Situation:
I am creating an app in C# that uses Oracle.DataAccess.Client (11g) to do certain operations on a Oracle database with stored procedures. I am aware that there is a certain enum (OracleDbType) that contains the Oracle data types, but I am not sure which one to use for certain types.
Questions:
What is the equivalent Oracle PL/SQL data type for each enumerated type in the OracleDbType enumeration?
There are three types of integer
(Int16, Int32, Int64) in the OracleDbType... how to know which one to use or are they all
suppose to work?
NUMBER(1,0) => Boolean
NUMBER(5,0) => Int16.MaxValue == 32767
NUMBER(10,0) => Int32.MaxValue == 2,147,483,647
NUMBER(19,0) => Int64.MaxValue == 9,223,372,036,854,775,807
NUMBER(19,0) => long.MaxValue == 9,223,372,036,854,775,807
The values of the OracleDbType Enumeration are defined in the documentation. Read the ODP for .NET Developer's Guide.
With regards to choosing between Int16, Int32 and Int64, they are all supposed to work. Choose the one which matches the expected size of your .Net variable: Int16 for values between -32768 and 32767, Int32 for values between -2147483648 and 2147483647, and Int64 for anything larger. There appear to be some funnies relating to converting Ints and PL/SQL data types. Check this blog post by Mark Williams.
Here's a method to convert C# types to the most common OracleDbTypes
Also, for very large character data values, you may want to use
OracleDbType.Clob
.Check APC's links out, they are what you are looking for : the mapping is quite straightforward according to the name of the enumeration.
But as you began to notice, there is something tricky about integers. Here is my mapping :
Int16
:NUMBER(5)
.Int32
:NUMBER(10)
.Int64
:NUMBER(19)
.The thing is that if you call
GetInt64
on aNUMBER(38)
column, you will get an exception even if the value is in the correct range...