Tudor Malene (Migrated from SEC-1423) said:
The ProtectPointcutPostProcessor.postProcessBeforeInitialization method parses each pointcut expression for each method of the present bean until it finds a match, which is highly inefficient.
The parsed pointcutExpressions can be cached in a LinkedHashSet when the pointcut is added ( see patch ), thus avoiding the parsing of the same expression for each method of the current bean.
Luke Taylor said:
Good idea. I've updated the code to add caching for both 3.1.x and 3.0.x.