Flexible utilities to control and animate zoom and translation of Views and much more - either programmatically or through touch events.

compile 'com.otaliastudios:zoomlayout:1.3.0'

Want to see ZoomLayout in action? Try the demo app or take a look at ViewPrinter, a printing library heavily based on this.

• ZoomLayout : a container that supports 2D pan and zoom to a View hierarchy, even supporting clicks.
• ZoomImageView : (yet another) ImageView that supports 2D pan and zoom.
• Lightweight, no dependencies
• API 16

In fact, both ZoomLayout and ZoomImageView are just very simple implementations of the internal ZoomEngine. The zoom engine lets you animate everything through constant updates, as long as you feed it with touch events, with a Matrix-based mechanism that makes it very flexible.

A container for view hierarchies that can be panned or zoomed.

<com.otaliastudios.zoom.ZoomLayout
    android:layout_width="match_parent"
    android:layout_height="match_parent"
    android:scrollbars="vertical|horizontal"                               
    app:overScrollHorizontal="true"
    app:overScrollVertical="true"
    app:overPinchable="true"
    app:horizontalPanEnabled="true"
    app:verticalPanEnabled="true"
    app:zoomEnabled="true"
    app:minZoom="0.7"
    app:minZoomType="zoom"
    app:maxZoom="3.0"
    app:maxZoomType="zoom"
    app:hasClickableChildren="false">

    <!-- Content here. -->

</com.otaliastudios.zoom.ZoomLayout>

ZoomLayout supports only a single child, but that child can have as many children as you wish. If any of these children is clickable or should react to touch events, you are required to set hasClickableChildren to true. This is off by default because it is more expensive in terms of performance.

The child view will be measured as wrap content with no limits in space, as in a 2D scroll view. So it can be as big as you want.

The zoom layout will forward all API calls to the internal engine. See engine docs. You can also get the backing engine using zoomLayout.getEngine().

zoomLayout.panTo(x, y, true); // Shorthand for zoomLayout.getEngine().panTo(x, y, true)
zoomLayout.panBy(deltaX, deltaY, true);
zoomLayout.zoomTo(zoom, true);
zoomLayout.zoomBy(factor, true);
zoomLayout.realZoomTo(realZoom, true);
zoomLayout.moveTo(zoom, x, y, true);

An ImageView implementation to control pan and zoom over its Drawable or Bitmap.

<com.otaliastudios.zoom.ZoomImageView
    android:layout_width="match_parent"
    android:layout_height="match_parent"
    android:scrollbars="vertical|horizontal"                                                                 
    app:overScrollHorizontal="true"
    app:overScrollVertical="true"
    app:overPinchable="true"
    app:horizontalPanEnabled="true"
    app:verticalPanEnabled="true"
    app:zoomEnabled="true"
    app:minZoom="0.7"
    app:minZoomType="zoom"
    app:maxZoom="3.0"
    app:maxZoomType="zoom"/>

There is nothing surprising going on. Just call setImageDrawable() and you are done.

Presumably ZoomImageView won't work if:

• the drawable has no intrinsic dimensions
• the view has wrap_content as a dimension
• you change the scaleType (read later to know more)

There are lots of libraries on this topic and this is not necessarily better, yet it is a natural implementations of the zoom engine. It is fast, lightweight and simple.

The zoom image view will forward all API calls to the internal engine. See engine docs. You can also get the backing engine using zoomImageView.getEngine().

zoomImageView.panTo(x, y, true); // Shorthand for zoomImageView.getEngine().panTo(x, y, true)
zoomImageView.panBy(deltaX, deltaY, true);
zoomImageView.zoomTo(zoom, true);
zoomImageView.zoomBy(factor, true);
zoomImageView.realZoomTo(realZoom, true);
zoomImageView.moveTo(zoom, x, y, true);

The low-level engine offers a Matrix-based stream of updates, as long as it is fed with touch events and knows the dimensions of your content.

There is no strict limit over what you can do with a Matrix,

• move Canvas objects around
• transform View hierarchies
• apply to ImageViews or Bitmap
• transform MotionEvents
• probably more

When the engine becomes aware of the content size, it will apply a base transformation to the content that can be controlled through setTransformation(int, int) or app:transformation and app:transformationGravity. It is applied only once, and defines the starting viewport over our content.

If, after applying the transformation (and any minZoom / maxZoom constraint), the content is partially cropped along some dimension, the engine will also apply a translation according to the given transformation gravity.

The base transformation makes the difference between zoom and realZoom. Since we have silently applied a base zoom to the content, we must introduce two separate types:

To make things clearer, when transformation is none, the zoom and the real zoom will be identical. The distinction is very useful when it comes to imposing min and max constraints to our zoom value.

Some of the zoom APIs will let you pass an integer (either TYPE_ZOOM or TYPE_REAL_ZOOM) to define the zoom type you are referencing to. Depending on the context, imposing restrictions on one type will make more sense than the other - e. g., in a PDF viewer, you might want to cap real zoom at 1.

The moveTo(float, float, float, boolean) API will let you animate both zoom and pan at the same time.

All pan APIs accept x and y coordinates. These refer to the top-left visible pixel of the content.

• If using ZoomLayout, the coordinate system is that of the inner view
• If using ZoomImageView, the coordinate system is that of the drawable intrinsic width and height
• If using the engine directly, the coordinate system is that of the rect you passed in setContentRect

In any case the current scale is not considered, so your system won't change if zoom changes.

The moveTo(float, float, float, boolean) API will let you animate both zoom and pan at the same time.

If you are interested in using the engine directly, I encourage you to take a look at the ZoomLayout or ZoomImageView implementations. It is extremely simple. Basically:

• You construct a ZoomEngine passing the View that acts as a container for your content
• As soon as you know it (and whenever it changes), you pass the content dimensions
• As soon as you receive them, you pass touch updates to onInterceptTouchEvent or onTouchEvent
• The ZoomEngine.Listener is fed with Matrix updates

You are welcome to contribute with suggestions or pull requests. I don't plan to add lots of features (specifically, I don't plan to have ZoomImageView compete with similar libraries that already do this very well), the plan is to keep this lightweight. But I welcome well-thought contributions.