从普及数据库读取日期字段会导致崩溃

Question

I am trying to read some data from a Pervasive database (through a C# Console app I wrote) and whenever I reach a row that contains information on a particular date column, the program throws an exception.

I managed to identify the problematic records:

DtTransDate        DtSystemTime           DtnLotteryDrawDate
7/15/2013          3:01:32 AM             9/1/8226

As you can see, the DtnLotteryDrawDate is a weird value but I should be able to read it as it is a valid date. I am using the ADO.NET Provider for Pervasive (which by the way, indicates that there's a bug on the Pervasive driver itself):

at Pervasive.Data.SqlClient.Lna.e.ak(u )
at Pervasive.Data.SqlClient.Lna.w.a(Int16 , Int32 , ad , f )
at Pervasive.Data.SqlClient.Lna.e.aa(Int32 , Boolean )
at Pervasive.Data.SqlClient.PsqlDataReader.e()
at Pervasive.Data.SqlClient.PsqlDataReader.Read()

So I tried the ODBC Driver and got the same problem but a tiny better message:

at System.DateTime.DateToTicks(Int32 year, Int32 month, Int32 day)
at System.DateTime..ctor(Int32 year, Int32 month, Int32 day, Int32 hour, Int32 minute, Int32 second, Int32 millisecond)
   at Pervasive.Data.SqlClient.Lna.u.r()
   at Pervasive.Data.SqlClient.Lna.aq.a(u )
   at Pervasive.Data.SqlClient.Lna.e.ak(u )
   at Pervasive.Data.SqlClient.Lna.w.a(Int16 , Int32 , ad , f )
   at Pervasive.Data.SqlClient.Lna.e.aa(Int32 , Boolean )
   at Pervasive.Data.SqlClient.PsqlDataReader.e()
   at Pervasive.Data.SqlClient.PsqlDataReader.Read()

So I tried converting the Date to a string to see if that would get rid of the issue but that didn't work. Basically, I wrote the following SQL statement:

 SELECT DtTransDate ,
        TmSystemTime ,
        CONVERT(DtnLotteryDrawDate, SQL_CHAR) as DtnLotteryDrawDate
        FROM    TICKHISH
        WHERE   ( DtTransDate >= { d '2013-07-15' }

The above SQL Statement would throw the following exception:

error in row and the stack trace I pasted above.

If I DO NOT try to convert the field to string, then the Exception is more helpful:

Year, Month, and Day parameters describe an un-representable DateTime.

So essentially, the ADO.NET driver is trying to construct a Date passing in incorrect values for either the month or the day. Probably is passing the 8226 as either the day or the month field.

I tried running the same SQL statement on the Pervasive Control Panel (PCC) and interestingly, the PCC program doesn't crash if I do not attempt to convert to string the date field but if I try any kind of conversion or calling a function like year(DtnLotteryDrawDate) immediately crashes as well.

Any ideas on how to work around this issue?

TINY UPDATE

I was able to set up a Linked Server using SSMS and connect to the Pervasive database through the linked server. When I attempt to run this simple query:

SELECT  * FROM OPENQUERY(linked_server, 'SELECT * FROM MyTable where DtTransDate  = ''2013-07-15'' ')

I get this error:

OLE DB provider "MSDASQL" for linked server "linked_server" returned message "[Pervasive][ODBC Client Interface]Data truncated column 186.".
OLE DB provider "MSDASQL" for linked server "linked_server" returned message "[Pervasive][ODBC Client Interface]Invalid date, time or timestamp value.".
Msg 7330, Level 16, State 2, Line 3
Cannot fetch a row from OLE DB provider "MSDASQL" for linked server "linked_server".

Answer 1

There seems to be something wrong with the actual data. I was able to insert, through the PCC, and view through both PCC and the ADO.NET provider a data value of 9/1/8226. If I changed the bytes of the date value through Function Executor to make the date appear invalid, I was able to get an "error in row" type message. There used to be a couple of programs that could fix invalid dates but I think one was pulled. The other is FixDates from Goldstar Software, but it is not free.