Update encoding_utf

file_ops.c

@@ -58,135 +58,42 @@ static int Buf_EnsureSize(CBuf *dest, size_t size)
    return Buf_Create(dest, size, &g_Alloc);
 }
 
-#ifndef _WIN32
-
-static uint8_t kUtf8Limits[5] = { 0xC0, 0xE0, 0xF0, 0xF8, 0xFC };
-
-static Bool Utf16_To_Utf8(uint8_t *dest, size_t *destLen,
-      const uint16_t *src, size_t srcLen)
+static bool Utf16_To_Char(CBuf *dest, const uint16_t *s, int fileMode)
 {
-   size_t destPos = 0;
-   size_t srcPos  = 0;
+   bool res;
+   size_t dest_len = 0;
+   unsigned len    = 0;
 
-   for (;;)
-   {
-      unsigned numAdds;
-      uint32_t value;
+   while (s[len] != '\0')
+      len++;
 
-      if (srcPos == srcLen)
-      {
-         *destLen = destPos;
-         return True;
-      }
-      value = src[srcPos++];
-      if (value < 0x80)
-      {
-         if (dest)
-            dest[destPos] = (char)value;
-         destPos++;
-         continue;
-      }
-      if (value >= 0xD800 && value < 0xE000)
-      {
-         uint32_t c2;
+   utf16_conv_utf8(NULL, &dest_len, s, len);
+   dest_len += 1;
 
-         if (value >= 0xDC00 || srcPos == srcLen)
-            break;
-         c2 = src[srcPos++];
-         if (c2 < 0xDC00 || c2 >= 0xE000)
-            break;
-         value = (((value - 0xD800) << 10) | (c2 - 0xDC00)) + 0x10000;
-      }
-      for (numAdds = 1; numAdds < 5; numAdds++)
-         if (value < (((uint32_t)1) << (numAdds * 5 + 6)))
-            break;
-      if (dest)
-         dest[destPos] = (char)(kUtf8Limits[numAdds - 1] 
-               + (value >> (6 * numAdds)));
-      destPos++;
-      do
-      {
-         numAdds--;
-         if (dest)
-            dest[destPos] = (char)(0x80 
-                  + ((value >> (6 * numAdds)) & 0x3F));
-         destPos++;
-      }while (numAdds != 0);
-   }
-   *destLen = destPos;
-   return False;
-}
-
-static SRes Utf16_To_Utf8Buf(CBuf *dest,
-      const uint16_t *src, size_t srcLen)
-{
-   Bool res;
-   size_t destLen = 0;
-
-   Utf16_To_Utf8(NULL, &destLen, src, srcLen);
-   destLen += 1;
-
-   if (!Buf_EnsureSize(dest, destLen))
+   if (!Buf_EnsureSize(dest, dest_len))
       return SZ_ERROR_MEM;
 
-   res = Utf16_To_Utf8(dest->data, &destLen, src, srcLen);
-   dest->data[destLen] = 0;
+   res = utf16_conv_utf8(dest->data, &dest_len, s, len);
+   dest->data[dest_len] = 0;
 
-   return res ? SZ_OK : SZ_ERROR_FAIL;
-}
-#endif
-
-static SRes Utf16_To_Char(CBuf *buf, const uint16_t *s, int fileMode)
-{
-   unsigned len = 0;
-
-   for (len = 0; s[len] != '\0'; len++);
-
-#ifdef _WIN32
-   {
-      int size = len * 3 + 100;
-      if (!Buf_EnsureSize(buf, size))
-         return SZ_ERROR_MEM;
-      {
-         char defaultChar = '_';
-         BOOL defUsed;
-         int numChars = WideCharToMultiByte(fileMode ?
-               (
-#ifdef UNDER_CE
-                     CP_ACP
-#else
-                     AreFileApisANSI() ? CP_ACP : CP_OEMCP
-#endif
-               ) : CP_OEMCP,
-               0, (LPCWSTR)s, len, (char *)buf->data,
-               size, &defaultChar, &defUsed);
-         if (numChars == 0 || numChars >= size)
-            return SZ_ERROR_FAIL;
-         buf->data[numChars] = 0;
-         return SZ_OK;
-      }
-   }
-#else
-   (void)fileMode;
-   return Utf16_To_Utf8Buf(buf, s, len);
-#endif
-}
-
-static SRes ConvertUtf16toCharString(const uint16_t *in, char *s, size_t len)
-{
-   CBuf buf;
-   SRes res;
-
-   Buf_Init(&buf);
-   res = Utf16_To_Char(&buf, in, 0);
-
-   if (res == SZ_OK)
-      strlcpy(s, (const char*)buf.data, len);
-
-   Buf_Free(&buf, &g_Alloc);
    return res;
 }
 
+static bool ConvertUtf16toCharString(const uint16_t *in, char *s, size_t len)
+{
+   CBuf dest;
+   bool ret;
+
+   Buf_Init(&dest);
+   ret = Utf16_To_Char(&dest, in, 0);
+
+   if (ret)
+      strlcpy(s, (const char*)dest.data, len);
+
+   Buf_Free(&dest, &g_Alloc);
+   return ret;
+}
+
 /* Extract the relative path relative_path from a 7z archive 
  * archive_path and allocate a buf for it to write it in.
  * If optional_outfile is set, extract to that instead and don't alloc buffer.
@@ -268,7 +175,7 @@ static int read_7zip_file(
          }
 
          SzArEx_GetFileNameUtf16(&db, i, temp);
-         res = ConvertUtf16toCharString(temp, infile, sizeof(infile));
+         res = ConvertUtf16toCharString(temp, infile, sizeof(infile)) ? SZ_OK : SZ_ERROR_FAIL;
 
          if (!strcmp(infile, relative_path))
          {
@@ -419,7 +326,7 @@ static struct string_list *compressed_7zip_file_list_new(
             }
          }
          SzArEx_GetFileNameUtf16(&db, i, temp);
-         res      = ConvertUtf16toCharString(temp, infile, sizeof(infile));
+         res      = ConvertUtf16toCharString(temp, infile, sizeof(infile)) ? SZ_OK : SZ_ERROR_FAIL;
          file_ext = path_get_extension(infile);
 
          if (string_list_find_elem_prefix(ext_list, ".", file_ext))

libretro-common/encodings/encoding_utf.c

@@ -75,9 +75,8 @@ size_t utf8_conv_utf32(uint32_t *out, size_t out_chars,
    return ret;
 }
 
-
 bool utf16_conv_utf8(uint8_t *out, size_t *out_chars,
-      const uint16_t *in, size_t in_size)
+     const uint16_t *in, size_t in_size)
 {
    static uint8_t kUtf8Limits[5] = { 0xC0, 0xE0, 0xF0, 0xF8, 0xFC };
    size_t out_pos = 0;
@@ -93,9 +92,7 @@ bool utf16_conv_utf8(uint8_t *out, size_t *out_chars,
          *out_chars = out_pos;
          return true;
       }
-
       value = in[in_pos++];
-
       if (value < 0x80)
       {
          if (out)
@@ -119,12 +116,10 @@ bool utf16_conv_utf8(uint8_t *out, size_t *out_chars,
       for (numAdds = 1; numAdds < 5; numAdds++)
          if (value < (((uint32_t)1) << (numAdds * 5 + 6)))
             break;
-
       if (out)
          out[out_pos] = (char)(kUtf8Limits[numAdds - 1] 
                + (value >> (6 * numAdds)));
       out_pos++;
-
       do
       {
          numAdds--;