Community
Participate
Working Groups
Build 20030129 The current implementation of the custom hashtable doesn't adjust the key if an entry with the same hash value already exists. IMO the following statement is missing: if (entry == null) { ... } Object result = entry.value; entry.key = key; <=== missing statement entry.value = value; Otherwise the hash table might hang on "old" keys which could be garbage collect. Additionally it is confusing to see "wrong" keys in the hash table. In JDT the hash table kept a pointer to a working copy element although it could be garbage collected since we exchanged the element with the original.
Good catch. Fixed.