Question

What are the advantages of replacing (int)reader[0] with reader.GetInt32(0)? I’m sure such casting functions are there for a reason, but other than it feeling more aesthetically pleasing to avoid the cast myself, I’m not sure what those reasons are.

Answer 1

In code….

        void OneWay()         {             System.Data.SqlClient.SqlDataReader reader = null;             int i = reader.GetInt32(0);         }          void OtherWay()         {             System.Data.SqlClient.SqlDataReader reader = null;             int i = (int)reader[0];         } 

In IL

.method private hidebysig instance void OneWay() cil managed {     .maxstack 2     .locals init (         [0] class [System.Data]System.Data.SqlClient.SqlDataReader reader,         [1] int32 i)     L_0000: nop      L_0001: ldnull      L_0002: stloc.0      L_0003: ldloc.0      L_0004: ldc.i4.0      L_0005: callvirt instance int32 [System.Data]System.Data.Common.DbDataReader::GetInt32(int32)     L_000a: stloc.1      L_000b: ret  }   .method private hidebysig instance void OtherWay() cil managed {     .maxstack 2     .locals init (         [0] class [System.Data]System.Data.SqlClient.SqlDataReader reader,         [1] int32 i)     L_0000: nop      L_0001: ldnull      L_0002: stloc.0      L_0003: ldloc.0      L_0004: ldc.i4.0      L_0005: callvirt instance object [System.Data]System.Data.Common.DbDataReader::get_Item(int32)     L_000a: unbox.any int32     L_000f: stloc.1      L_0010: ret  } 

So, the IL is different, but I doubt that there is any noticeable difference between them. Maybe after a million iterations you will see the difference, but not likely.