Attributes representing information about the job and document(s) or resources required and/or consumed.

Each entry in the jmAttributeTable is a per-job entry with an extra index for each type of attribute (jmAttributeTypeIndex) that a job can have and an additional index (jmAttributeInstanceIndex) for those attributes that can have

multiple instances per job. The jmAttributeTypeIndex object SHALL contain an enum type that indicates the type of attribute (see the JmAttributeTypeTC textual-convention). The value of the attribute SHALL be represented in either the jmAttributeValueAsInteger or jmAttributeValueAsOctets objects, and/or both, as specified in the JmAttributeTypeTC textual- convention.

The agent SHALL create rows in the jmAttributeTable as the server or device is able to discover the attributes either from the job submission protocol itself or from the document PDL. As the documents are interpreted, the interpreter MAY discover additional attributes and so the agent adds additional rows to this table. As the attributes that represent resources are actually consumed, the usage counter contained in the jmAttributeValueAsInteger object is incremented according to the units indicated in the description of the JmAttributeTypeTC enum.

The agent SHALL maintain each row in the jmAttributeTable for at least the minimum time after a job completes as specified by the jmGeneralAttributePersistence object.

Zero or more entries SHALL exist in this table for each job in a job set.

See Section 3.3 entitled 'The Attribute Mechanism' for a description of the jmAttributeTable.