ContentProvider

Construct a ContentProvider instance. Content providers must be declared in the manifest, accessed with ContentResolver, and created automatically by the system, so applications usually do not create ContentProvider instances directly.

At construction time, the object is uninitialized, and most fields and methods are unavailable. Subclasses should initialize themselves in onCreate(), not the constructor.

Content providers are created on the application main thread at application launch time. The constructor must not perform lengthy operations, or application startup will be delayed.

Override this to handle requests to perform a batch of operations, or the default implementation will iterate over the operations and call apply(ContentProvider, ContentProviderResult[], int) on each of them. If all calls to apply(ContentProvider, ContentProviderResult[], int) succeed then a ContentProviderResult array with as many elements as there were operations will be returned. If any of the calls fail, it is up to the implementation how many of the others take effect. This method can be called from multiple threads, as described in Processes and Threads.

After being instantiated, this is called to tell the content provider about itself.

Override this to handle requests to insert a set of new rows, or the default implementation will iterate over the values and call insert(Uri, ContentValues) on each of them. As a courtesy, call notifyChange() after inserting. This method can be called from multiple threads, as described in Processes and Threads.

Call a provider-defined method. This can be used to implement interfaces that are cheaper and/or unnatural for a table-like model.

WARNING: The framework does no permission checking on this entry into the content provider besides the basic ability for the application to get access to the provider at all. For example, it has no idea whether the call being executed may read or write data in the provider, so can't enforce those individual permissions. Any implementation of this method must do its own permission checks on incoming calls to make sure they are allowed.

Implement this to support canonicalization of URIs that refer to your content provider. A canonical URI is one that can be transported across devices, backup/restore, and other contexts, and still be able to refer to the same data item. Typically this is implemented by adding query params to the URI allowing the content provider to verify that an incoming canonical URI references the same data as it was originally intended for and, if it doesn't, to find that data (if it exists) in the current environment.

For example, if the content provider holds people and a normal URI in it is created with a row index into that people database, the cananical representation may have an additional query param at the end which specifies the name of the person it is intended for. Later calls into the provider with that URI will look up the row of that URI's base index and, if it doesn't match or its entry's name doesn't match the name in the query param, perform a query on its database to find the correct row to operate on.

If you implement support for canonical URIs, all incoming calls with URIs (including this one) must perform this verification and recovery of any canonical URIs they receive. In addition, you must also implement uncanonicalize(Uri) to strip the canonicalization of any of these URIs.

The default implementation of this method returns null, indicating that canonical URIs are not supported.

Implement this to handle requests to delete one or more rows. The implementation should apply the selection clause when performing deletion, allowing the operation to affect multiple rows in a directory. As a courtesy, call notifyChange() after deleting. This method can be called from multiple threads, as described in Processes and Threads.

The implementation is responsible for parsing out a row ID at the end of the URI, if a specific row is being deleted. That is, the client would pass in content://contacts/people/22 and the implementation is responsible for parsing the record number (22) when creating a SQL statement.

Print the Provider's state into the given stream. This gets invoked if you run "adb shell dumpsys activity provider <provider_component_name>".

Return the package name of the caller that initiated the request being processed on the current thread. The returned package will have been verified to belong to the calling UID. Returns null if not currently processing a request.

This will always return null when processing getType(Uri) or getStreamTypes(Uri, String) requests.

Retrieves the Context this provider is running in. Only available once onCreate() has been called -- this will return null in the constructor.

Return the path-based permissions required for read and/or write access to this content provider. This method can be called from multiple threads, as described in Processes and Threads.

Return the name of the permission required for read-only access to this content provider. This method can be called from multiple threads, as described in Processes and Threads.

Called by a client to determine the types of data streams that this content provider supports for the given URI. The default implementation returns null, meaning no types. If your content provider stores data of a particular type, return that MIME type if it matches the given mimeTypeFilter. If it can perform type conversions, return an array of all supported MIME types that match mimeTypeFilter.

Implement this to handle requests for the MIME type of the data at the given URI. The returned MIME type should start with vnd.android.cursor.item for a single record, or vnd.android.cursor.dir/ for multiple items. This method can be called from multiple threads, as described in Processes and Threads.

Note that there are no permissions needed for an application to access this information; if your content provider requires read and/or write permissions, or is not exported, all applications can still call this method regardless of their access permissions. This allows them to retrieve the MIME type for a URI when dispatching intents.

Return the name of the permission required for read/write access to this content provider. This method can be called from multiple threads, as described in Processes and Threads.

Implement this to handle requests to insert a new row. As a courtesy, call notifyChange() after inserting. This method can be called from multiple threads, as described in Processes and Threads.

Called by the system when the device configuration changes while your component is running. Note that, unlike activities, other components are never restarted when a configuration changes: they must always deal with the results of the change, such as by re-retrieving resources.

At the time that this function has been called, your Resources object will have been updated to return resource values matching the new configuration.

For more information, read Handling Runtime Changes. This method is always called on the application main thread, and must not perform lengthy operations.

The default content provider implementation does nothing. Override this method to take appropriate action. (Content providers do not usually care about things like screen orientation, but may want to know about locale changes.)

Implement this to initialize your content provider on startup. This method is called for all registered content providers on the application main thread at application launch time. It must not perform lengthy operations, or application startup will be delayed.

You should defer nontrivial initialization (such as opening, upgrading, and scanning databases) until the content provider is used (via query(Uri, String[], String, String[], String), insert(Uri, ContentValues), etc). Deferred initialization keeps application startup fast, avoids unnecessary work if the provider turns out not to be needed, and stops database errors (such as a full disk) from halting application launch.

If you use SQLite, SQLiteOpenHelper is a helpful utility class that makes it easy to manage databases, and will automatically defer opening until first use. If you do use SQLiteOpenHelper, make sure to avoid calling getReadableDatabase() or getWritableDatabase() from this method. (Instead, override onOpen(SQLiteDatabase) to initialize the database when it is first opened.)

This is called when the overall system is running low on memory, and actively running processes should trim their memory usage. While the exact point at which this will be called is not defined, generally it will happen when all background process have been killed. That is, before reaching the point of killing processes hosting service and foreground UI that we would like to avoid killing.

You should implement this method to release any caches or other unnecessary resources you may be holding on to. The system will perform a garbage collection for you after returning from this method.

Preferably, you should implement onTrimMemory(int) from ComponentCallbacks2 to incrementally unload your resources based on various levels of memory demands. That API is available for API level 14 and higher, so you should only use this onLowMemory() method as a fallback for older versions, which can be treated the same as onTrimMemory(int) with the TRIM_MEMORY_COMPLETE level.

The default content provider implementation does nothing. Subclasses may override this method to take appropriate action.

Called when the operating system has determined that it is a good time for a process to trim unneeded memory from its process. This will happen for example when it goes in the background and there is not enough memory to keep as many background processes running as desired. You should never compare to exact values of the level, since new intermediate values may be added -- you will typically want to compare if the value is greater or equal to a level you are interested in.

To retrieve the processes current trim level at any point, you can use ActivityManager.getMyMemoryState(RunningAppProcessInfo).

This is like openFile(Uri, String), but can be implemented by providers that need to be able to return sub-sections of files, often assets inside of their .apk. This method can be called from multiple threads, as described in Processes and Threads.

If you implement this, your clients must be able to deal with such file slices, either directly with openAssetFileDescriptor(Uri, String), or by using the higher-level ContentResolver.openInputStream or ContentResolver.openOutputStream methods.

The returned AssetFileDescriptor can be a pipe or socket pair to enable streaming of data.

If you are implementing this to return a full file, you should create the AssetFileDescriptor with UNKNOWN_LENGTH to be compatible with applications that cannot handle sub-sections of files.

For use in Intents, you will want to implement getType(Uri) to return the appropriate MIME type for the data returned here with the same URI. This will allow intent resolution to automatically determine the data MIME type and select the appropriate matching targets as part of its operation.

For better interoperability with other applications, it is recommended that for any URIs that can be opened, you also support queries on them containing at least the columns specified by OpenableColumns.

This is like openFile(Uri, String), but can be implemented by providers that need to be able to return sub-sections of files, often assets inside of their .apk. This method can be called from multiple threads, as described in Processes and Threads.

If you implement this, your clients must be able to deal with such file slices, either directly with openAssetFileDescriptor(Uri, String), or by using the higher-level ContentResolver.openInputStream or ContentResolver.openOutputStream methods.

The returned AssetFileDescriptor can be a pipe or socket pair to enable streaming of data.

If you are implementing this to return a full file, you should create the AssetFileDescriptor with UNKNOWN_LENGTH to be compatible with applications that cannot handle sub-sections of files.

For use in Intents, you will want to implement getType(Uri) to return the appropriate MIME type for the data returned here with the same URI. This will allow intent resolution to automatically determine the data MIME type and select the appropriate matching targets as part of its operation.

For better interoperability with other applications, it is recommended that for any URIs that can be opened, you also support queries on them containing at least the columns specified by OpenableColumns.

Override this to handle requests to open a file blob. The default implementation always throws FileNotFoundException. This method can be called from multiple threads, as described in Processes and Threads.

This method returns a ParcelFileDescriptor, which is returned directly to the caller. This way large data (such as images and documents) can be returned without copying the content.

The returned ParcelFileDescriptor is owned by the caller, so it is their responsibility to close it when done. That is, the implementation of this method should create a new ParcelFileDescriptor for each call.

If opened with the exclusive "r" or "w" modes, the returned ParcelFileDescriptor can be a pipe or socket pair to enable streaming of data. Opening with the "rw" or "rwt" modes implies a file on disk that supports seeking.

If you need to detect when the returned ParcelFileDescriptor has been closed, or if the remote process has crashed or encountered some other error, you can use open(File, int, android.os.Handler, android.os.ParcelFileDescriptor.OnCloseListener), createReliablePipe(), or createReliableSocketPair().

For use in Intents, you will want to implement getType(Uri) to return the appropriate MIME type for the data returned here with the same URI. This will allow intent resolution to automatically determine the data MIME type and select the appropriate matching targets as part of its operation.

For better interoperability with other applications, it is recommended that for any URIs that can be opened, you also support queries on them containing at least the columns specified by OpenableColumns. You may also want to support other common columns if you have additional meta-data to supply, such as DATE_ADDED in MediaStore.MediaColumns.

Override this to handle requests to open a file blob. The default implementation always throws FileNotFoundException. This method can be called from multiple threads, as described in Processes and Threads.

This method returns a ParcelFileDescriptor, which is returned directly to the caller. This way large data (such as images and documents) can be returned without copying the content.

The returned ParcelFileDescriptor is owned by the caller, so it is their responsibility to close it when done. That is, the implementation of this method should create a new ParcelFileDescriptor for each call.

If opened with the exclusive "r" or "w" modes, the returned ParcelFileDescriptor can be a pipe or socket pair to enable streaming of data. Opening with the "rw" or "rwt" modes implies a file on disk that supports seeking.

If you need to detect when the returned ParcelFileDescriptor has been closed, or if the remote process has crashed or encountered some other error, you can use open(File, int, android.os.Handler, android.os.ParcelFileDescriptor.OnCloseListener), createReliablePipe(), or createReliableSocketPair().

For use in Intents, you will want to implement getType(Uri) to return the appropriate MIME type for the data returned here with the same URI. This will allow intent resolution to automatically determine the data MIME type and select the appropriate matching targets as part of its operation.

For better interoperability with other applications, it is recommended that for any URIs that can be opened, you also support queries on them containing at least the columns specified by OpenableColumns. You may also want to support other common columns if you have additional meta-data to supply, such as DATE_ADDED in MediaStore.MediaColumns.

A helper function for implementing openTypedAssetFile(Uri, String, Bundle), for creating a data pipe and background thread allowing you to stream generated data back to the client. This function returns a new ParcelFileDescriptor that should be returned to the caller (the caller is responsible for closing it).

Called by a client to open a read-only stream containing data of a particular MIME type. This is like openAssetFile(Uri, String), except the file can only be read-only and the content provider may perform data conversions to generate data of the desired type.

The default implementation compares the given mimeType against the result of getType(Uri) and, if they match, simply calls openAssetFile(Uri, String).

See ClipData for examples of the use and implementation of this method.

The returned AssetFileDescriptor can be a pipe or socket pair to enable streaming of data.

For better interoperability with other applications, it is recommended that for any URIs that can be opened, you also support queries on them containing at least the columns specified by OpenableColumns. You may also want to support other common columns if you have additional meta-data to supply, such as DATE_ADDED in MediaStore.MediaColumns.

Called by a client to open a read-only stream containing data of a particular MIME type. This is like openAssetFile(Uri, String), except the file can only be read-only and the content provider may perform data conversions to generate data of the desired type.

The default implementation compares the given mimeType against the result of getType(Uri) and, if they match, simply calls openAssetFile(Uri, String).

See ClipData for examples of the use and implementation of this method.

The returned AssetFileDescriptor can be a pipe or socket pair to enable streaming of data.

For better interoperability with other applications, it is recommended that for any URIs that can be opened, you also support queries on them containing at least the columns specified by OpenableColumns. You may also want to support other common columns if you have additional meta-data to supply, such as DATE_ADDED in MediaStore.MediaColumns.

Implement this to handle query requests from clients. This method can be called from multiple threads, as described in Processes and Threads.

Example client call:

// Request a specific record.
 Cursor managedCursor = managedQuery(
                ContentUris.withAppendedId(Contacts.People.CONTENT_URI, 2),
                projection,    // Which columns to return.
                null,          // WHERE clause.
                null,          // WHERE clause value substitution
                People.NAME + " ASC");   // Sort order.

// SQLiteQueryBuilder is a helper class that creates the
        // proper SQL syntax for us.
        SQLiteQueryBuilder qBuilder = new SQLiteQueryBuilder();

        // Set the table we're querying.
        qBuilder.setTables(DATABASE_TABLE_NAME);

        // If the query ends in a specific record number, we're
        // being asked for a specific record, so set the
        // WHERE clause in our query.
        if((URI_MATCHER.match(uri)) == SPECIFIC_MESSAGE){
            qBuilder.appendWhere("_id=" + uri.getPathLeafId());
        }

        // Make the query.
        Cursor c = qBuilder.query(mDb,
                projection,
                selection,
                selectionArgs,
                groupBy,
                having,
                sortOrder);
        c.setNotificationUri(getContext().getContentResolver(), uri);
        return c;

Implement this to handle query requests from clients with support for cancellation. This method can be called from multiple threads, as described in Processes and Threads.

Example client call:

// Request a specific record.
 Cursor managedCursor = managedQuery(
                ContentUris.withAppendedId(Contacts.People.CONTENT_URI, 2),
                projection,    // Which columns to return.
                null,          // WHERE clause.
                null,          // WHERE clause value substitution
                People.NAME + " ASC");   // Sort order.

// SQLiteQueryBuilder is a helper class that creates the
        // proper SQL syntax for us.
        SQLiteQueryBuilder qBuilder = new SQLiteQueryBuilder();

        // Set the table we're querying.
        qBuilder.setTables(DATABASE_TABLE_NAME);

        // If the query ends in a specific record number, we're
        // being asked for a specific record, so set the
        // WHERE clause in our query.
        if((URI_MATCHER.match(uri)) == SPECIFIC_MESSAGE){
            qBuilder.appendWhere("_id=" + uri.getPathLeafId());
        }

        // Make the query.
        Cursor c = qBuilder.query(mDb,
                projection,
                selection,
                selectionArgs,
                groupBy,
                having,
                sortOrder);
        c.setNotificationUri(getContext().getContentResolver(), uri);
        return c;

If you implement this method then you must also implement the version of query(Uri, String[], String, String[], String) that does not take a cancellation signal to ensure correct operation on older versions of the Android Framework in which the cancellation signal overload was not available.

Implement this to shut down the ContentProvider instance. You can then invoke this method in unit tests.

Android normally handles ContentProvider startup and shutdown automatically. You do not need to start up or shut down a ContentProvider. When you invoke a test method on a ContentProvider, however, a ContentProvider instance is started and keeps running after the test finishes, even if a succeeding test instantiates another ContentProvider. A conflict develops because the two instances are usually running against the same underlying data source (for example, an sqlite database).

Implementing shutDown() avoids this conflict by providing a way to terminate the ContentProvider. This method can also prevent memory leaks from multiple instantiations of the ContentProvider, and it can ensure unit test isolation by allowing you to completely clean up the test fixture before moving on to the next test.

Remove canonicalization from canonical URIs previously returned by canonicalize(Uri). For example, if your implementation is to add a query param to canonicalize a URI, this method can simply trip any query params on the URI. The default implementation always returns the same url that was passed in.

Implement this to handle requests to update one or more rows. The implementation should update all rows matching the selection to set the columns according to the provided values map. As a courtesy, call notifyChange() after updating. This method can be called from multiple threads, as described in Processes and Threads.

Returns true if this instance is a temporary content provider.

Convenience for subclasses that wish to implement openFile(Uri, String) by looking up a column named "_data" at the given URI.

Change the path-based permission required to read and/or write data in the content provider. This is normally set for you from its manifest information when the provider is first created.

Change the permission required to read data from the content provider. This is normally set for you from its manifest information when the provider is first created.

Change the permission required to read and write data in the content provider. This is normally set for you from its manifest information when the provider is first created.