DEV Community: himaratsu

How to use Siri Shortcuts in your app

himaratsu — Fri, 05 Oct 2018 09:58:28 +0000

Introduction

Siri shortcut is one of the biggest update in iOS 12.
If you said to Siri about your task, Siri responds with your apps installed.
SiriKit is published in iOS 12, which make developers enable to adapt Siri to their apps.

Recently, my app had adaptive SiriKit. I introduce my code.

For using Siri Shortcut, We have tw ways. NSUserActivity, or Intents.
Today I introduce NSUserActivity because it is very simple.

How to use Siri Shortcuts

1. Regist Siri Shortcut

My code:

import UIKit
import Intents  // 1. import Intents framework

class CheckinViewController: UIViewController {

   override func viewDidAppear(_ animated: Bool) {
        super.viewDidAppear(animated)

        // 2. register Siri shortcut action when user did some of action
        userActivity = NSUserActivity.checkinActivity
    }
}


extension NSUserActivity {

    public static let checkinType = "com.example.checkin"

    public static var checkinActivity: NSUserActivity {
        let userActivity = NSUserActivity(activityType: checkinType)

        userActivity.isEligibleForSearch = true

        // 3. set `isEligibleForPrediction` true
        //    you should set recommend phrase to `suggestedInvocationPhrase`
        if #available(iOS 12.0, *) {
            userActivity.isEligibleForPrediction = true
            userActivity.suggestedInvocationPhrase = "Check in"
        }
        userActivity.title = "Check in to the nearest station"

        return userActivity
    }
}

This code is mostly same to Apple sample code.
When user action something, you register it to iOS with NSUserActivity.

2. Use NSUserActivity

Once registering Siri, user can set up at Settings app > Siri shortcuts section.

Handling user action by Siri, you implemented delegate method.

func application(_ application: UIApplication, continue userActivity: NSUserActivity, restorationHandler: @escaping ([Any]?) -> Void) -> Bool {

        if userActivity.activityType == NSUserActivity.checkinType {
            // Handle Siri Action Here
            return true
        }

        return false
    }

And write definition in your info.plist.

<key>NSUserActivityTypes</key>
<array>
    <string>com.example.checkin</string>
</array>

Recap

In this article, I wrote how to use Siri Shortcut in your app with my actual code.
The way of using NSUserActivity is very simple and easy.

Hope this article helps anyone!

———

Use Carthage effectively

himaratsu — Wed, 26 Sep 2018 15:51:12 +0000

Carthage

Carthage is one of the most popular dependency management tool for iOS.

In my article, I introduce several convenient command to save your time.

Command

carthage update —platform ios

When executing carthage update, carthage create Cartfile.resolved from Cartfile you prepared. Then,
build frameworks.

In this command, you should specify platform you want to use frameworks in.

carthage update —platform ios

Carthage usually create multiple platforms frameworks(iOS, watchOS, tvOS, and macOS), but maybe not use these all.

carthage bootstrap —platform ios —cache-builds

When executing carthage bootstrap, carthage build frameworks from Cartfile.resolved file. Not from Cartfile.

carthage bootstrap —-platform iOS —cache-builds

—cache-builds means not rebuilding frameworks if you already had built frameworks in Carthage/Checkouts/
.

Fortunely, carthage automatically check if framework is built in same Swift version. If not, rebuild it.

Specify libraries

Cartfile is share between multi platforms. For example, iOS and watchOS.

Basically, command is here:
carthage bootstrap —platform iOS,watchOS

But it has problem. In many cases, watchOS doesn’t use all dependencies defined in Cartfile.

So, my recommend is:

carthage bootstrap —platform iOS
carthage bootstrap —platform watchos SWXMLHash

Separate command, and pass library name to build command for watchOS.

Use latest version of OSS

To shorten build time, you can use framework binary. It’s provided by Library Author, so you can use implicitly.

To use it, your project swift version needs to be same as swift version of that binary. If you use Swift 4.2, using library of version build in Swift 4.2.

Carthage skip checkout and build steps, so it is very efficiently.

Recap

Carthage is great tool, but build time is bottleneck. I introduce tips to resolve it.

Make hope this article helps you!

—-

*I'm studying English. If you notice English mistake, please tell me in the comments.

Self-Introduction

himaratsu — Tue, 11 Sep 2018 08:25:23 +0000

Hi, I'm mobile developer in Japan.

I make iPhone/iPad apps as my office work and private work.

My representative work is ALPACA, photo-related app.
ALPACA is the app that helps you delete photos in iPhone/iPad quickly.
This app group similar photos automatically. You can delete photo grouped except you picked as left.

This app is loved by people that take many pictures of babies, pets, and selfies.

I'd like to post iOS dev tips in English. (I'm studying English!)
I'll write about UIKit, Swift, Layout System(Auto Layout&UIStackView), and etc.

Please follow me if you have interested in iOS dev. Thank you!

*I'm studying English. If you notice English mistake, please tell me in the comments.

TensorFlowを読んでいく Getting Started

himaratsu — Sat, 10 Feb 2018 02:16:32 +0000

Getting Started

TensorFlowは機械学習のためのツールです。幅広い機能が含まれていますが、主な機能はディープニューラルネットワークモデルのために用意されています。

TensorFlowは多くのAPIを提供しています。このセクションではハイレベルなAPIを紹介します。TensorFlowの利用がはじめてなら、以下のドキュメントから読んでください。

Getting Started for ML Beginners
- 機械学習がはじめての人向け
Getting Started with TensorFlow
- 機械学習やったことある人向け

次に、以下のドキュメントを読んで、ハイレベルAPIの特徴を理解しましょう。

Checkpoints
- 学習の進捗を保存し、中断した場所から再開する方法の説明
Feature Columns
- Estimatorが、モデルを変更せずにさまざまな入力データ型を処理する方法の説明
Datasets Quick Start
- TensorFlowの入力パイプラインの説明
Creating Custom Estimators
- モデルを作り、学習させる方法を紹介

さらに詳しい情報は以下にあります。

The Low Level Introduction demonstrates how to use TensorFlow outside of the Estimator framework, for debugging and experimentation.

Low Level Introductionでは、デバッグや実験のために、Estimatorフレームワークの外側でTensorFlowを使う方法が書かれています(https://www.tensorflow.org/versions/master/programmers_guide/low_level_intro)
Programmer's Guideには、主なTensolFlowのコンポーネントの機能の紹介が書かれています
Tutorialsでは、TensorFlowの様々なモデルを使ったウォークスルーを読むことができます。