Handle both an ASCII (UTF8) and a WIDE (UTF16) version of image/vm name/path

Note:
the Microsoft windows API mostly uses the W version (for enabling internationalized image name/path)
the image uses UTF8 encoded bytes string for communication with the VM (this is best for compatibility with Unix/Mac)
The idea here is that the implementation maintains both versions of the UTF8 and UTF16 path/name

The appropriate macro returning a TCHAR * are also provided.
This is in order to support the generic version using TCHAR, which are normally used to ease transition to UNICODE.

Note about string length:
No effort has been made so far to support long path names for image and VM.
The path is limited to MAX_PATH in UTF16.
UTF8 can eventually consume more characters than UTF16 (not necessarily more bytes).
Thus, the ASCII version has been made longer (MAX_PATH_UTF8) in order to avoid an even more restrictive limit.

Author: nicolas-cellier-aka-nice

platforms/win32/vm/sqWin32.h

@@ -259,21 +259,46 @@ int reverse_image_words(unsigned int *dst, unsigned int *src,int depth, int widt
 /********************************************************/
 /* Declarations we may need by other modules            */
 /********************************************************/
-extern char imageName[];		/* full path and name to image */
-extern TCHAR imagePath[];		/* full path to image */
-extern TCHAR vmPath[];		    /* full path to interpreter's directory */
-extern TCHAR vmName[];		    /* name of the interpreter's executable */
-extern char windowTitle[];             /* window title string */
-extern char vmBuildString[];            /* the vm build string */
-extern TCHAR windowClassName[];    /* class name for the window */
+
+/* Note: a character can require up to 4 bytes in UTF8 encoding
+   But the expansion from UTF16 -> UTF8 is never more than 3 bytes for 1 short
+   U+ 0000-U+  007F - 1byte in utf8, 1 short in utf16.
+   U+ 0080-U+  07FF - 2byte in utf8, 1 short in utf16.
+   U+ 0800-U+  FFFF - 3byte in utf8, 1 short in utf16.
+   U+10000-U+10FFFF - 4byte in utf8, 2 short in utf16.
+*/
+#define MAX_PATH_UTF8 (MAX_PATH*3)
+
+extern char  imageName [];       /* full path and name to image - UTF8 */
+extern WCHAR imageNameW[];       /* full path and name to image - UTF16 */
+extern char  imagePathA[];       /* full path to image - UTF8 */
+extern WCHAR imagePathW[];       /* full path to image - UTF16 */
+extern char  vmPathA[];          /* full path to interpreter's directory - UTF8 */
+extern WCHAR vmPathW[];          /* full path to interpreter's directory - UTF16 */
+extern char  vmNameA[];          /* name of the interpreter's executable - UTF8 */
+extern WCHAR vmNameW[];          /* name of the interpreter's executable - UTF16 */
+extern char windowTitle[];       /* window title string - UTF8 */
+extern char vmBuildString[];     /* the vm build string */
+extern TCHAR windowClassName[];  /* class name for the window */
+
+#ifdef UNICODE
+#define imageNameT imageNameW /* define the generic TCHAR* version */
+#define imagePath  imagePathW
+#define vmName vmNameW
+#define vmPath vmPathW
+#else
+#define imageNameT imageNameA
+#define imagePath  imagePathA
+#define vmName vmNameA
+#define vmPath vmPathA
+#endif
 
 extern UINT SQ_LAUNCH_DROP;
 
 extern const TCHAR U_ON[];
 extern const TCHAR U_OFF[];
 extern const TCHAR U_GLOBAL[];
-extern const TCHAR U_SLASH[];
-extern const TCHAR U_BACKSLASH[];
+extern const WCHAR W_BACKSLASH[];
 
 #ifndef NO_PREFERENCES
 extern HMENU vmPrefsMenu;         /* preferences menu */
@@ -380,7 +405,7 @@ void vprintLastError(TCHAR *fmt, ...);
 /******************************************************/
 /* Misc functions                                     */
 /******************************************************/
-DWORD SqueakImageLength(TCHAR *fileName);
+DWORD SqueakImageLength(WCHAR *fileName);
 int isLocalFileName(TCHAR *fileName);
 
 #ifndef NO_PLUGIN_SUPPORT

platforms/win32/vm/sqWin32Main.c

@@ -130,7 +130,7 @@ TCHAR consoleBuffer[4096];
 char stderrName[MAX_PATH+1];
 char stdoutName[MAX_PATH+1];
 
-static  char vmLogDirA[MAX_PATH];
+static  char vmLogDirA[MAX_PATH_UTF8];
 static WCHAR vmLogDirW[MAX_PATH];
 
 TCHAR *logName = TEXT("");             /* full path and name to log file */
@@ -462,7 +462,7 @@ char *ioGetLogDirectory(void) {
 }
 
 sqInt ioSetLogDirectoryOfSize(void* lblIndex, sqInt sz) {
-  if (sz >= MAX_PATH) return 0;
+  if (sz >= MAX_PATH_UTF8) return 0;
   strncpy(vmLogDirA, lblIndex, sz);
   vmLogDirA[sz] = 0;
   MultiByteToWideChar(CP_UTF8, 0, vmLogDirA, -1, vmLogDirW, MAX_PATH);
@@ -1405,7 +1405,7 @@ sqImageFile findEmbeddedImage(void) {
 	int start;
 	int length;
 
-	f = sqImageFileOpen(vmName, "rb");
+	f = sqImageFileOpen(vmNameA, "rb");
 	if(!f) {
 		MessageBox(0,"Error opening VM",VM_NAME,MB_OK);
 		return 0;
@@ -1441,6 +1441,7 @@ sqImageFile findEmbeddedImage(void) {
 	/* Gotcha! */
 	sqImageFileSeek(f, sqImageFilePosition(f) - 4);
 	strcpy(imageName, name);
+	MultiByteToWideChar(CP_UTF8,0,imageName,-1,imageNameW,MAX_PATH,NULL,NULL);
 	imageSize = endMarker - sqImageFilePosition(f);
 	return f;
 }
@@ -1577,7 +1578,7 @@ sqMain(int argc, char *argv[])
   }
 
   SetupFilesAndPath();
-  ioSetLogDirectoryOfSize(vmPath, strlen(vmPath));
+  ioSetLogDirectoryOfSize(vmPathA, strlen(vmPathA));
 
   /* release resources on exit */
   atexit(Cleanup);
@@ -1620,7 +1621,7 @@ sqMain(int argc, char *argv[])
 
 
   if(!imageFile) {
-    imageSize = SqueakImageLength(toUnicode(imageName));
+    imageSize = SqueakImageLength(imageNameW);
     if(imageSize == 0) printUsage(2);
   }
 
@@ -1650,13 +1651,13 @@ sqMain(int argc, char *argv[])
 
 #if !NewspeakVM
     /* set the CWD to the image location */
-    if(*imageName) {
-      char path[MAX_PATH+1], *ptr;
-      strcpy(path,imageName);
-      ptr = strrchr(path, '\\');
+    if(*imageNameW) {
+      WCHAR path[MAX_PATH+1], *ptr;
+      wcsncpy(path,imageNameW,MAX_PATH);
+      ptr = wcsrchr(path, '\\');
       if(ptr) {
-	*ptr = 0;
-	SetCurrentDirectory(path);
+        *ptr = 0;
+        SetCurrentDirectoryW(path);
       }
     }
 #endif /* !NewspeakVM */
@@ -1723,9 +1724,8 @@ WinMain(HINSTANCE hInst, HINSTANCE hPrevInstance, LPSTR lpCmdLine, int nCmdShow)
 
   /* fetch us the name of the executable */
   {
-    WCHAR vmNameW[MAX_PATH];
     GetModuleFileNameW(hInst, vmNameW, MAX_PATH);
-    WideCharToMultiByte(CP_UTF8, 0, vmNameW, -1, vmName, MAX_PATH, NULL, NULL);
+    WideCharToMultiByte(CP_UTF8, 0, vmNameW, -1, vmNameA, MAX_PATH_UTF8, NULL, NULL);
   }
 	/* parse the command line into the unix-style argc, argv, converting to
 	 * UTF-8 on the way. */
@@ -2086,7 +2086,7 @@ SubsystemType()
     IMAGE_OPTIONAL_HEADER image_optional_header;
 
     /* Open the reference file. */ 
-    hImage = CreateFile(vmName,
+    hImage = CreateFileW(vmNameW,
                         GENERIC_READ,
                         FILE_SHARE_READ,
                         NULL,
@@ -2150,7 +2150,8 @@ parseGenericArgs(int argc, char *argv[])
 
 	if (*imageName == 0) { /* only try to use image name if none is provided */
 		if (*argv[0] && IsImage(argv[0])) {
-			strcpy(imageName, argv[0]);
+			strncpy(imageName, argv[0],MAX_PATH_UTF8);
+			MultiByteToWideChar(CP_UTF8, 0, imageName, -1, imageNameW, MAX_PATH);
 			/* if provided, the image is a vm argument. */
 			vmOptions[numOptionsVM++] = argv[0];
 		}

platforms/win32/vm/sqWin32Prefs.c

@@ -194,15 +194,17 @@ void LoadPreferences()
 
   /* get image file name from ini file */
   size = GetPrivateProfileString(U_GLOBAL, TEXT("ImageFile"), 
-			 TEXT(""), imageName, MAX_PATH, squeakIniName);
+			 TEXT(""), imageNameT, MAX_PATH, squeakIniName);
   if(size > 0) {
     if( !(imageName[0] == '\\' && imageName[1] == '\\') && !(imageName[1] == ':' && imageName[2] == '\\')) {
       /* make the path relative to VM directory */
-      lstrcpy(imageName, vmName);
-      (lstrrchr(imageName,U_BACKSLASH[0]))[1] = 0;
-      size = lstrlen(imageName);
+      strcpy(imageName , vmNameA);
+	  wcscpy(imageNameW, vmNameW);
+      (strrchr(imageName ,'\\'           ))[1] = 0;
+	  (wcsrchr(imageNameW, W_BACKSLASH[0]))[1] = 0;
+      size = lstrlen(imageNameT);
       size = GetPrivateProfileString(U_GLOBAL, TEXT("ImageFile"), 
-			 TEXT(""), imageName + size, MAX_PATH - size, squeakIniName);
+			 TEXT(""), imageNameT + size, MAX_PATH - size, squeakIniName);
 	}
   }
 

platforms/win32/vm/sqWin32Service.c

@@ -413,12 +413,12 @@ void sqServiceInstall(void)
       return;
     }
   /* Add the image name to the subkey. */
-  ok = RegSetValueEx(hk,        /* subkey handle         */
-      TEXT("Image"),            /* value name            */
-      0,                        /* must be zero          */
-      REG_EXPAND_SZ,            /* value type            */
-      (LPBYTE) imageName,       /* address of value data */
-      strlen(imageName) + 1);   /* length of value data  */
+  ok = RegSetValueExW(hk,                     /* subkey handle         */
+      L"Image",                               /* value name            */
+      0,                                      /* must be zero          */
+      REG_EXPAND_SZ,                          /* value type            */
+      (LPBYTE) imageNameW,                    /* address of value data */
+      (wcslen(imageNameW)+1)*sizeof(WCHAR));  /* length of value data  */
   if(ok != ERROR_SUCCESS)
     {
       printLastError(TEXT("RegSetValueEX failed"));
@@ -517,14 +517,15 @@ DWORD WINAPI sqThreadMain(DWORD ignored)
       TerminateThread(GetCurrentThread(), 0);
     }
   /* Read the image name from the subkey. */
-  dwSize = MAX_PATH;
-  ok = RegQueryValueEx(hk,TEXT("Image"),NULL, &dwType, (LPBYTE) imageName, &dwSize);
+  dwSize = MAX_PATH*sizeof(WCHAR);
+  ok = RegQueryValueExW(hk,L"Image",NULL, &dwType, (LPBYTE) imageNameW, &dwSize);
   if(ok != ERROR_SUCCESS || dwType != REG_EXPAND_SZ)
     {
       sqStopService(TEXT("Failed to read the image name from registry"));
       TerminateThread(GetCurrentThread(), 0);
     }
-  imageName[dwSize] = 0;
+  imageNameW[(dwSize+1)/2] = 0;
+  WideCharToMultiByte(CP_UTF8, 0, imageNameW, -1, imageName, MAX_PATH_UTF8, NULL, NULL);
   /* Read the log file name from the subkey. */
   dwSize = MAX_PATH;
   ok = RegQueryValueEx(hk,TEXT("Log"),NULL, &dwType, (LPBYTE) &tmpString, &dwSize);