I was looking at a patch for FormatDateSupport [1] to remove its dependency on a copy of LRUMap taken from Commons and have a question on the thread safety involved as, although the cache is synchronized, the formatters it contains are used outside that block.
This was discussed in bug 32311 [2] but the synchronization code suggested there wasn't added. Am I missing something, or should I reopen 32311 and add a patch? -- Jeremy [1] http://svn.apache.org/repos/asf/tomcat/taglibs/standard/trunk/impl/src/main/java/org/apache/taglibs/standard/tag/common/fmt/FormatDateSupport.java [2] https://issues.apache.org/bugzilla/show_bug.cgi?id=32311#c4