截圖原理(一)――Android自動化測試學習歷程
把兩節的內容彙總起來,第一節講的是如何在apk中直接進行截圖,用到了Robotium的Solo類的takeScreenShot方法,有一個小的 android demo,以及從方法一直往裡鑽,知道它具體是怎麼進行截圖的。
第二節講的是脫離apk,直接在PC端截圖,通過的是adb橋接的方式,呼叫ddmlib.jar包中的AndroidDebugBridge和IDevice的類,對其進行截圖,並儲存到我想要的位置,是可以寫成一個小工具的。
視訊地址:http://study.163.com/course/courseLearn.htm?courseId=712011#/learn/video?lessonId=877120&courseId=712011
一、面試問題引入:
1、怎樣在一個app崩潰前復現bug操作步驟?(非手工和人眼操作)
答:可以通過截圖實現,在關鍵步驟處均進行截圖操作,這樣app崩潰了也能夠根據之前的截圖進行現場確認和步驟復現。那麼如何實現截圖?
可以通過:
1、monkeyrunner裡面——device.takeSnapshot()
2、Robotium裡面——solo.takeScreenshot(String pictureName)
面試問題:
(1)takeScreenshot的實現原理?通過哪些方法得到截圖?是單執行緒還是多執行緒?得到的檢視物件是單一View還是View陣列?如果沒有裝載sdk卡,或者說想要儲存在PC端,該如何處理呢?
二、Robotium實現截圖操作,及原理
具體的screenshot以及robotium在有原始碼的情況下的一個具體testcase類就是如下這樣的示例:
package com.li.xiami.test; import static org.junit.Assert.*; import org.junit.After; import org.junit.Before; import org.junit.Test; import com.android.robotium.solo.Solo; import com.li.xiami.MainActivity; import android.test.ActivityInstrumentationTestCase2;public class ScreenShot extends ActivityInstrumentationTestCase2<MainActivity> { //包名 static String packageName = "com.li.xiami"; //宣告一個robotium的solo類 private Solo solo; private static String tag = "xiami"; //構造方法中寫好包名和類名,讓ActivityInstrumentationTestCase2能夠找到被測試的app //的MainActivity @SuppressWarnings("deprecation") public ScreenShot(){ //super(packageName, MainActivity.class); super(MainActivity.class); } @Before protected void setUp() throws Exception { super.setUp(); //初始化solo物件 solo = new Solo(getInstrumentation(), getActivity()); } @After protected void tearDown() throws Exception { solo.finishOpenedActivities(); } @Test public void test() { solo.clickOnButton("OK"); solo.sleep(1000); solo.takeScreenshot("123picture"); solo.sleep(3000); } }
第一次執行:
但是第一次執行的時候出現了這樣的問題:提示:Test Run Failed:java.lang.ClassNotFoundException
但是我該配置的都配置了(包括bulid path的配置,solo包的匯入以及jnuit4的包的匯入等,以及類名也檢查了好幾遍都是對的啊),後來才找到了問題的原因:
我的project.properties中的target=android-18,然後我的AndroidManifest.xml中配置的uses-sdk的targetSdkVersion是寫的17:
<uses-sdk
android:minSdkVersion="8"
android:targetSdkVersion="17" />
所以就出現了這個問題,把這個也修改成18之後,程式就能跑通了。。。
問題總結:
1、robotium中可能出現的Test run failed:classnotfoundexception的可能原因:
(1)jar包的匯入有問題,需要確認build path的Libraries和Order and Export,都需要勾選上
(2)真的是待測的apk的MainActivity的類沒找到,比如說有原始碼的情況,類名寫錯了;或者是無原始碼的情況,MainActivity的類名獲取錯誤了進而也寫錯了導致出現的這個問題
(3)就是剛才出現的這個project.properties中的target與androidManifest.xml中配置的targetSdkVersion不匹配
所有說各種問題啊,不一定報的這個exception,就一定是你class not found。。。
第二次執行:
第二次執行好不容易跑通了,但是通過DDMS裡面的File Explorer工具檢視mnt/sdcard/Robotium-Screenshots目錄下檢視是否生成了我想要的檔案,結果發現根本就沒有Robotium-Screenshots資料夾,也就是說當第一次往sdk卡里面寫東西的時候,竟然連資料夾都沒有建立起來,那就要想到是不是許可權問題?
然後就需要配置uses-permission節點:
<uses-permission android:name="android.permission.WRITE_EXTERNAL_STORAGE"></uses-permission>
配置這個節點的具體位置在:manifest裡面,但是在Application節點之外,而且在Application節點以上,否則會報錯。。。
三、Robotium的截圖處理的程式碼分析
步驟:
(1)
程式碼分析:
追本溯源,開始找路。。。
第一步跳轉到的函式:takeScreenshot(String name)
/** * Takes a screenshot and saves it with the specified name in "/sdcard/Robotium-Screenshots/". * Requires write permission (android.permission.WRITE_EXTERNAL_STORAGE) in AndroidManifest.xml of the application under test. * * @param name the name to give the screenshot * */ //上面的話翻譯下來就是:儲存的位置確定了,就是在mnt/sdcard/Robotium-Screenshots/目錄下 //但是需要寫sd卡的許可權,需要給under test的application在AndroidManifest.xml中配置permission,那麼這裡也就解釋了我上面的執行過程中第二個問題 public void takeScreenshot(String name){ takeScreenshot(name, 100); }
第二步跳轉到的函式:takeScreenshot(String name, int quality)
/** * Takes a screenshot and saves the image with the specified name in "/sdcard/Robotium-Screenshots/". * Requires write permission (android.permission.WRITE_EXTERNAL_STORAGE) in AndroidManifest.xml of the application under test. * * @param name the name to give the screenshot * @param quality the compression rate. From 0 (compress for lowest size) to 100 (compress for maximum quality) * */ //上面的話翻譯下來就是:圖片儲存位置以及讀寫許可權與第一步中相同 //引數分別表示picture的name,以及清晰度(從0到100),預設是100,當然你也可以直接在函式中呼叫這個函式,然後設定這個quality的值 public void takeScreenshot(String name, int quality){ screenshotTaker.takeScreenshot(name, quality); }
第三步跳轉到的函式:screenshotTaker.takeScreenshot(String name, int quality)
/** * Takes a screenshot and saves it in "/sdcard/Robotium-Screenshots/". * Requires write permission (android.permission.WRITE_EXTERNAL_STORAGE) in AndroidManifest.xml of the application under test. * * @param view the view to take screenshot of * @param name the name to give the screenshot image * @param quality the compression rate. From 0 (compress for lowest size) to 100 (compress for maximum quality). */ //第三步走到了一個新的類中,是screenShotTaker的類 //這個才是真正的執行Screenshot的函式,這個才是截圖的邏輯
public void takeScreenshot(final String name, final int quality) { //1、得到目前螢幕所有檢視 View decorView = getScreenshotView(); if(decorView == null) return; //2、初始化 initScreenShotSaver(); //3、例項化截圖物件 ScreenshotRunnable runnable = new ScreenshotRunnable(decorView, name, quality); //4、呼叫截圖物件的run方法 activityUtils.getCurrentActivity(false).runOnUiThread(runnable); }
第四步(1 得到螢幕所有檢視)跳轉到的函式:getScreenshotView()
/** * Gets the proper view to use for a screenshot. */ private View getScreenshotView() { //獲取到螢幕上的view View decorView = viewFetcher.getRecentDecorView(viewFetcher.getWindowDecorViews()); final long endTime = SystemClock.uptimeMillis() + Timeout.getSmallTimeout(); while (decorView == null) { final boolean timedOut = SystemClock.uptimeMillis() > endTime; if (timedOut){ return null; } sleeper.sleepMini(); decorView = viewFetcher.getRecentDecorView(viewFetcher.getWindowDecorViews()); } wrapAllGLViews(decorView); return decorView; }
第五步跳轉到的函式:viewFetcher.getWindowDecorViews()
/** * Returns the WindorDecorViews shown on the screen. * * @return the WindorDecorViews shown on the screen */ //翻譯下來就是:獲取到展示在screen上的所有WindowDecorViews,是一個View的陣列,然後這個view的陣列返回後,再作為viewFetcher.getRecentDecorView的引數 //用反射方法去獲取 View 檢視陣列 @SuppressWarnings("unchecked") public View[] getWindowDecorViews() { Field viewsField; Field instanceField; try { viewsField = windowManager.getDeclaredField("mViews"); instanceField = windowManager.getDeclaredField(windowManagerString); viewsField.setAccessible(true); instanceField.setAccessible(true); Object instance = instanceField.get(null); View[] result; if (android.os.Build.VERSION.SDK_INT >= 19) { result = ((ArrayList<View>) viewsField.get(instance)).toArray(new View[0]); } else { result = (View[]) viewsField.get(instance); } return result; } catch (SecurityException e) { e.printStackTrace(); } catch (NoSuchFieldException e) { e.printStackTrace(); } catch (IllegalArgumentException e) { e.printStackTrace(); } catch (IllegalAccessException e) { e.printStackTrace(); } return null; }
第六步跳轉到的函式:viewFetcher.getRecentDecorView(View[] views)
/** * Returns the most recent DecorView * * @param views the views to check * @return the most recent DecorView */ //翻譯下來就是:返回最近的DecorView public final View getRecentDecorView(View[] views) { if(views == null) return null; final View[] decorViews = new View[views.length]; int i = 0; View view; //通過遍歷View陣列,來得到most recent DecorView for (int j = 0; j < views.length; j++) { view = views[j]; if (view != null && view.getClass().getName() .equals("com.android.internal.policy.impl.PhoneWindow$DecorView")) { decorViews[i] = view; i++; } } return getRecentContainer(decorViews); }
第七步:(1中的獲取螢幕已經結束,看2的init操作)
/** * This method initializes the aysnc screenshot saving logic */ //翻譯下來就是:初始化一個aysnc(非同步)的sreenshot的儲存邏輯 private void initScreenShotSaver() { if(screenShotSaverThread == null || screenShotSaver == null) { //宣告一個HandlerThread物件 screenShotSaverThread = new HandlerThread("ScreenShotSaver"); screenShotSaverThread.start(); //把screenShotSaverThread捆綁到handler screenShotSaver = new ScreenShotSaver(screenShotSaverThread); } }
但是這裡用到了HandlerThread和Handler,看之。。。
第八步跳轉的函式:ScreenShotSaver(HandlerThread thread)
/** * This class is a Handler which deals with saving the screenshots on a separate thread. * * The screenshot logic by necessity has to run on the ui thread. However, in practice * it seems that saving a screenshot (with quality 100) takes approx twice as long * as taking it in the first place. * * Saving the screenshots in a separate thread like this will thus make the screenshot * process approx 3x faster as far as the main thread is concerned. * */ //翻譯下來就是:這是一個繼承自Handler,在一個單獨的thread上處理如何儲存sreenchots的類 //screenshot的邏輯必須要跑在ui執行緒上,然而,事實上,好像這個儲存screenshot反而花費了將近2倍的時間 //儲存這個screenshots在另一個執行緒中,就會使得這個處理能夠快三倍,當然是與跑在主執行緒上相比而言 private class ScreenShotSaver extends Handler { public ScreenShotSaver(HandlerThread thread) { super(thread.getLooper()); }
第九步跳轉到的函式:(3、例項化截圖物件)ScreenshotRunnable(View view, String name, int quality)
這個ScreenshotRunnable類是實現了Runnable介面中的run方法,在其中根據不同的view型別進行不同的bitmap的轉換,得到bitmap物件,之後若該bitmap不為空,則儲存到sd卡中(呼叫的 screenShotSaver.saveBitmap(BitMap b, String name, int quality)),然後這裡的這個screenShotSaver是一個繼承自Handler的類
/** * Here we have a Runnable which is responsible for taking the actual screenshot, * and then posting the bitmap to a Handler which will save it. * 這是把runnable物件放進Handler物件裡面通過得到的view去變成bitmap * 把runnable的run方法實現,首先把view轉成bitmap物件,之後呼叫之前的screenShotSaver的 * Handler物件save這個bitmap的物件 * This Runnable is run on the UI thread. */ private class ScreenshotRunnable implements Runnable { private View view; private String name; private int quality; public ScreenshotRunnable(final View _view, final String _name, final int _quality) { view = _view; name = _name; quality = _quality; } public void run() { if(view !=null){ Bitmap b; //根據是否是WebView做出不同的處理 if(view instanceof WebView){ b = getBitmapOfWebView((WebView) view); } else{ b = getBitmapOfView(view); } if(b != null) //如果bitmap物件不為空,就存到sd卡里 screenShotSaver.saveBitmap(b, name, quality); else Log.d(LOG_TAG, "NULL BITMAP!!"); } } }
第十步跳轉到的函式:saveBitmap(Bitmap bitmap, String name, int quality),這裡會產生一個message,然後通過handlemessage來處理這個message
/** * This method posts a Bitmap with meta-data to the Handler queue. * * @param bitmap the bitmap to save * @param name the name of the file * @param quality the compression rate. From 0 (compress for lowest size) to 100 (compress for maximum quality). */ public void saveBitmap(Bitmap bitmap, String name, int quality) { //獲取一個message物件,然後把bitmap的資訊儲存到這個message中 //將這個message發出去,傳送到looper,然後這個message會被handleMessage接收 //這裡沒有直接儲存,而使用message,是想要用到looper,使用looper的好處是:可以短時間內放10個左右的截圖
Message message = this.obtainMessage(); message.arg1 = quality; message.obj = bitmap; message.getData().putString("name", name); this.sendMessage(message); }
具體的handleMessage函式如下所示,也是位於這個screenShotSaver的類中:
/** * Here we process the Handler queue and save the bitmaps. * * @param message A Message containing the bitmap to save, and some metadata. */ public void handleMessage(Message message) { //複寫Handler的handleMessage方法,然後獲取到message物件,之後呼叫saveFile方法方法儲存bitmap物件 String name = message.getData().getString("name"); int quality = message.arg1; Bitmap b = (Bitmap)message.obj; if(b != null) { saveFile(name, b, quality); b.recycle(); } else { Log.d(LOG_TAG, "NULL BITMAP!!"); } }
接下來就到了saveFile的函式中:
/** * Saves a file. * * @param name the name of the file * @param b the bitmap to save * @param quality the compression rate. From 0 (compress for lowest size) to 100 (compress for maximum quality). * */ private void saveFile(String name, Bitmap b, int quality){ //構造一個File輸出流,寫bitmap物件到sd卡 FileOutputStream fos = null; String fileName = getFileName(name); // File directory = new File(Environment.getExternalStorageDirectory() + "/Robotium-Screenshots/"); directory.mkdir(); File fileToSave = new File(directory,fileName); try { //初始化一個File的輸入輸出類,用以進行file的儲存,之後呼叫compress方法寫入 fos = new FileOutputStream(fileToSave); if (b.compress(Bitmap.CompressFormat.JPEG, quality, fos) == false) Log.d(LOG_TAG, "Compress/Write failed"); fos.flush(); fos.close(); } catch (Exception e) { Log.d(LOG_TAG, "Can‘t save the screenshot! Requires write permission (android.permission.WRITE_EXTERNAL_STORAGE) in AndroidManifest.xml of the application under test."); e.printStackTrace(); } }