Marcono1234 created LOG4J2-2948:
-----------------------------------
Summary: Use IdentityHashMap for ParameterFormatter deja vu
Key: LOG4J2-2948
URL: https://issues.apache.org/jira/browse/LOG4J2-2948
Project: Log4j 2
Issue Type: Improvement
Reporter: Marcono1234
{{org.apache.logging.log4j.message.ParameterFormatter}} currently uses its
{{identityToString(Object)}} method to create identifier strings for detecting
recursive objects.
However, as noted by the description of that method, this approach is not
completely foolproof:
{code}
* As much as is reasonably practical, the hashCode method defined by
* class {@code Object} does return distinct integers for distinct
* objects. (This is typically implemented by converting the internal
* address of the object into an integer, but this implementation
* technique is not required by the Java™ programming language.)
{code}
Would it make sense to switch to JDK's
[{{java.util.IdentityHashMap}}|https://docs.oracle.com/javase/8/docs/api/java/util/IdentityHashMap.html]
(converted to {{Set}} using {{Collections.newSetFromMap}})? I would assume,
but have not tested it, that performance will also be better because it simply
tests for equality by checking for reference equality whereas the current
approach needs to compare strings.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
