XPath heaven and Namespace hell

XPath is fairly easy to learn by example (great examples here) but namespaces in the XML can be maddening when they cause data to be alluded.  Some things to check before you resort to using simple string matching or regular expressions to get to your XML data:

• XPath expressions are case-sensitive
• Indexing is 1-based rather than 0-based, e.g. /MSG.2[1] refers to the first element (not the second element)
• Default namespaces (e.g. xmlns="urn:hl7-org:v3" vs. xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance", the former having a default namespace) requires that you set up a prefix for the namespace if you want to write XPath expressions against the XML content.  Good article about default namespaces 
• When writing XPath expressions where namespaces are involved, you need to prefix each element in the expression (with the appropriate namespace prefix).  E.g. /hl7v3:PRPM_IN301030CA/hl7v3:controlActEvent/hl7v3:effectiveTime is required, /hl7v3:PRPM_IN301030CA/controlActEvent/effectiveTime will not work (there is no inheritance down the branch)