A
Rectangle specifies an area in a coordinate space that is enclosed by the
Rectangle object's upper-left point
(x,y) in the coordinate space, its width, and its height.
A Rectangle object's width and height are public fields. The constructors that create a Rectangle, and the methods that can modify one, do not prevent setting a negative value for width or height.
A Rectangle whose width or height is exactly zero has location along those axes with zero dimension, but is otherwise considered empty. The isEmpty() method will return true for such a Rectangle. Methods which test if an empty Rectangle contains or intersects a point or rectangle will always return false if either dimension is zero. Methods which combine such a Rectangle with a point or rectangle will include the location of the Rectangle on that axis in the result as if the add(Point) method were being called.
A Rectangle whose width or height is negative has neither location nor dimension along those axes with negative dimensions. Such a Rectangle is treated as non-existent along those axes. Such a Rectangle is also empty with respect to containment calculations and methods which test if it contains or intersects a point or rectangle will always return false. Methods which combine such a Rectangle with a point or rectangle will ignore the Rectangle entirely in generating the result. If two Rectangle objects are combined and each has a negative dimension, the result will have at least one negative dimension.
Methods which affect only the location of a Rectangle will operate on its location regardless of whether or not it has a negative or zero dimension along either axis.
Note that a Rectangle constructed with the default no-argument constructor will have dimensions of 0x0 and therefore be empty. That Rectangle will still have a location of (0,0) and will contribute that location to the union and add operations. Code attempting to accumulate the bounds of a set of points should therefore initially construct the Rectangle with a specifically negative width and height or it should use the first point in the set to construct the Rectangle. For example:
Rectangle bounds = new Rectangle(0, 0, -1, -1);
for (int i = 0; i < points.length; i++) {
bounds.add(points[i]);
}
or if we know that the points array contains at least one point:
Rectangle bounds = new Rectangle(points[0]);
for (int i = 1; i < points.length; i++) {
bounds.add(points[i]);
}
This class uses 32-bit integers to store its location and dimensions. Frequently operations may produce a result that exceeds the range of a 32-bit integer. The methods will calculate their results in a way that avoids any 32-bit overflow for intermediate results and then choose the best representation to store the final results back into the 32-bit fields which hold the location and dimensions. The location of the result will be stored into the x and y fields by clipping the true result to the nearest 32-bit value. The values stored into the width and height dimension fields will be chosen as the 32-bit values that encompass the largest part of the true result as possible. Generally this means that the dimension will be clipped independently to the range of 32-bit integers except that if the location had to be moved to store it into its pair of 32-bit fields then the dimensions will be adjusted relative to the "best representation" of the location. If the true result had a negative dimension and was therefore non-existent along one or both axes, the stored dimensions will be negative numbers in those axes. If the true result had a location that could be represented within the range of 32-bit integers, but zero dimension along one or both axes, then the stored dimensions will be zero in those axes.