Updater swallowing important exceptions
In order to prevent network errors causing crashes when updating repositories, the repo updater catches exceptions (and shows them in a Toast). I suspect that it catches very general exceptions (i.e. Exception) because there are many different types of network problems which could be of interest. However, it is currently also swallowing SQLiteConstraintExceptions such as the following:
SQLiteLog E (1555) abort at 56 in [INSERT INTO fdroid_apk SELECT * FROM temp_update_db.temp_fdroid_apk]: UNIQUE constr
aint failed: fdroid_apk.appId, fdroid_apk.vercode, fdroid_apk.repo
UpdateService E Exception during update processing
E android.database.sqlite.SQLiteConstraintException: UNIQUE constraint failed: fdroid_apk.appId, fdroid_apk.
vercode, fdroid_apk.repo (code 1555)
E at android.database.sqlite.SQLiteConnection.nativeExecuteForChangedRowCount(Native Method)
E at android.database.sqlite.SQLiteConnection.executeForChangedRowCount(SQLiteConnection.java:736)
E at android.database.sqlite.SQLiteSession.executeForChangedRowCount(SQLiteSession.java:754)
E at android.database.sqlite.SQLiteStatement.executeUpdateDelete(SQLiteStatement.java:64)
E at android.database.sqlite.SQLiteDatabase.executeSql(SQLiteDatabase.java:1676)
E at android.database.sqlite.SQLiteDatabase.execSQL(SQLiteDatabase.java:1605)
E at org.fdroid.fdroid.data.TempAppProvider.commitTable(TempAppProvider.java:182)
E at org.fdroid.fdroid.data.TempAppProvider.insert(TempAppProvider.java:116)
E at android.content.ContentProvider$Transport.insert(ContentProvider.java:239)
E at android.content.ContentResolver.insert(ContentResolver.java:1207)
E at org.fdroid.fdroid.data.TempAppProvider$Helper.commitAppsAndApks(TempAppProvider.java:99)
E at org.fdroid.fdroid.data.RepoPersister.commit(RepoPersister.java:70)
E at org.fdroid.fdroid.RepoUpdater.commitToDb(RepoUpdater.java:237)
E at org.fdroid.fdroid.RepoUpdater.processDownloadedFile(RepoUpdater.java:213)
E at org.fdroid.fdroid.RepoUpdater.update(RepoUpdater.java:150)
E at org.fdroid.fdroid.UpdateService.onHandleIntent(UpdateService.java:355)
E at android.app.IntentService$ServiceHandler.handleMessage(IntentService.java:65)
E at android.os.Handler.dispatchMessage(Handler.java:102)
E at android.os.Looper.loop(Looper.java:135)
E at android.os.HandlerThread.run(HandlerThread.java:61)In this particular case, I investigated and it seems to be that my database got itself into a sad state, quite likely while I was toying with development of the database. I don't think it is a bug in production.
However, problems such as this that do make it into production should be logged to ACRA at least. I suspect this could be done by instead catching only IOException.
Thoughts @mvdan, @eighthave , others?
Activity
There was just an ACRA report that might be related:
USER_COMMENT=update hung during unpacking of update, force-stopped, then restarted ANDROID_VERSION=6.0.1 APP_VERSION_NAME=0.100.1 BRAND=google PHONE_MODEL=Nexus 5X CUSTOM_DATA= STACK_TRACE=java.lang.RuntimeException: An error occurred while executing doInBackground() at android.os.AsyncTask$3.done(AsyncTask.java:309) at java.util.concurrent.FutureTask.finishCompletion(FutureTask.java:354) at java.util.concurrent.FutureTask.setException(FutureTask.java:223) at java.util.concurrent.FutureTask.run(FutureTask.java:242) at android.os.AsyncTask$SerialExecutor$1.run(AsyncTask.java:234) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1113) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:588) at java.lang.Thread.run(Thread.java:818) Caused by: android.database.sqlite.SQLiteDatabaseLockedException: database is locked (code 5) at android.database.sqlite.SQLiteConnection.nativeExecute(Native Method) at android.database.sqlite.SQLiteConnection.execute(SQLiteConnection.java:553) at android.database.sqlite.SQLiteSession.beginTransactionUnchecked(SQLiteSession.java:323) at android.database.sqlite.SQLiteSession.beginTransaction(SQLiteSession.java:298) at android.database.sqlite.SQLiteDatabase.beginTransaction(SQLiteDatabase.java:505) at android.database.sqlite.SQLiteDatabase.beginTransaction(SQLiteDatabase.java:416) at org.fdroid.fdroid.data.FDroidProvider.applyBatch(FDroidProvider.java:64) at android.content.ContentProvider$Transport.applyBatch(ContentProvider.java:315) at android.content.ContentProviderClient.applyBatch(ContentProviderClient.java:419) at android.content.ContentResolver.applyBatch(ContentResolver.java:1270) at org.fdroid.fdroid.data.InstalledAppCacheUpdater.updateCache(InstalledAppCacheUpdater.java:106) at org.fdroid.fdroid.data.InstalledAppCacheUpdater.update(InstalledAppCacheUpdater.java:72) at org.fdroid.fdroid.data.InstalledAppCacheUpdater.access$100(InstalledAppCacheUpdater.java:26) at org.fdroid.fdroid.data.InstalledAppCacheUpdater$PostponedWorker.doInBackground(InstalledAppCacheUpdater.java:180) at org.fdroid.fdroid.data.InstalledAppCacheUpdater$PostponedWorker.doInBackground(InstalledAppCacheUpdater.java:173) at android.os.AsyncTask$2.call(AsyncTask.java:295) at java.util.concurrent.FutureTask.run(FutureTask.java:237) ... 4 more android.database.sqlite.SQLiteDatabaseLockedException: database is locked (code 5) at android.database.sqlite.SQLiteConnection.nativeExecute(Native Method) at android.database.sqlite.SQLiteConnection.execute(SQLiteConnection.java:553) at android.database.sqlite.SQLiteSession.beginTransactionUnchecked(SQLiteSession.java:323) at android.database.sqlite.SQLiteSession.beginTransaction(SQLiteSession.java:298) at android.database.sqlite.SQLiteDatabase.beginTransaction(SQLiteDatabase.java:505) at android.database.sqlite.SQLiteDatabase.beginTransaction(SQLiteDatabase.java:416) at org.fdroid.fdroid.data.FDroidProvider.applyBatch(FDroidProvider.java:64) at android.content.ContentProvider$Transport.applyBatch(ContentProvider.java:315) at android.content.ContentProviderClient.applyBatch(ContentProviderClient.java:419) at android.content.ContentResolver.applyBatch(ContentResolver.java:1270) at org.fdroid.fdroid.data.InstalledAppCacheUpdater.updateCache(InstalledAppCacheUpdater.java:106) at org.fdroid.fdroid.data.InstalledAppCacheUpdater.update(InstalledAppCacheUpdater.java:72) at org.fdroid.fdroid.data.InstalledAppCacheUpdater.access$100(InstalledAppCacheUpdater.java:26) at org.fdroid.fdroid.data.InstalledAppCacheUpdater$PostponedWorker.doInBackground(InstalledAppCacheUpdater.java:180) at org.fdroid.fdroid.data.InstalledAppCacheUpdater$PostponedWorker.doInBackground(InstalledAppCacheUpdater.java:173) at android.os.AsyncTask$2.call(AsyncTask.java:295) at java.util.concurrent.FutureTask.run(FutureTask.java:237) at android.os.AsyncTask$SerialExecutor$1.run(AsyncTask.java:234) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1113) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:588) at java.lang.Thread.run(Thread.java:818)
The above crash is unrelated. This issue is specifically about the repo update process swallowing exceptions (and hence not getting to ACRA) whereas the report above is from the "update installed app cache" process which doesn't swallow exceptions (hence the bug report in ACRA).
I started on a MR, but then realised it was a little more tricky because the notification for updating repositories hangs around like a bad smell if you crash hard during a repo update. It is a sticky update so the user can't clear it manually either. Given it is a bit more work than I'd've liked, lets ignore this for now and I'll see if we can come up with something to address it for a later release.
Milestone changed to %0.102
Milestone changed to %0.103 - UX Overhaul
changed milestone to %1.1
