Java Enterprise in a Nutshell, Second Edition

Identifiers

An identifier refers to either a standard JMS header field name or a custom message property name. Any JMS header field name can be used as an identifier, except for JMSDestination, JMSExpiration, JMSRedelivered, and JMSReplyTo, which can be used as identifiers in a message selector. JMSDestination and JMSReplyTo are Destination values, and message selector operators support only numeric, boolean or string values. The time at which message selectors are applied to messages isn’t specified in the JMS specification, so using the value of JMSExpiration doesn’t provide a consistent, well-defined result. Using JMSRedelivered in a selector can result in unexpected results. If, for example, a selector checks for JMSRedelivered being true, the first delivery attempt by a provider will fail the selector because the redelivered flag should be false, but the provider can then immediately redeliver the message and pass the selector, making the redelivered part of the predicate ineffective.

Identifier names are case-sensitive and follow the same general rules as Java identifiers. They must start with a valid Java identifier start character as determined by the java.lang.Character.isJavaIdentifierStart( ) method. For example, a letter, currency symbol, or connecting punctuation character such as an underscore _ contain valid Java identifier characters as determined by the Character.isJavaIdentifierPart( ) method. You can’t use these reserved words for identifiers: NULL, TRUE ...