DownloadProvider 源码详细分析

DownloadProvider 简介
DownloadProvider 是Android提供的DownloadManager的增强版，亮点是支持断点下载，提供了“开始下载”，“暂停下载”，“重新下载”，“删除下载”接口。源码下载地址

DownloadProvider 详细分析

DownloadProvider开始下载的是由DownloadManager 的 enqueue方法启动的，启动一个新的下载任务的时序图 


开始新的下载时候会调用DownloadManager的enqueue方法，然后再执行DownloadProvider的insert方法，将下载信息写入数据库，包括下载链接地址等，然后再调用DownloadService的onCreate或者onStartCommand方法。 DownloadProvider类是非常重要的类，所有操作都跟此类有关，因为要保存下载状态。 在分析DownloadProvider的insert方法前，先看看insert方法的源码
?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
@Override
public Uri insert(final Uri uri, final ContentValues values) {
    checkInsertPermissions(values);
    SQLiteDatabase db = mOpenHelper.getWritableDatabase();
 
    // note we disallow inserting into ALL_DOWNLOADS
    int match = sURIMatcher.match(uri);
    if (match != MY_DOWNLOADS) {
        Log.d(Constants.TAG, "calling insert on an unknown/invalid URI: "
                + uri);
        throw new IllegalArgumentException("Unknown/Invalid URI " + uri);
    }
 
    ContentValues filteredValues = new ContentValues();
 
    ......
 
    Integer dest = values.getAsInteger(Downloads.COLUMN_DESTINATION);
    if (dest != null) {
     
        ......
         
    }
    Integer vis = values.getAsInteger(Downloads.COLUMN_VISIBILITY);
    if (vis == null) {
        if (dest == Downloads.DESTINATION_EXTERNAL) {
            filteredValues.put(Downloads.COLUMN_VISIBILITY,
                    Downloads.VISIBILITY_VISIBLE_NOTIFY_COMPLETED);
        } else {
            filteredValues.put(Downloads.COLUMN_VISIBILITY,
                    Downloads.VISIBILITY_HIDDEN);
        }
    } else {
        filteredValues.put(Downloads.COLUMN_VISIBILITY, vis);
    }
    ......
 
    String pckg = values.getAsString(Downloads.COLUMN_NOTIFICATION_PACKAGE);
    String clazz = values.getAsString(Downloads.COLUMN_NOTIFICATION_CLASS);
    if (pckg != null && (clazz != null || isPublicApi)) {
        ......
    }
    ......
 
    //启动下载服务
    Context context = getContext();
    context.startService(new Intent(context, DownloadService.class));
 
    //插入数据库
    long rowID = db.insert(DB_TABLE, null, filteredValues);
    if (rowID == -1) {
        Log.d(Constants.TAG, "couldn't insert into downloads database");
        return null;
    }
 
    insertRequestHeaders(db, rowID, values);
    //启动下载服务
    context.startService(new Intent(context, DownloadService.class));
    notifyContentChanged(uri, match);
    return ContentUris.withAppendedId(Downloads.CONTENT_URI, rowID);
}
每次开始一个新的下载任务，都会插入数据库，然后启动启动下载服务类DownloadService，所以真正处理下载的是后台服务DownloadService，DownloadService中有个下载线程类DownloadThread，DownloadService时序图 


如果DownloadService没有启动将会执行onCreate()------>onStartCommand()方法，否则执行onStartCommand()方法。然后执行updateFromProvider()方法启动UpdateThread线程，准备启动DownloadThread线程。 分析UpdateThread的run方法前先看看run方法的源码：
?
1
2
3
4
5
6
7
8
9
10
11
public void run() {
    Process.setThreadPriority(Process.THREAD_PRIORITY_BACKGROUND);
 
    //如果数据里的存储的达到了1000以上时候，将会删除status>200即失败的记录
    trimDatabase();
    removeSpuriousFiles();
 
    boolean keepService = false;
    // for each update from the database, remember which download is
    // supposed to get restarted soonest in the future
    long wakeUp = Long.MAX_VALUE;
?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
        //会一直在此循环，直到启动完所有下载任务
    for (;;) {
        synchronized (DownloadService.this) {
            if (mUpdateThread != this) {
                throw new IllegalStateException(
                        "multiple UpdateThreads in DownloadService");
            }
            if (!mPendingUpdate) {
                mUpdateThread = null;
                if (!keepService) {
                    stopSelf();
                }
                if (wakeUp != Long.MAX_VALUE) {
                    scheduleAlarm(wakeUp);
                }
                return;
            }
            mPendingUpdate = false;
        }
 
        long now = mSystemFacade.currentTimeMillis();
        keepService = false;
        wakeUp = Long.MAX_VALUE;
        Set<long> idsNoLongerInDatabase = new HashSet<long>(
                mDownloads.keySet());
 
        Cursor cursor = getContentResolver().query(
                Downloads.ALL_DOWNLOADS_CONTENT_URI, null, null, null,
                null);
        if (cursor == null) {
            continue;
        }
        try {
            DownloadInfo.Reader reader = new DownloadInfo.Reader(
                    getContentResolver(), cursor);
            int idColumn = cursor.getColumnIndexOrThrow(Downloads._ID);
 
            for (cursor.moveToFirst(); !cursor.isAfterLast(); cursor
                    .moveToNext()) {
                long id = cursor.getLong(idColumn);
                idsNoLongerInDatabase.remove(id);
                DownloadInfo info = mDownloads.get(id);
                if (info != null) {
                    updateDownload(reader, info, now);
                } else {
                    info = insertDownload(reader, now);
                }
                if (info.hasCompletionNotification()) {
                    keepService = true;
                }
                long next = info.nextAction(now);
                if (next == 0) {
                    keepService = true;
                } else if (next > 0 && next < wakeUp) {
                    wakeUp = next;
                }
            }
        } finally {
            cursor.close();
        }
 
        for (Long id : idsNoLongerInDatabase) {
            deleteDownload(id);
        }
 
        // is there a need to start the DownloadService? yes, if there
        // are rows to be deleted.
 
        for (DownloadInfo info : mDownloads.values()) {
            if (info.mDeleted) {
                keepService = true;
                break;
            }
        }
 
        mNotifier.updateNotification(mDownloads.values());
 
        // look for all rows with deleted flag set and delete the rows
        // from the database
        // permanently
        for (DownloadInfo info : mDownloads.values()) {
            if (info.mDeleted) {
                Helpers.deleteFile(getContentResolver(), info.mId,
                        info.mFileName, info.mMimeType);
            }
        }
    }
}</long></long>
UpdateThread线程负责从数据库中获取下载任务，该线程不会每次都新建新的对象，只有UpdateThread为空时候才会新建，在此线程的run()方法中有两个for循环，外循环是从数据获取下载任务，确保插入数据库的任务都能被启动，直到启动完所有的下载任务才会退出。内循环是遍历从数据库获取的没完成或者刚开始的下载任务，启动下载。 DownloadThrea线程是真正负责下载的线程，每次启动一个任务都会创建一个新的下载线程对象（对手机来说会耗很大的CPU,所有要加上同步锁或者线程池来控制同时下载的数量），看看DownloadThread run方法的时序图： 
从这个时序图可以看出，这里涉及的源码比较多，在这没有写，看的时候一定要对照的源码来看。其实在下载的时候会发生很多的异常，如网络异常，内存卡容量不足等，所以捕获异常很重要的，捕获后进行相关的处理，这也是体现出考虑全面的地方。 


【推广】 免费学中医，健康全家人
原文地址：https://www.cnblogs.com/bigben0123/p/4300671.html