Windows 95 (and following versions Windows 98 and ME) implemented the A functions and provided the W functions as aliases that do Wide to Ansi conversion. Basically, each API function that takes strings as parameters has two separate versions, while all of the functions that do not refer to strings have a single one, of course. Notice that the plain version with no string format indication is just a placeholder for one of them, invariably the 'A' version in past versions of Delphi (where this code is taken from), while in Delphi 2009 (as we'll see) the default becomes the 'W' version.

The declarations are identical but use either PAnsiChar or PWideChar to refer to strings.

Name 'GetWindowTextA' function GetWindowTextA name 'GetWindowTextA' function GetWindowTextW name 'GetWindowTextW' NMaxCount: Integer): Integer stdcall function GetWindowTextW(hWnd: HWND IpString: PWideChar nMaxCount: Integer): Integer stdcall NMaxCount: Integer): Integer stdcall function GetWindowTextA(hWnd: HWND IpString: PAnsiChar Most Windows API functions have two versions available, an Ansi version marked with the letter A and a wide-string version marked with the letter W.ġ5 As we'll see in the next chapter, the new Character unit includes methods for checking whether a Unicode code point is a symbol, a punctuation mark, a space.Īs an example, see the following lines from Windows.pas of Delphi 2007:įunction GetWindowText(hWnd: HWND IpString: PChar Since the early days, the Win32 API (which dates back to Windows NT) has included support for Unicode characters.

Getwindowtext windows api reference