Handle unicode file names on windows

All file names should be in UTF-8 within libavformat.

This is handled by mapping the open() function to an internal one
in os_support.h for windows.

fopen() could be overridden in the same way, but if that would be
used from ffmpeg.c, it would add a dependency on an ff prefixed
internal lavf function.

Signed-off-by: Martin Storsjö <martin@martin.st>

cmdutils.c

@@ -155,6 +155,66 @@ static const OptionDef* find_option(const OptionDef *po, const char *name){
     return po;
 }
 
+#if defined(_WIN32) && !defined(__MINGW32CE__)
+/* Will be leaked on exit */
+static char** win32_argv_utf8 = NULL;
+static int win32_argc = 0;
+
+/**
+ * Prepare command line arguments for executable.
+ * For Windows - perform wide-char to UTF-8 conversion.
+ * Input arguments should be main() function arguments.
+ * @param argc_ptr Arguments number (including executable)
+ * @param argv_ptr Arguments list.
+ */
+static void prepare_app_arguments(int *argc_ptr, char ***argv_ptr)
+{
+    char *argstr_flat;
+    wchar_t **argv_w;
+    int i, buffsize = 0, offset = 0;
+
+    if (win32_argv_utf8) {
+        *argc_ptr = win32_argc;
+        *argv_ptr = win32_argv_utf8;
+        return;
+    }
+
+    win32_argc = 0;
+    argv_w = CommandLineToArgvW(GetCommandLineW(), &win32_argc);
+    if (win32_argc <= 0 || !argv_w)
+        return;
+
+    /* determine the UTF-8 buffer size (including NULL-termination symbols) */
+    for (i = 0; i < win32_argc; i++)
+        buffsize += WideCharToMultiByte(CP_UTF8, 0, argv_w[i], -1,
+                                        NULL, 0, NULL, NULL);
+
+    win32_argv_utf8 = av_mallocz(sizeof(char*) * (win32_argc + 1) + buffsize);
+    argstr_flat     = (char*)win32_argv_utf8 + sizeof(char*) * (win32_argc + 1);
+    if (win32_argv_utf8 == NULL) {
+        LocalFree(argv_w);
+        return;
+    }
+
+    for (i = 0; i < win32_argc; i++) {
+        win32_argv_utf8[i] = &argstr_flat[offset];
+        offset += WideCharToMultiByte(CP_UTF8, 0, argv_w[i], -1,
+                                      &argstr_flat[offset],
+                                      buffsize - offset, NULL, NULL);
+    }
+    win32_argv_utf8[i] = NULL;
+    LocalFree(argv_w);
+
+    *argc_ptr = win32_argc;
+    *argv_ptr = win32_argv_utf8;
+}
+#else
+static inline void prepare_app_arguments(int *argc_ptr, char ***argv_ptr)
+{
+    /* nothing to do */
+}
+#endif /* WIN32 && !__MINGW32CE__ */
+
 void parse_options(int argc, char **argv, const OptionDef *options,
                    void (* parse_arg_function)(const char*))
 {
@@ -162,6 +222,9 @@ void parse_options(int argc, char **argv, const OptionDef *options,
     int optindex, handleoptions=1;
     const OptionDef *po;
 
+    /* perform system-dependent conversions for arguments list */
+    prepare_app_arguments(&argc, &argv);
+
     /* parse options */
     optindex = 1;
     while (optindex < argc) {

libavformat/os_support.c

@@ -28,6 +28,34 @@
 #include "avformat.h"
 #include "os_support.h"
 
+#if defined(_WIN32) && !defined(__MINGW32CE__)
+#include <windows.h>
+
+#undef open
+int ff_win32_open(const char *filename_utf8, int oflag, int pmode)
+{
+    int fd;
+    int num_chars;
+    wchar_t *filename_w;
+
+    /* convert UTF-8 to wide chars */
+    num_chars = MultiByteToWideChar(CP_UTF8, 0, filename_utf8, -1, NULL, 0);
+    if (num_chars <= 0)
+        return -1;
+    filename_w = av_mallocz(sizeof(wchar_t) * num_chars);
+    MultiByteToWideChar(CP_UTF8, 0, filename_utf8, -1, filename_w, num_chars);
+
+    fd = _wopen(filename_w, oflag, pmode);
+    av_freep(&filename_w);
+
+    /* filename maybe be in CP_ACP */
+    if (fd == -1 && !(oflag & O_CREAT))
+        return open(filename_utf8, oflag, pmode);
+
+    return fd;
+}
+#endif
+
 #if CONFIG_NETWORK
 #include <fcntl.h>
 #include <unistd.h>

libavformat/os_support.h

@@ -45,6 +45,11 @@ static inline int is_dos_path(const char *path)
     return 0;
 }
 
+#if defined(_WIN32) && !defined(__MINGW32CE__)
+int ff_win32_open(const char *filename, int oflag, int pmode);
+#define open ff_win32_open
+#endif
+
 #if CONFIG_NETWORK
 #if !HAVE_SOCKLEN_T
 typedef int socklen_t;

libavformat/version.h

@@ -25,7 +25,7 @@
 
 #define LIBAVFORMAT_VERSION_MAJOR 53
 #define LIBAVFORMAT_VERSION_MINOR  0
-#define LIBAVFORMAT_VERSION_MICRO  2
+#define LIBAVFORMAT_VERSION_MICRO  3
 
 #define LIBAVFORMAT_VERSION_INT AV_VERSION_INT(LIBAVFORMAT_VERSION_MAJOR, \
                                                LIBAVFORMAT_VERSION_MINOR, \