| Ticket Hash: | 49e30d76126fb32ad22333e4b6363400fcff4b46 | |||
| Title: | Error with NULL columns | |||
| Status: | Closed | Type: | Code_Defect | |
| Severity: | Important | Priority: | Medium | |
| Subsystem: | Db_Type_Conversion | Resolution: | Duplicate | |
| Last Modified: | 2013-04-29 23:03:36 | |||
| Version Found In: | 1.0.85 | |||
| User Comments: | ||||
anonymous added on 2013-04-27 20:49:25:
Hello, A query like "SELECT *, NULL AS C1 FROM T1;" returns an error "DataException: Invalid storage type: DBNull" So a query like "SELECT T1.* FROM T1 LEFT JOIN (SELECT C1, Max(C2) AS C2_Max FROM T2 GROUP BY C1) T3 ON T1.C1 = T3.C1;" returns the same error if T2 has no record, because "(SELECT C1, Max(C2) AS C2_Max FROM T2 GROUP BY C1)" is null My code in VB.net: Dim oCommand As New SQLiteCommand(stSQL) Dim oDataAdapter As New SQLiteDataAdapter(oCommand) oCommand.Connection() = oConnection oDataAdapter.Fill(oDataset, stTableDTS) '=> Error oDataAdapter.Dispose() oCommand.Dispose() I used the package sqlite-netFx40-static-binary-bundle-Win32-2010-1.0.85.0.zip, It's ok with the previous version 1.0.84 mistachkin added on 2013-04-29 20:14:19: Would it be possible to see the full schema for the tables involved in this query? mistachkin added on 2013-04-29 20:41:42: It seems to me that the query "SELECT *, NULL AS C1 FROM T1;" should consider the "type" of column C1 to be DBNull.Value. Am I missing something here? mistachkin added on 2013-04-29 22:27:23: This issue appears to be related to ticket [48a6b8e4ca].
mistachkin added on 2013-04-29 23:03:36: Preliminary fix checked-in on trunk [dcc277f29a].
Is it possible for you to verify that this fixes the issue seen in your
environment?
Closing this ticket as 'duplicate'. Further comments on this issue should be
added to ticket [48a6b8e4ca] instead of this one.
| ||||
