Please explain the CDA entryRelationship element

I have access to the HL7 Clinical Document Architecture, Release 2.0, which states that it is used essentially to link entries with each other in a CDA document. Specifically, it links between what is called the "source" and the "target" entries. I also read about the different types of relationships (CAUS, COMP, GEVL, MFST, REFR, RSON, SAS, SPRT, SUBJ, XCRPT) and somewhat understand those.

My main question: what are the "source" and "target" elements? Are they the element containing the entryRelationship, and the element contained by entryRelationship?

For example:

<entry typeCode="DRIV">
    <act classCode="ACT" moodCode="EVN">
        ...
        <entryRelationship ty开发者_如何学编程peCode="SUBJ">
            <observation classCode="OBS" moodCode="EVN">
                ...
                <entryRelationship typeCode="REFR">
                    <observation classCode="OBS" moodCode="EVN">
                    ...
                    </observation>
                </entryRelationship>
            </observation>
        </entryRelationship>
    </act>  
</entry>

In the above snippet, according to my understanding, there is a SUBJ relationship between the act the the first observation, and there is a REFR relationship between the two observations. Is this correct?

The source of a entryRelationship is the entry that contains in its body the element in your example the source entry is

 <act classCode="ACT" moodCode="EVN"> 

and the target is

 <observation classCode="OBS" moodCode="EVN">  

Its is posible to indicate a inverse relationship using the "InversionInd" attribute of the entryRelationship elemet. If this attribute is set to true source and target are inverted.