无法将“const char*"转换为“LPCWSTR {aka const wchar_t*}"

2021-12-09 00:00:00 qt char type-conversion c++ wchar-t

我在我的 C++ 代码中遇到一个我无法理解的错误.精简后的代码位在这里:

I'm getting an error in my C++ code that I can't quite make sense of. The stripped down code bits are here:

RS232Handle=OpenRS232("COM1", 9600);

HANDLE OpenRS232(const char* ComName, DWORD BaudRate)
{
    ComHandle=CreateFile(ComName, GENERIC_READ|GENERIC_WRITE, 0, NULL, OPEN_EXISTING, FILE_ATTRIBUTE_NORMAL, NULL);
}

我收到以下错误:

error: cannot convert 'const char*' to 'LPCWSTR {aka const wchar_t*}' for argument '1' to 'void* CreateFileW(LPCWSTR, DWORD, DWORD, LPSECURITY_ATTRIBUTES, DWORD, DWORD, HANDLE)'
     ComHandle=CreateFile(ComName, GENERIC_READ|GENERIC_WRITE, 0, NULL, OPEN_EXISTING, FILE_ATTRIBUTE_NORMAL, NULL);

代码取自 VS 代码，我现在使用 Qt creator.

The code was taken from VS code and I am now using Qt creator.

我该如何解决这个问题?谢谢！

How can I fix this issue? Thanks!

推荐答案

Windows CreateFile 函数实际上是一个扩展为以下之一的宏:

The Windows CreateFile function is actually a macro that expands to one of:

  • CreateFileA，它采用 const char*
    • 类型的文件路径
  • CreateFileW，它采用 const wchar_t* 类型的文件路径.
  • CreateFileA, which takes a file path of type const char*
  • CreateFileW, which takes a file path of type const wchar_t*.

(Windows API 中大多数采用字符串的函数也是如此.)

(The same is true for most of the functions in the Windows API that take a string.)

您声明了参数 const char* ComName，但显然编译时定义了 UNICODE，因此它调用了 W 版本的功能.没有从 const wchar_t* 到 const char* 的自动转换，因此出现错误.

You're declaring the parameter const char* ComName, but apparently compiling with UNICODE defined, so it's calling the W version of the function. There's no automatic conversion from const wchar_t* to const char*, hence the error.

您的选择是:

  1. 将函数参数更改为 UTF-16 (const wchar_t*) 字符串.
  2. 保留 char* 参数，但让您的函数使用 MultiByteToWideChar.
  3. 显式调用 CreateFileA 而不是 CreateFile.
  4. 在不使用 UNICODE 的情况下编译您的程序，以便宏默认扩展为 A 版本.
  5. 绑架一位著名的 Microsoft 开发人员并强迫他阅读UTF-8 Everywhere，直到他同意让 Windows 完全支持 UTF-8 作为ANSI"代码页，从而使各地的 Windows 开发人员摆脱这种宽字符的束缚.
  1. Change the function parameter to a UTF-16 (const wchar_t*) string.
  2. Keep the char* parameter, but have your function explicitly convert it to a UTF-16 string with a function like MultiByteToWideChar.
  3. Explicitly call CreateFileA instead of CreateFile.
  4. Compile your program without UNICODE, so that the macros expand to the A versions by default.
  5. Kidnap a prominent Microsoft developer and force him to read UTF-8 Everywhere until he agrees to have Windows fully support UTF-8 as an "ANSI" code page, thus freeing Windows developers everywhere from this wide-character stuff.

我不知道是否涉及绑架，但 Windows 10 1903 终于添加了支持 对于 UTF-8 作为 ANSI 代码页.

I don't know if a kidnapping was involved, but Windows 10 1903 finally added support for UTF-8 as an ANSI code page.

相关文章