Working with Borders

Imagine a label whose text runs right up against the edge of its box, or a panel whose inner controls sit flush against its outer boundary. The layout works, but it looks cramped. Swing lets you fix this by attaching a border to a component. A border is a decorative or spacing region that surrounds the component's edges. Depending on the kind of border you choose, that region might be invisible padding, a single colored outline, or a frame with different thicknesses on each side.

As you read, keep one question in mind: what does this border add to the component's area, and what does the learner see in that added area? The three border types covered here answer that question in three different ways. Following each one step by step makes their differences much easier to hold onto than memorizing their method names.

Start with a bare component that has no border at all. Its visible region is just its own background, and its content reaches the outermost edge. Now attach a border, and ask two things. First, does the border occupy any width along each edge? Second, if it occupies space, is anything drawn into that space or is it left alone? Those two questions separate the three types. One type reserves space and draws nothing. Another draws a visible line of equal width on every side. The third draws a line too, but lets each side have its own width.

The before and after columns show the same label. In the second column, the border zone is carved out of the label's original bounds. Whatever the label was drawing now has to fit inside the smaller inner area, so the visible content is pushed inward and the content area shrinks by the width the border reserves on each side.

Every bordered component can be read as three nested regions. The outer bounds are the full rectangle the component occupies. Inside that sits the border zone, the strip reserved along the edges. The innermost region is the content area, where the component's own text, controls, or painting actually lands. The border zone is always between the other two.

The key feature is that nothing is drawn in the reserved space. A reader who expects a visible outline will be surprised: the empty border is purely a spacing tool. Its usefulness comes from what it does to the layout around the component rather than from anything visible on its own.

The source names two common uses. On a JPanel that holds other components, an empty border puts space between those components and the panel's edge. On a JLabel, it adds breathing room between the text and the edge of the label, which otherwise would have none.

A line border draws a line of one color and one thickness in pixels around all four edges of a component. Unlike the empty border, it does not simply reserve space; it fills that space with a visible line. Every side receives the same thickness, because a single integer governs all four edges. If you want a frame that looks identical on all sides, this is the natural choice.

Think of the matte border as the line border generalized. Where the line border uses one thickness for all four sides, the matte border accepts four separate widths, one for each edge. This lets you build frames that are heavier on one side than another, or open up a side entirely by giving it a width of zero. The four widths are followed by one color, which applies to the drawn border.

Placing the three methods side by side shows how their parameter lists encode their behavior. The empty border takes four widths and no color, because it draws nothing. The line border takes a color and one thickness, because it draws a uniform line. The matte border takes four widths and a color, because it draws a line whose thickness can differ by edge. Once you know that, the parameter list tells you whether a call draws anything at all.

Choosing the right type comes down to one question: does the area need to be blank, drawn with one uniform line, or drawn with a line whose sides differ? The comparison below lays out how each type answers it.