From summer 2013, I made the patches and sent them, to enable the Doxygen to accept non-ASCII identifiers: The names of functions, files, classes, macros or namespaces. I hope these patches or functionalities will be merged into official version , but it may take months or years. I distribute the patched version of the doxygen on the DLSite.com. The product includes doxygen.exe and doxywizard.exe (for Microsoft Windows), the source codes, the Unified-Diff file to patch against offcial Git hub, some documentations to use and build, and etc.
This product includes the way to localize doxywizard GUI. You can localize doxywizard by using free Qt Linguist Tools.
As told above, You can buy from DLSite.com, If you are in Japan, you can buy from Guide page for original(Japanese)/using Yen directly. The both products are completely same.
After all patches are included into official version, the payment for this product is simply means the donation for me.
You can check example output on this site. The same exapmle is included with both the product version and trial version, given on DLSite.com(see above section). Check クラス(Classes) or ファイル(Files) and also Class Index and Class members. The entries on the indices are sorted by the sort-key, not by the name themselves.
Currently, about some weeks or some months. But many bugs are reported almost every day, the enhancement patches are given lesser priority than fixing one. It can be taken for years to make available the enhancement patches on official version.
You can see it to go Gnome Bug Tracker: search h12u.com as a reporter or a commenter within the product Doxygen, then you can see some issues got FIXED indicators.
On the Visual Studio, we can use non-ASCII identifiers for years(Microsoft .NET Framework accepts such identifiers). On the other environments, we caught up with the Visual Studio at 2013. we can do that with Clang 3.3, the part of LLVM. The release note of CLang 3.3 tells the new feature about "Extended Identifiers: Unicode Support and Universal Character Names".
Currently, the doxygen, except this product gives, does not support such identifiers. I fixed it in the end of the troubles. When using such characters on the head of the identifiers, the name will appear after the ASCII alphabets. It is not good for several cases, and I add the new command
<sort-key>. The command will work like Wikipedia/Mediawiki.
Currently, the doxywizard GUI can only show English. I translate both the GUI elements and the help documents of the options into Japanese. You can also translate by using Qt linguist.