramName]=self[keyPath:keypath]
}
return parameters
}
}
Use this protocol to create a DoctorsFilter implementation:
struct DoctorsFilter: Parameterable {
var mainCategoryId: String?
var page: String?
var specialtyId: String?
var cityID: String?
var regionId: String?
var name: String?
var companyId: String?
var orderBy: String?
let paramNames:[String:KeyPath<Self,String?>] = [
"main_category_id":\.mainCategoryId,
"page":\.page,
"specialty_id":\.specialtyId,
"city_id":\.cityID,
"region_id":\.regionId,
"name":\.name,
"company_id":\.companyId,
"order_by":\.orderBy]
}
private var parameters: Parameters? {
switch self {
case .searchForDoctors(let doctorsFilter):
return doctorsFilter.parameters()
case .someOtherThing(let someOtherThing):
return someOtherThing.parameters()
default:
return nil
}
}
}
The other approach is to simply split your creation of the parameters dictionary into multiple lines; If you assign nil against a dictionary key then there is no key/value pair stored in the dictionary for that key. In this case I have left your tuple approach in place, but you could use the struct (and I strongly suggest you do so)
private var parameters: Parameters? {
switch self {
case .searchForDoctors(let doctorsFilter):
var params: Parameters()
params["main_category_id"] = doctorsFilter.0
params["page"] = doctorsFilter.1
params["specialty_id"] = doctorsFilter.2
params["city_id"] = doctorsFilter.3
params["region_id"] = doctorsFilter.4
params["name"] = doctorsFilter.5
params["company_id"] = doctorsFilter.6
params["order_by"] = doctorsFilter.7
return params
default:
return nil
}
}
If we want to handle mixed properties, rather than just optional strings, we need to modify the code slightly. We need to use PartialKeyPath. This makes the code a little more complex since the subscript operator for a PartialKeyPath returns a double optional. This needs to be handled.
protocol Parameterable {
var paramNames: [String:PartialKeyPath<Self>] {get}
func parameters() -> [String:Any]
}
extension Parameterable {
func parameters() -> [String:Any] {
var parameters = [String:Any]()
for (paramName,keypath) in self.paramNames {
let value = self[keyPath:keypath] as? Any?
if let value = value {
parameters[paramName] = value
}
}
return parameters
}
}
struct DoctorsFilter:Parameterable {
var mainCategoryId: String?
var page: String?
var specialtyId: String?
var cityID: Int
var regionId: String?
var name: String?
var companyId: String?
var orderBy: String?
let paramNames:[String:PartialKeyPath<Self>] =
["main_category_id":\Self.mainCategoryId,
"page":\Self.page,
"specialty_id":\Self.specialtyId,
"city_id":\Self.cityID,
"region_id":\Self.regionId,
"name":\Self.name,
"company_id":\Self.companyId,
"order_by":\Self.orderBy]
}
There are three primary ways to safely unwrap an optional. You can also provide default values if you wish to unwrap an optional.
Guard Statement Unwrapping
var firstString: String?
// In some cases you might performa a break, continue or a return.
guard let someString = firstString else { return }
print(someString)
If Let Unwrapping
var secondString: String?
var thirdString: String?
thirdString = "Hello, World!"
// Notice that we are able to use the same "IF LET" to unwrap
// multiple values. However, if one fails, they all fail.
// You can do the same thing with "Guard" statements.
if let someString = secondString,
let someOtherString = thirdString {
print(someString)
print(someOtherString)
} else {
// With this code snippet, we will ALWAYS hit this block.
// because secondString has no value.
print("We weren't able to unwrap.")
}
Default Value Unwrapping
var fourthString: String?
// The ?? is telling the compiler that if it cannot be unwrapped,
// use this value instead.
print(fourthString ?? "Hello, World")
In Swift it is recommended that anytime you see a ! that you use some form of unwrapping. Swift is very "Type Safe".
Here's a resource you can use for Optionals.
https://docs.swift.org/swift-book/LanguageGuide/TheBasics.html
Your Solution
Your solution might look something like this.
private var parameters: Parameters? {
switch self {
case .searchForDoctors(let doctorsFilter):
if let mainCatID = doctorsFilter.0,
let page = doctorsFilter.1,
let specialtyID = doctorsFilter.2,
let cityID = doctorsFilter.3,
let regionID = doctorsFilter.4,
let name = doctorsFilter.5,
let companyID = doctorsFilter.6,
let orderBy = doctorsFilter.7 {
params: Parameters = ["main_category_id": mainCatID,
"page": page,
"specialty_id": specialtyID,
"city_id": cityID,
"region_id": regionID,
"name": name,
"company_id": companyID,
"order_by": orderBy]
return params
} else {
//Unable to safely unwrap, return nothing.
return nil
}
default:
return nil
}
}
Goal of the code:
To assign a struct dictionary with Strings as Keys and String Arrays as values to a variable and then pull one (can be at random) specific String key value in the String Array and return that one String element in the underlying String Array so that it can be used elsewhere (potentially assigned to a label.text)
Essentially (please reference code below), I want to access one value at random in myDictionary using a specific key ("keyOne"), and pull, let's say, "Value2" then return only the string "Value2" from the underlying String Array associated with "keyOne" using indexing.
Errors are in the code below.
The issue I'm thinking is that I haven't figured out how to turn my final var Testing = dict["keyOne"] into an Int compatible index... if it was an index, the code would pull an Int value and the corresponding String from the three Strings in the underlying value array (due to the three String values associated with "keyOne").
Also, variableView() just inherits the datasource from several other containers, but the var dataSource : Structure? is the main reference, so that is what I included.
Code so far:
let myDictionary = [Structure(name: "keyOne", text: ["Value1", "Value2", "Value3"]), Structure(name: "keyTwo", text: ["Value4", "Value5", "Value6"])]
lazy var dict = Dictionary(uniqueKeysWithValues: myDictionary.lazy.map { ($0.name, $0.text) })
struct Structure: Hashable {
var name: String
var text: [String]
init(name: String, text: [String]){
self.name = name
self.text = text
}
}
func variable(at index: Int) -> variableView {
let variable = variableView()
var Testing = dict["keyOne"]
variable.dataSource = Testing![index] <- Cannot assign value of type 'String' to type 'structure'
return variable
var dataSource : Structure? {
didSet {
label.text = "This is a test"
} else {
// n/a
}
}
Please note that the error message is above in the code for variable.dataSource = Testing![index].
I am also suspecting that my issue lies in the "looping" logic of how I am assigning a variable with a struct, to a datasource which references that same struct.
Any help is appreciated as I have been stuck on this for legitimately a week (I truly have exhausted every single StackOverflow answer/question pair I could find).
THANK YOU!
EDIT:
I found this documentation to assist me greatly with this, and I recommend anyone with a similar question as mine to reference this: https://swift.org/blog/dictionary-and-set-improvements/
Given the question and the discussion in the comments I would add a mutating func to the struct that removes and returns a random string
mutating func pullText() -> String? {
guard let index = text.indices.randomElement() else {
return nil
}
return text.remove(at: index)
}
Example
if let index = myDictionary.firstIndex(where: { $0.name == "keyOne" }),
let text = myDictionary[index].pullText() {
someLabel.text = text
}
Here is another example based on the code in the question
Assuming VariableView looks something like this
struct VariableView: View {
var dataSource : Structure?
var word: String?
var body: some View {
Text(word ?? "")
}
}
Then the func variable can be changed to
func variable() -> VariableView {
var variable = VariableView()
if let index = dict.firstIndex(where: { $0.name == "keyOne" }) {
variable.dataSource = dict[index]
variable.word = dict[index].pullText()
}
return variable
}
I have an array consists of Dictionary.
I need to group them by Key in a dictionary.
I tried the line, but do not know what to write in handler. I am trying
globalArray.groupBy(handler: {$0["Name"]})
it gives error;
Cannot convert value of type "String?" to closure result type "_"
my group by extension is as follows;
extension Sequence {
// Using a `typealias` because it's shorter to write `E`
// Think of it as a shortcut
typealias E = Iterator.Element
// Declaring a `K` generic that we'll use as the type of the key
// for the resulting dictionary. The only restriction is having
// it conforming to the `Hashable` protocol
func groupBy<K: Hashable>(handler: (E) -> K) -> [K: [E]] {
// Creating the resulting dictionary
var grouped = [K: [E]]()
// Iterating over our elements
self.forEach { item in
// Retrieving the key based on the current item
let key = handler(item)
if grouped[key] == nil {
grouped[key] = []
}
grouped[key]?.append(item)
}
return grouped
}
}
Could you please show me the right usage?
BR,
Erdem
I am using this extension to group array , and it is working superbly
extension Array {
func grouped<T>(by criteria: (Element) -> T) -> [T: [Element]] {
var groups = [T: [Element]]()
for element in self {
let key = criteria(element)
if groups.keys.contains(key) == false {
groups[key] = [Element]()
}
groups[key]?.append(element)
}
return groups
}
}
How I use
array.grouped { (object:MyObjectClass) -> String in
return object.location?.name ?? "EmptyKey"
//Here you need to return your key
}
Hope it is helpful to you
Below is my custom object class.
class UserGroups: NSObject {
let groupName: String
let users: [CheckIn]?
init(json:JSON) {
self.groupName = json[Constants.Models.UserGroups.groupName].stringValue
self.users = UserGroups.getUserGroupsList(jsonArray: json[Constants.Models.UserGroups.users].arrayValue)
}
class func getUserGroupsList(jsonArray: [JSON]) -> [CheckIn]{
return jsonArray.flatMap({ (jsonItem: JSON) -> CheckIn in
return CheckIn(json: jsonItem)
})
}
}
I've an array of above custom objects. How can I combine 2 or more custom objects into a single object by merging users of every object having same groupName.
Below is my CheckIn Model:
class CheckIn: NSObject {
let id: String
let firstName: String
let lastName: String
let latitude: String
let longitude: String
let hint: String
init(json: JSON) {
self.id = json[Constants.Models.CheckIn.id].stringValue
self.firstName = json[Constants.Models.CheckIn.firstName].stringValue
self.lastName = json[Constants.Models.CheckIn.lastName].stringValue
self.hint = json[Constants.Models.CheckIn.hint].stringValue
self.latitude = json["location"][Constants.Models.CheckIn.latitude].stringValue
self.longitude = json["location"][Constants.Models.CheckIn.longitude].stringValue
}
}
id field is not unique in CheckIn.
Here's a slightly simplified example that shows how to combine groups that have the same group name.
Here is the UserGroup class. users is now a variable (var) because we will be adding elements to groups to combine them.
class UserGroups: NSObject {
let groupName: String
var users: [String]?
init(groupName: String, users: [String]?) {
self.groupName = groupName
self.users = users
}
}
Here are three groups, two of the share the same group name, Blues.
let group1 = UserGroups(groupName: "Blues", users: ["Tom", "Huck", "Jim"])
let group2 = UserGroups(groupName: "Reds", users: ["Jo", "Ben", "Tommy"])
let group3 = UserGroups(groupName: "Blues", users: ["Polly", "Watson", "Douglas"])
Next, we'll put all the groups in an array.
let allGroups = [group1, group2, group3]
Here, we use Swift's reduce function to allow us to reduce the array to only groups with unique group names.
let compacted = allGroups.reduce([UserGroups](), { partialResult, group in
var dupe = partialResult.filter {$0.groupName == group.groupName }.first
if let dupeGroup = dupe {
dupeGroup.users?.append(contentsOf: group.users ?? [])
return partialResult
} else {
var newPartialResult = partialResult
newPartialResult.append(group)
return newPartialResult
}
})
The array is now reduced to unique groups, we print out all the groups and their users with the help of Swift's map function.
print(compacted.map { $0.users })
// Prints [
Optional(["Tom", "Huck", "Jim", "Polly", "Watson", "Douglas"]),
Optional(["Jo", "Ben", "Tommy"])
]
The Solution
You did not include the CheckIn model, but I will assume that it has some sort of an id field unique to each user. We will use this to make the object Hashable:
// Add this to your file outside of the UserGroups class
extension CheckIn: Hashable {
var hashValue: Int { return self.id }
}
Making it Hashable allows you to convert the Array to a Set, which does not allow duplicates and will remove them in a very efficient way.
// Change getUserGroupsList as follows
class func getUserGroupsList(jsonArray: [JSON]) -> [CheckIn] {
return Array(Set(jsonArray.flatMap({ (jsonItem: JSON) -> CheckIn in
return CheckIn(json: jsonItem)
})))
}
Optional Considerations
As an aside, in case you're coming from another language, Swift gives you nice type inference and default names for closure arguments ($0 is the first argument). You can probably make the code a little less verbose, but it's a matter of taste which is preferred.
class func getUserGroupsList(jsonArray: [JSON]) -> [CheckIn] {
return Array(Set(jsonArray.flatMap { CheckIn(json: $0) }))
}
Also consider whether you really want the return value to be an array. If you want the list to always have unique users, it is a bit more efficient to use a Set as your return type and forgo the conversion back to an Array like this:
class func getUserGroupsList(jsonArray: [JSON]) -> Set<CheckIn> {
return Set(jsonArray.flatMap { CheckIn(json: $0) })
}
Finally, consider whether you really need the users property to be optional. With sequence types, it is often sufficient to use an empty sequence to denote absence of a value. Depending on your situation, this may simplify your code. The final version looks like this:
class UserGroups: NSObject {
let groupName: String
let users: Set<CheckIn>
init(json:JSON) {
self.groupName = json[Constants.Models.UserGroups.groupName].stringValue
self.users = UserGroups.getUserGroupsList(jsonArray: json[Constants.Models.UserGroups.users].arrayValue)
}
class func getUserGroupsList(jsonArray: [JSON]) -> Set<CheckIn> {
return Set(jsonArray.flatMap { CheckIn(json: $0) })
}
}
Maintaining Order
The caveat is that Set does not maintain the order of the items. If the order of the groups does matter, we can use this solution instead:
class func getUserGroupsList(jsonArray: [JSON]) -> [CheckIn] {
var encountered: Set<CheckIn> = []
return jsonArray.flatMap { CheckIn(json: $0) }.filter { encountered.update(with: $0) == nil }
}
In this version, we still use a set, but only to maintain a set of items we've encountered already. The update method on a set returns the same value if it's already in the set or returns nil if it's being inserted for the first time. We use that to filter our array to those items being encountered for the first time while adding them to the set of encountered items to filter them out when they are subsequently encountered again.
I tried to create a custom iterator which returns wrapper abcContainer over raw data class abc
// raw data class
class abc {
var name : String = "";
init( _ value : String) {
name = value;
}
}
// with container, only "name" is to be visible
class abcContainer {
private var _abc : abc;
init( _ obj : abc) {
_abc = obj;
}
// + extra methods here
func getName() -> String {
return _abc.name
}
}
The point would be that the dictionary would return instances of abcContainer instead of just the plain raw abc class.
I wanted to use the sequence protocol to make the conversion automatic, but I was not able to transform the [String:abc] into [String:abcContainer] automatically like this:
// the iterator is implemented just iterating the inner basic dict
// but wrapping the result value as abcContainer
class abcIterator : Sequence, IteratorProtocol {
private var __source : [String:abc]?;
var index = 0
var myIterator : DictionaryIterator<String, abc>;
init(_ ctxArray: [String:abc]) {
self.__source = ctxArray
index = 0;
myIterator = (__source?.makeIterator())!
}
func next() -> abcContainer? {
let nextItem = myIterator.next();
if(nextItem != nil) {
return abcContainer((nextItem?.value)!);
}
return nil;
}
}
// this was supposed to be the wrapper over the collection
class abcCollection : Sequence {
private var __source : [String:abc]?;
init(_ list: [String:abc]) {
self.__source = list
}
func makeIterator() -> abcIterator {
return abcIterator(self.__source!);
}
}
I'm probably missing something very basic here. When I try to use the collection like this:
var dict : [String:abc] = [String:abc]();
dict["abba"] = abc("John Smith");
for (key,value) in abcCollection(dict) {
print(key, value.getName());
}
I get error: Expression type "abcCollection" is ambiguous without more context
Does anyone have idea how to make it work? What is missing? I have a feeling that this answer has the information I need...
Swift 2 to 3 Migration for Swift Sequence Protocol
The problem in your original code is that abcCollection(dict)
returned a sequence of abcContainer objects, and those cannot
be assigned to a (key, value) tuple.
You can achieve your goal with
class abcCollection : Sequence {
private var __source : [String:abc]
init(_ list: [String:abc]) {
self.__source = list
}
public func makeIterator() -> AnyIterator<(AnyObject,abcContainer)> {
let mapped = self.__source.lazy.map {
($0.key as AnyObject, abcContainer($0.value))
}
return AnyIterator(mapped.makeIterator())
}
}
Making __source non-optional makes all the (optional) unwrappings
redundant, and lazy.map { ... } returns a lazily evaluated
sequence of key/value pairs which is then type-erased.
Ok, perhaps the answer was abcIterator was not necessary, you could have defined the iterator directly just like done in the linked answer like this:
class abcCollection : Sequence {
private var __source : [String:abc]?;
init(_ list: [String:abc]) {
self.__source = list
}
public func makeIterator() -> AnyIterator<(AnyObject,abcContainer)> {
var it = self.__source?.makeIterator();
return AnyIterator {
let n = it?.next();
if n == nil { return nil }
return (n?.key as AnyObject, abcContainer((n?.value)!))
}
}
}
After that, the custom collection returned wrapped objects correctly.