Category Archives: iOS

Handling Swift JSONDecoder typeMismatch error while using Codable protocol

If you have a problematic JSON, which can contain number or string for some keys, you can decode object without that property and manually set that property after decoding.

For example, I have a Vehicle class inside HistoryItem. In Vehicle model_year can be empty String or non empty Int. Here I am decoding modelYear manually using NSDictionary and trying to get Int. Swift 4 cannot do it automatically.

This is my Vehicle class that is contained inside HistoryItem:

As you see, there is no model_year key in CodingKeys.

Why Swift at least should be faster than Objective C

In sections Optionals and Chaining and Value types there is an answer to a post topic question – Why Swift at least should be faster than Objective C

Using guard has three benefits. While the syntax can act as an if statement, its primary benefit is inferring non-nullability. Where an if statement requires a case, guard assumes the case based on the condition provided. Also, since guard contains no scope, with exception of the else closure, leaseStart is presented as an unwrapped optional to the guard’s super-scope. Lastly, if the guard statement’s test fails, Swift requires the else to exit the current method or loop, ensuring leaseStart never is accessed when nil. This is performed with the keywords return, continue, break, or throw.
ObjC was weakly typed, and allowed any method to be called on any object at any time. If the method call failed, there was a default handler in the runtime that returned nil. That meant that no unwrapping or testing was needed, the equivalent statement in ObjC:

leaseStart = [[[aBuilding tenantList:5] leaseDetails] startDate]

would return nil and this could be tested. However, this also demanded that all method calls be dynamic, which introduces significant overhead. Swift’s use of optionals provides a similar mechanism for testing and dealing with nils, but does so in a way that allows the compiler to use static dispatch because the unwrapping action is called on a defined instance (the wrapper), versus occurring in the runtime dispatch system.

The programmer is free to choose which semantics are more appropriate for each data structure in the application. Larger structures like windows would be defined as classes, allowing them to be passed around as pointers. Smaller structures, like a 2D point, can be defined as structs, which will be pass-by-value and allow direct access to their internal data with no dereference. The performance improvement inherent to the pass-by-value concept is such that Swift uses these types for almost all common data types, including Int and Double, and types normally represented by objects, like String and Array.[41] Using value types can result in significant performance improvements in user applications also.[42]

To ensure that even the largest structs do not cause a performance penalty when they are handed off, Swift uses copy on write so that the objects are copied only if and when the program attempts to change a value in them. This means that the various accessors have what is in effect a pointer to the same data storage, but this takes place far below the level of the language, in the computer’s memory management unit (MMU). So while the data is physically stored as one instance in memory, at the level of the application, these values are separate, and physical separation is enforced by copy on write only if needed.[43]

Creating a cold signal in ReactiveCocoa 7

For any case, there is an arbitrarily complex way to do it.

This prints “HELLO WORLD!”.

AppCode vs XCode

AppCode has features that are not present in Xcode. At a minimum, AppCode will be a good addition to XCode.

The full list of features is here.

Here I will show examples of using AppCode features. Inspired by this article, that was very usefull for me.

1. Quick Fix feature in AppCode is more convenient. Refactoring to Modern Objective C syntax using Quick Fix (Alt + Enter).



2. Simplifying expressions using Quick Fix.


3. AppCode can generate code for you. Use Command + N.
Generated code – isEqual and hash methods for a class that has properties – objectID, name, selected:

Description method generated for a class with many properties:

4. Imports do not block other code and are folded. Typos in names are shown and renaming works better than in XCode.
These are small features, but very useful. Here defered should be renamed to deferred.

5. Unused imports and properties are shown.
XCode doesn’t do that.

6. Scroll from source feature in project navigator.
You can open a source file by name as in XCode, but after that you can also navigate to that source file in a project tree in one click.

Updating View With Two Data Sources in iOS

In this article I am going to describe a robust, simple and clean app architecture that allows to update a single iOS tableView from two different data sources asynchronously.

Let’s say, we have two sources of data, that need to be displayed in a tableView. When user pulls to refresh a tableView, two separate asynchronous requests are performed. While updating continues we need to show activity indicator and we can dismiss it only when both requests are finished.

This can be cleanly accomplished using ReactiveCocoa + MVVM. One of the benefits of ReactiveCocoa is in the easiness of synchronization of asynchronous data streams.

One would say, that this task can be also performed by using of dispatch groups. I agree, but it is more complex and not easy to do when the nature of data sources is different. For example, when one of the datasources is HTTP API and the second is a WebSocket.

Architecture looks like this:

In a ViewModel we have 3 arrays – cloudData1 (or any other name you want, that better describes your domain), cloudData2 and a union single datasource of a TableView – tableData. When cloudData1 or cloudData2 updates, the update of tableData is triggered and then it triggers an update of the TableView itself by bindings.

In the init method of a ViewModel we create a binding of cloudData1 and cloudData2 to tableData:

Also we combine two flags of progress cloud1RequestInProgress and cloud2RequestInProgress to a single one – inProgress.

LoadData method of a ViewModel starts the loading of data from each of two datasources:

And in a TableView we create a binding between ViewModels tableData and a View:

Pull to refresh triggers loadData of a ViewModel:

In a view we also create a binding between inProgress property of a ViewModel and the need to dismiss refreshing indicator:

Simple and brief -> easy to support, extend and reuse.

Почему нужно писать юнит-тесты, пример

Юнит тесты реально спасают, я вот написал тест:

и он не прошел. Оказалось, что я забыл проставить boolValue тут:

Что интересно, без boolValue никаких warnings и тем более ошибок сборки.

iOS Unit Test Examples / Snippets

In this article I will post common iOS Unit testing examples for testing of:

  1. Asynchronous loading of data using REST API
  2. Loading of a UIView
  3. Method call after UITableViewCell is clicked
  4. Static UITableView sections have correct number of rows
  5. Loading of a UIView from UIStoryboard
  6. IBOutlet connection
  7. Model logic
  8. UITableViewCell subclass
  9. IBAction / Method is implemented
  10. Testing bundle name

Asynchronous loading of data using REST API

Here [RSHTTPClient sharedClient] is an instance of AFHTTPSessionManager subclass. But it doesn’t matter, the main is the principle.

Loading of a UIView

Method call after UITableViewCell is clicked

Here we verify that method restoreHomeKitClicked is called when user selects a specific row in UITableView.

Static UITableView sections have correct number of rows

Loading of a UIView from UIStoryboard

IBOutlet connection

Model logic

Here we test that an array, that is formed inside our model has correct number of items. Model logic testing obviously can be different depending of logic itself.

UITableViewCell subclass

IBAction / Method is implemented

Testing bundle name

The last is a rare case, but helps to prevent someone changing a bundle name of project accidentally.

How to wait until several AFNetworking requests finish

Use dispatch_group_t, dispatch_group_enter dispatch_group_leave, dispatch_group_notify.

Also, below is code to send a huge array to server by dividing ti to parts. It is actual when a single request is too big for server – you get “URL too long error” or HTTP 414.

How to create semi-transparent outside border for UIImageView

Just in case if somebody looks for an UIImage with transparent outside border. If you simply set a layer border, you will get a transparent border, but you will see inside image behind that border, not outside image. I managed to create an ImageView with transparent outside border.

It looks like this:
cool image

The idea is simple. I save UIImage from UIImageView. Then I remove UIImage and set initial layer as border layer. Then I put a new smaller sublayer above it and set a saved UIImage as contents of it.

The same approach can be used for UIButton or UIView.

Что нового в iOS 9

Новинки в iOS 9:

1. Новое встроенное приложение News, улучшения Notes, Maps, Wallet, CarPlay.

2. На iPad появится Split View, можно будет пользоваться несколькими приложениями одновременно, например, Safari и Twitter.

3. Siri станет умнее. Например, можно будет попросить Siri напомнить о чем-то, когда ты будешь дома. Она сама создает уведомление с привязкой к локализации.

4. Улучшения производительности. Батарея на 1 час дольше будет держаться. Встроенные приложения начнут использовать Metal, то есть по сути графический процессор. Для обновления системы достаточно будет не 4, а 1 гб свободной памяти.

#iOS9, #Apple