Recovering design patterns can enhance existing source code analysis tools by bringing program understanding tothe design level. This paper presents a new, fully automatedpattern detection approach. The new approach is based onour reclassification of the GoF patterns by their pattern in-tent. We argue that the GoF pattern catalog classifies de-sign patterns in the forward-engineering sense; our reclas-sification is better suited for reverse engineering. Our ap-proach uses lightweight static program analysis techniquesto capture program intent. This paper also describes ourtool, PINOT, that implements this new approach. PINOTdetects all the GoF patterns that have concrete definitionsdriven by code structure or system behavior. Our tool isfaster, more accurate, and targets more patterns than exist-ing pattern detection tools. PINOT has been used success-fully in detecting patterns in Java AWT, JHotDraw, Swing,Apache Ant, and many other programs and packages.