SQL Server has a built in method for easily pulling one value from an XML. To do this you use the method called .value(). The .value method will run an XQuery against the XML specified in the query. This method is scalar, so it will only return 1 value. You cannot use this to return multiple values.
-- Load the XML data in to a variable to work with.
-- This would typically be passed as a parameter to a stored procedure
In the query above, you can see that you can just pass the xpath and the datatype that you want the output to be. This will pull that value out of the XML and put it in the specified format. You can see that after the xpath, there is a funny syntax . This tells SQL Server to grab the first instance of the AnimalName node. You need to do this because the .value() method only returns one value and will not work with a repeating node.
This part says to parse the @XMLToParse variable (previously filled with the XML file) using the method built in to SQL Server called .nodes. It specifies the repeating node to be Animals/LandAnimals/Animal. It assigns an aliases this XML parsed records as a table named xmlData. The (A) is the column name of the rowset. This will be referenced in the select part of the statement.
The select part of this statement references xmlData (which is the table aliases) and A (which is the column named for that table). It calls the .value function to return the value from the table/column. For the .value function, you pass in 2 elements.
The first element is the field. In this case we are passing in just a period. We do this because there is no node below the Animal node. If there was, we would need to specify it here..
The second parameter is the datatype that you would like the value to be casted as. Then I always aliases the field to something relevant.
Although parsing XML can be a little confusing in SQL Server, it is very powerful. This is a great way to pass bulk data to a stored procedure from any type of client application.