ArcSDESQLExecute is not able to handle unique identifiers

[Cristian_Galindo]

Using ArcGIS Pro 2.9.5 I found that ArcSDESQLExecute is not able to handle unique identifiers in a table.

I tried to run the following Script:

database = 'C:\\Installation\\updm.sde'

egdb_conn = arcpy.ArcSDESQLExecute(database)
sql_clause = "SELECT PROJECT_ID from DataOwner.My_PROJECTS;"
print(sql_query)

query_return = egdb_conn.execute(sql_clause)

len(query_return)

the table My_PROJECTS has a column PROJECT_ID defined as a uniqueidentifier., the result odf the execution is:

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
In  [39]:
Line 12:    query_return = egdb_conn.execute(sql_clause)

File C:\ArcGISPro\Resources\ArcPy\arcpy\arcobjects\arcobjects.py, in execute:
Line 44:    return convertArcObjectToPythonObject(self._arc_object.Execute(*gp_fixargs(args)))

AttributeError: ArcSDESQLExecute: StreamBindOutputColumn ArcSDE Error -65 Invalid pointer argument to function

[BlakeTerhune]

Try without the ending semicolon in the SQL statement.

[Cristian_Galindo]

Thanks for the suggestion.

As a matter of fact I did it before of making the post, but I did it again just to be able to copy and paste the output from the execution:

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
In  [48]:
Line 7:     query_return = egdb_conn.execute(sql_clause)

File C:\ArcGISPro\Resources\ArcPy\arcpy\arcobjects\arcobjects.py, in execute:
Line 44:    return convertArcObjectToPythonObject(self._arc_object.Execute(*gp_fixargs(args)))

AttributeError: ArcSDESQLExecute: StreamBindOutputColumn ArcSDE Error -65 Invalid pointer argument to function
---------------------------------------------------------------------------

As I already state in my original post this happens with columns defined as unique identifier. If I try to make the query to an INT, VARCHAR or DECIMAL column, the query is performed with out any issues.

[Cristian_Galindo · Accepted solution]

Despite the fact that unique identifiers are supported, I went for a casting approach

database = 'C:\\Installation\\updm.sde'

egdb_conn = arcpy.ArcSDESQLExecute(database)
# Getting the ID's froma table
sql_clause = "SELECT CAST(PROJECT_ID AS VARCHAR(100)) from DataOwner.My_PROJECTS;"

query_return = egdb_conn.execute(sql_clause)

# Whatever operation you want to do with the results, in this case
# is just the verification of how many results, but they will be 
# iterated later

len(query_return)

* by the way the semicolon does not make any difference.