Usage

XPath querying

One can query an XML document using XPath 1.0 queries using omni_xml.xpath function. The function takes an XML document as text and an XPath query as text.

It returns a table with path (path to the node) and value (node value).

Element

When searching for elements, it will return entire nodes serialized as XML.

select * from omni_xml.xpath('<node>value</node>', '/node')

Results in:

 path  |       value        
-------+--------------------
 /node | <node>value</node>
(1 row)

Text

When searching for textual data, one can use text() in their queries.

select * from omni_xml.xpath('<node>value</node>', '/node/text()')

Results in:

  path  | value 
--------+-------
 /node/ | value
(1 row)

Namespaces

It is possible to query using an explicit namespace identifier.

select * from omni_xml.xpath('<ns:node>value</ns:node>', '/ns:node')

Results in:

   path   |          value           
----------+--------------------------
 /ns:node | <ns:node>value</ns:node>

Proper namespace support is lacking

This is not ideal if the identifier used is unknown ahead of time, but sufficient if used for API response handling when it is typically the same.