.Net 字符集与编解码
0 .NET 字符集编解码
.Net 内部使用的字符集是Unicode，如果需要编码为其他诸如GBK、UTF8编码，可以通过Encoding 类来实现。
using System.Text;
void PrintBytes(byte[] bytes)
{
foreach (var b in bytes)
{
Console.Write("{0:X} ", b);
}
Console.WriteLine();
}
Encoding.RegisterProvider(CodePagesEncodingProvider.Instance);
string str = "主账号";
var gbkBytes = Encoding.GetEncoding("gbk").GetBytes(str); //获取GBK编码
var utf8Bytes = Encoding.UTF8.GetBytes(str); //获取UTF8编码
var unicodeBytes = Encoding.Unicode.GetBytes(str); //获取Unicode编码
PrintBytes(gbkBytes);
PrintBytes(utf8Bytes);
PrintBytes(unicodeBytes);
var gbkStr = Encoding.GetEncoding("gbk").GetString(gbkBytes); //使用GBK解码
var utf8Str = Encoding.UTF8.GetString(utf8Bytes); //使用UTF8解码
var unicodeStr = Encoding.Unicode.GetString(unicodeBytes); //使用Unicode解码
Console.WriteLine(gbkStr);
Console.WriteLine(utf8Str);
Console.WriteLine(unicodeStr);
输出：
D6 F7 D5 CB BA C5
E4 B8 BB E8 B4 A6 E5 8F B7
3B 4E 26 8D F7 53
主账号
主账号
主账号在使用C++API时，当遇到字符串处理时难免会需要处理字符编码的问题。这里主要针对于使用C++ API是遇到的一些编码被封送的情况测试。
1 Windows环境下
这里首先测试了.Net 在Windows环境下运行情况下，.Net 默认使用ANSI 编解码，其中在 DllImport 中指定的 CharSet 对导出函数的直接字符串参数生效。CharSet 取值与C++ API 端接收到的字符串情况对应如下：
1.1 请求函数：C# => C++
1.1.1 测试函数字符串参数编码
输入文字：主账号
| CharSet | C++API端接收到的字符集 | 输出Byte值 |
|---|---|---|
| 不设置 | GBK | D6 F7 D5 CB BA C5 |
| Ansi | GBK | D6 F7 D5 CB BA C5 |
| Unicode | Unicode | 3B 4E 26 8D F7 53 |
| Auto | Unicode | 3B 4E 26 8D F7 53 |
| None | GBK | D6 F7 D5 CB BA C5 |
测试接口代码：
//对这个接口分别设置为以下5种情况进行测试
[DllImport(LibName, CallingConvention = CallingConvention.StdCall)]
public static extern void RegisterFront(string ip, int port);
[DllImport(LibName, CharSet=CharSet.Ansi, CallingConvention = CallingConvention.StdCall)]
public static extern void RegisterFront(string ip, int port);
[DllImport(LibName, CharSet=CharSet.Unicode, CallingConvention = CallingConvention.StdCall)]
public static extern void RegisterFront(string ip, int port);
[DllImport(LibName, CharSet=CharSet.Auto, allingConvention = CallingConvention.StdCall)]
public static extern void RegisterFront(string ip, int port);
[DllImport(LibName, CharSet=CharSet.None, CallingConvention = CallingConvention.StdCall)]
public static extern void RegisterFront(string ip, int port);1.1.2 测试函数结构体参数中的字符串编码
在这个 DllImport 中设置的CharSet 仅对接口函数的直接字符串类型生效。如果参数是一个对象，而对象中的字符串类型需要在定义封装对象的位置，通过StructLayout 属性的CharSet 来设置。我这里测试下来，CharSet 取值与C++ API 端接收到的字符串情况对应如下：
输入文字：主账号
| CharSet | C++API端接收到的字符集 | 输出Byte值 |
|---|---|---|
| 不设置 | GBK | D6 F7 D5 CB BA C5 |
| Ansi | GBK | D6 F7 D5 CB BA C5 |
| Unicode | null | |
| Auto | null | |
| None | GBK | D6 F7 D5 CB BA C5 |
跟上面类似，但是Unicode 传送的不成功，想必是类型问题，Unicode 对应C++中应该对应使用wchar* 数组。
测试接口及定义结构体的代码：
[StructLayout(LayoutKind.Sequential, CharSet=CharSet.Ansi)]
public class StepReqAddPrimaryAccount
{
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 9)]
public string? TradingDay;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 32)]
public string? PrimaryAccountID;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? PrimaryAccountName;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? BrokerPassword;
public int ChannelID;
public bool IsAllowLogin;
public AccountStatusType AccountStatus;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? Password;
public int RiskGroupID;
public int CommissionGroupID;
}
[StructLayout(LayoutKind.Sequential, CharSet = CharSet.Unicode)]
public class StepReqUpdatePrimaryAccount
{
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 9)]
public string? TradingDay;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 32)]
public string? PrimaryAccountID;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? PrimaryAccountName;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? BrokerPassword;
public int ChannelID;
public bool IsAllowLogin;
public AccountStatusType AccountStatus;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? Password;
public int RiskGroupID;
public int CommissionGroupID;
}
[StructLayout(LayoutKind.Sequential, CharSet = CharSet.None)]
public class StepReqAddAccount
{
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 9)]
public string? TradingDay;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 32)]
public string? AccountID;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? AccountName;
public AccountStatusType AccountStatus;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? Password;
public int TradeGroupID;
public int RiskGroupID;
public int CommissionGroupID;
}
[StructLayout(LayoutKind.Sequential, CharSet = CharSet.Auto)]
public class StepReqUpdateAccount
{
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 9)]
public string? TradingDay;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 32)]
public string? AccountID;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? AccountName;
public AccountStatusType AccountStatus;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? Password;
public int TradeGroupID;
public int RiskGroupID;
public int CommissionGroupID;
} [DllImport(LibName, CharSet=CharSet.Ansi, CallingConvention = CallingConvention.StdCall)]
public static extern int ReqAddPrimaryAccount(StepReqAddPrimaryAccount reqAddPrimaryAccount, int requestID);
[DllImport(LibName, CharSet = CharSet.Unicode, CallingConvention = CallingConvention.StdCall)]
public static extern int ReqUpdatePrimaryAccount(StepReqUpdatePrimaryAccount reqUpdatePrimaryAccount, int requestID);
[DllImport(LibName, CharSet = CharSet.Auto, CallingConvention = CallingConvention.StdCall)]
public static extern int ReqAddAccount(StepReqAddAccount reqAddAccount, int requestID);
[DllImport(LibName, CharSet = CharSet.None, CallingConvention = CallingConvention.StdCall)]
public static extern int ReqUpdateAccount(StepReqUpdateAccount reqUpdateAccount, int requestID);1.2 回调函数：C++ => C#
返回值：正确
C++中编码：GBK
| CharSet | C#回调函数接收到字符集 | 解码情况 |
|---|---|---|
| 不设置 | GBK | 正确 |
| Ansi | GBK | 正确 |
| Unicode | null | 乱码 |
| Auto | null | 乱码 |
| None | GBK | 正确 |
C++中编码：Utf8
相关测试代码：
//依次将CharSet 设置为：不设置值、Ansi、Unicode、Auto、None，进行测试
[StructLayout(LayoutKind.Sequential, CharSet=CharSet.Ansi)]
public class StepRspInfo
{
public int ErrorID;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 256)]
public string? ErrorMsg;
}回调委托定义：
[UnmanagedFunctionPointer(CallingConvention.StdCall)]
public delegate void OnRspAdminUserLogin(StepRspAdminUserLogin? rspAdminUserLogin, StepRspInfo? rspInfo, int requestID, bool isLast);2 Linux环境下
2.1 请求函数： C# => C++
2.1.1 测试函数字符串参数编码
输入文字：主账号
| CharSet | C++API端接收到的字符集 | 输出Byte值 |
|---|---|---|
| 不设置 | UTF8 | E4 B8 BB E8 B4 A6 E5 8F B7 |
| Ansi | UTF8 | E4 B8 BB E8 B4 A6 E5 8F B7 |
| Unicode | Unicode | 3B 4E 26 8D F7 53 |
| Auto | UTF8 | E4 B8 BB E8 B4 A6 E5 8F B7 |
| None | UTF8 | E4 B8 BB E8 B4 A6 E5 8F B7 |
测试接口代码：
//对这个接口分别设置为以下5种情况进行测试
[DllImport(LibName, CallingConvention = CallingConvention.StdCall)]
public static extern void RegisterFront(string ip, int port);
[DllImport(LibName, CharSet=CharSet.Ansi, CallingConvention = CallingConvention.StdCall)]
public static extern void RegisterFront(string ip, int port);
[DllImport(LibName, CharSet=CharSet.Unicode, CallingConvention = CallingConvention.StdCall)]
public static extern void RegisterFront(string ip, int port);
[DllImport(LibName, CharSet=CharSet.Auto, allingConvention = CallingConvention.StdCall)]
public static extern void RegisterFront(string ip, int port);
[DllImport(LibName, CharSet=CharSet.None, CallingConvention = CallingConvention.StdCall)]
public static extern void RegisterFront(string ip, int port);2.1.2 测试函数结构体参数中的字符串编码
通过StructLayout 属性的CharSet 来设置结构体中的字符串编码，CharSet 取值与C++ API 端接收到的字符串情况对应如下：
输入文字：主账号
| CharSet | C++API端接收到的字符集 | 输出Byte值 |
|---|---|---|
| 不设置 | UTF8 | E4 B8 BB E8 B4 A6 E5 8F B7 |
| Ansi | UTF8 | E4 B8 BB E8 B4 A6 E5 8F B7 |
| Unicode | null | |
| Auto | UTF8 | E4 B8 BB E8 B4 A6 E5 8F B7 |
| None | UTF8 | E4 B8 BB E8 B4 A6 E5 8F B7 |
跟上面类似，但是Unicode 传送的不成功，想必是类型问题，Unicode 对应C++中应该对应使用wchar* 数组。
测试接口及定义结构体的代码：
[StructLayout(LayoutKind.Sequential, CharSet=CharSet.Ansi)]
public class StepReqAddPrimaryAccount
{
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 9)]
public string? TradingDay;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 32)]
public string? PrimaryAccountID;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? PrimaryAccountName;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? BrokerPassword;
public int ChannelID;
public bool IsAllowLogin;
public AccountStatusType AccountStatus;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? Password;
public int RiskGroupID;
public int CommissionGroupID;
}
[StructLayout(LayoutKind.Sequential, CharSet = CharSet.Unicode)]
public class StepReqUpdatePrimaryAccount
{
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 9)]
public string? TradingDay;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 32)]
public string? PrimaryAccountID;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? PrimaryAccountName;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? BrokerPassword;
public int ChannelID;
public bool IsAllowLogin;
public AccountStatusType AccountStatus;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? Password;
public int RiskGroupID;
public int CommissionGroupID;
}
[StructLayout(LayoutKind.Sequential, CharSet = CharSet.None)]
public class StepReqAddAccount
{
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 9)]
public string? TradingDay;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 32)]
public string? AccountID;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? AccountName;
public AccountStatusType AccountStatus;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? Password;
public int TradeGroupID;
public int RiskGroupID;
public int CommissionGroupID;
}
[StructLayout(LayoutKind.Sequential, CharSet = CharSet.Auto)]
public class StepReqUpdateAccount
{
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 9)]
public string? TradingDay;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 32)]
public string? AccountID;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? AccountName;
public AccountStatusType AccountStatus;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 64)]
public string? Password;
public int TradeGroupID;
public int RiskGroupID;
public int CommissionGroupID;
} [DllImport(LibName, CharSet=CharSet.Ansi, CallingConvention = CallingConvention.StdCall)]
public static extern int ReqAddPrimaryAccount(StepReqAddPrimaryAccount reqAddPrimaryAccount, int requestID);
[DllImport(LibName, CharSet = CharSet.Unicode, CallingConvention = CallingConvention.StdCall)]
public static extern int ReqUpdatePrimaryAccount(StepReqUpdatePrimaryAccount reqUpdatePrimaryAccount, int requestID);
[DllImport(LibName, CharSet = CharSet.Auto, CallingConvention = CallingConvention.StdCall)]
public static extern int ReqAddAccount(StepReqAddAccount reqAddAccount, int requestID);
[DllImport(LibName, CharSet = CharSet.None, CallingConvention = CallingConvention.StdCall)]
public static extern int ReqUpdateAccount(StepReqUpdateAccount reqUpdateAccount, int requestID);2.2 回调函数： C++ => C#
返回值：正确
C++中编码：GBK
| CharSet | C#解码情况 |
|---|---|
| 不设置 | 乱码 |
| Ansi | 乱码 |
| Unicode | 乱码 |
| Auto | 乱码 |
| None | 乱码 |
C++中编码：Utf8
相关测试代码：
//依次将CharSet 设置为：不设置值、Ansi、Unicode、Auto、None，进行测试
[StructLayout(LayoutKind.Sequential, CharSet=CharSet.Ansi)]
public class StepRspInfo
{
public int ErrorID;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 256)]
public string? ErrorMsg;
}回调委托定义：
[UnmanagedFunctionPointer(CallingConvention.StdCall)]
public delegate void OnRspAdminUserLogin(StepRspAdminUserLogin? rspAdminUserLogin, StepRspInfo? rspInfo, int requestID, bool isLast);3 结论
当在与C++ API 交互时，如果在windows平台运行，建议使用GBK编码进行通信；而在Linux平台运行的话，建议使用 UTF8编码进行通信。