A little late to this party - but this also works well:
CREATE VIEW [ABC].[View_SomeDataUniqueKey]
AS
SELECT
CAST(CONCAT(CAST([ID] AS VARCHAR(4)),
CAST(ROW_NUMBER() OVER(ORDER BY [ID] ASC) as VARCHAR(4))
) AS int) AS [UniqueId]
,[ID]
FROM SOME_TABLE JOIN SOME_OTHER_TABLE
GO
In my case the join resulted in [ID] - the primary key being repeated up to 5 times (associated different unique data) The nice trick with this is that the original ID can be determined from each UniqueID effectively [ID]+RowNumber() = 11, 12, 13, 14, 21, 22, 23, 24 etc. If you add RowNumber() and [ID] back into the view - you can easily determine your original key from the data. But - this is not something that should be committed to a table because I am fairly sure that the RowNumber() of a view will never be reliably the same as the underlying data alters, even with the OVER(ORDER BY [ID] ASC) to try and help it.
Example output ( Select UniqueId, ID, ROWNR, Name from [REF].[View_Systems] ) :
UniqueId ID ROWNR Name
11 1 1 Amazon A
12 1 2 Amazon B
13 1 3 Amazon C
14 1 4 Amazon D
15 1 5 Amazon E
Table1:
[ID] [Name]
1 Amazon
Table2:
[ID] [Version]
1 A
1 B
1 C
1 D
1 E
CREATE VIEW [REF].[View_Systems]
AS
SELECT
CAST(CONCAT(CAST(TABA.[ID] AS VARCHAR(4)),
CAST(ROW_NUMBER() OVER(ORDER BY TABA.[ID] ASC) as VARCHAR(4))
) AS int) AS [UniqueId]
,TABA.[ID]
,ROW_NUMBER() OVER(ORDER BY TABA.[ID] ASC) AS ROWNR
,TABA.[Name]
FROM [Ref].[Table1] TABA LEFT JOIN [Ref].[Table2] TABB ON TABA.[ID] = TABB.[ID]
GO