Repopulate missing data

Sample Data

Query

;WITH CTE AS (        SELECT                olnID               ,cpID               ,cpIDInstance               ,cpValue        FROM        (               VALUES                (1, 68,   1, N'New Bern')               ,(1, 5016, 1, N'MT')               ,(1, 5066, 1, N'OAK')               ,(2, 68,   1, N'New Bern')               ,(2, 5016, 1, N'MT')               ,(2, 5066, 1, N'CHY')               ,(2, 5066, 2, N'MPL')               ,(2, 5066, 3, N'OAK')        ) cpv (olnID,cpID,cpIDInstance,cpValue) ) SELECT DISTINCT        a.olnID ,      cpID = CASE WHEN a.cpID = b.cpID THEN a.cpID ELSE b.cpID END ,      cpIDInstance = CASE WHEN a.cpID = b.cpID THEN a.cpIDInstance ELSE b.cpIDInstance END ,      N = a.cpIDInstance ,      cpValue = CASE WHEN a.cpID = b.cpID THEN a.cpValue ELSE b.cpValue END FROM   CTE a CROSS JOIN CTE b WHERE  a.olnID = b.olnID               AND    a.cpID >= b.cpID ORDER BY 1, 4

Desired Output

Reference
http://social.msdn.microsoft.com/Forums/sqlserver/en-US/214fe2b6-b826-4057-8a09-5a760bf0df9a/help-needed-in-duplicating-data-subset#c46f989c-581d-4bc4-8214-6405bb87297f