Conservar el orden de elementos al triturar datos XML en filas SQL

Question

¿Cómo devuelvo la secuencia de elementos al triturar XML en filas en una vista de SQL Server?

Ejemplo de entrada:

<ol> 
    <li>Smith</li> 
    <li>Jones</li> 
    <li>Brown</li> 
</ol> 

salida deseada:

Sequence Name 
-------- ----------- 
    1  Smith 
    2  Jones 
    3  Brown 

vista existente:

CREATE VIEW OrderedList 
AS 
SELECT [Sequence] = CAST(NULL AS int) -- TODO: Get ordinal position 
     [Name] = b.b.value('.', 'nvarchar(max)') 
FROM 
(
    SELECT a = CAST('<ol><li>Smith</li><li>Jones</li><li>Brown</li></ol>' AS xml) 
) a 
CROSS APPLY a.a.nodes('/ol/li') b (b) 

Answer 1

Puede utilizar row_number() en el nodo XML.

CREATE VIEW OrderedList 
AS 
SELECT [Sequence] = ROW_NUMBER() OVER(ORDER BY b.b), 
     [Name] = b.b.value('.', 'nvarchar(max)') 
FROM 
(
    SELECT a = CAST('<ol><li>Smith</li><li>Jones</li><li>Brown</li></ol>' AS xml) 
) a 
CROSS APPLY a.a.nodes('/ol/li') b (b) 

Ref: Uniquely Identifying XML Nodes with DENSE_RANK de Adam Machanic.