views:

824

answers:

2

Hi!

I know there are several threads and posts regarding this issue in the internet and I've read them (not every article, I have to admit) but none of them did fully satisfy me.

My situation:
I'm using ODP.net (dll version 2.111.6.0) to access the Oracle DB (version 10 + 11) and a DataReader to retrieve the data (.NET 3.5, C#).

Using this code results in a 'System.OverflowException (Arithmetic operation resulted in an overflow.)'


decimal.TryParse(oraReader.GetOracleDecimal(0).Value.ToString(), 
  NumberStyles.Any, null, out parsedOraDecimal)

and this one results in a value of '3,000000000000000000000000000000000000000000000000000000000E-126'


decimal.TryParse(oraReader.GetOracleValue(0).ToString(), 
  NumberStyles.Any, null, out parsedOraDecimal)

Now I have to find some way to retrieve and evaluate this value properly - the DB is also used from other apps which are out of my control so changes there are not possible.

Converting the types in my C# code from 'decimal' to 'double' is also not really an option.

Any ideas?

Thanks for help and advice!

Regards, robert.oh.

A: 

OracleDecimal has a larger precision than decimal. For that reason, you have to reduce the precision first. Forget all the parsing, use implicit conversion. Try something along the lines of (untested):

decimal d = (decimal)(OracleDecimal.SetPrecision(oraReader.GetOracleDecimal(0), 28));
ammoQ
Thanks for help!
robert.oh.
+1  A: 

I just had a similar issue, and tried the approach of changing the OracleDataAdapter to return Oracle specific types ( data_adapter.ReturnProviderSpecificTypes = true; ), but this is just a PITA, you end up casting OracleStrings back to strings, etc.

In the end I solved it by doing the precision rounding in the SQL statement using Oracle's round function:

SELECT round( myfield, 18 ) FROM mytable

Dotnet will then happily convert the figure to a decimal.

Dave Glassborow
Hi Dave,thanks for the hint!
robert.oh.