For those who works like me with both languages, sometimes we can encounter some subtle differences between VB and C#. Take this code in VB
Module Module1 Private Function NormalizeRank(ByVal userRank As Integer) As Integer Return CType(Math.Ceiling(Math.Min(userRank / 20, 5.0)), Integer) End Function Sub Main() For value As Integer = 0 To 101 Console.WriteLine("{0} normalized is {1}", value, NormalizeRank(value)) Next End Sub End Module
Is a very stupid piece of code that is supposed to take a value in the range 0..100 and normalize in the new interval [0..5], if the user pass a value higher than 100 the result must be 5. We expect value from 1 to 20 to be transformed into 1, values from 21 to 40 to be transformed to 2 and so on. If you run this code you find that the result is expected one. Now suppose you need to translate this function to C#, probably the first thing you can write is this one.
private static Int32 Normalize(Int32 rank) { return (Int32)Math.Ceiling(Math.Min(rank / 20, 5.0)); } static void Main(string[] args) { for (int value = 0; value < 101; value++) { Console.WriteLine("{0} normalized is {1}", value, Normalize(value)); } }
If you think that this is correct, try to run the code, and find this result.
0 normalized is 0 1 normalized is 0 2 normalized is 0 3 normalized is 0 4 normalized is 0 5 normalized is 0 6 normalized is 0 7 normalized is 0 8 normalized is 0 9 normalized is 0 10 normalized is 0 11 normalized is 0 12 normalized is 0 13 normalized is 0 14 normalized is 0 15 normalized is 0 16 normalized is 0 17 normalized is 0 18 normalized is 0 19 normalized is 0 20 normalized is 1 21 normalized is 1 22 normalized is 1 23 normalized is 1 24 normalized is 1
If you wonder why the normalization of a number like 2 is zero you need to think to the meaning of the operator / in both languages. In Visual Basic it means “Floating point division” while in C# the type of the operation is determined by the operands. so 2 / 20 in vb is 0.1 while 2 / 20 in C# is 0. To solve this problem you need to write something like this
private static Int32 Normalize(Int32 rank) { return (Int32)Math.Ceiling(Math.Min(rank / 20.0, 5.0)); }
Now since 20.0 is a double, the compiler will emit a floating point division.
Alk.
Tags: C#
Leave a Reply