Re: patch: let msdosfs(vfat)/ntfs to support UTF-8 locale well

En. It is a good idea. When to patch msdosfs, I first try to use
kiconv and found that non-ascii code always convert fail. I have the
source code a quickview and sure that kiconv not support UTF-8. Not
familiar with kiconv, I hardcode Into the msdosfs source code. If
kiconv support UTF-8, msdosfs only needs a few change.
It seems someone also  do similar job. the link:
http://people.freebsd.org/~imura/kiconv/
the related archive message:
http://www.mail-archive.com/freebsd-hackers_at_freebsd.org/msg57285.html

2009/2/13 Gleb Kurtsou <gleb.kurtsou_at_gmail.com>:
> On (10/02/2009 17:12), 裴国兴 wrote:
>> I write a patch to support UTF-8 locale well. I think it maybe help
>> for some FreeBSD user.
>> follow link is the patch (base on FreeBSD 7.1):
>> http://btload.googlegroups.com/web/msdosfs.patch?gda=MzIscT8AAABs_gmy4a1S9lRiXjEy-V5OpwtI67JnIGlz0zr18tjObOtoi5oIt3BJMRGeqGBbbj-ccyFKn-rNKC-d1pM_IdV0
>> the full tar.bz2 package:
>> http://btload.googlegroups.com/web/msdosfs.tar.bz2?gda=IG1pBkEAAABs_gmy4a1S9lRiXjEy-V5OpwtI67JnIGlz0zr18tjObNLRc95Ps2S1UISaL0WhuitTCT_pCLcFTwcI3Sro5jAzlXFeCn-cdYleF-vtiGpWAA
>>
>> I also will patch for ntfs driver
>> http://btload.googlegroups.com/web/ntfs.patch?gda=OqsHoDwAAABs_gmy4a1S9lRiXjEy-V5O7RN7t-m4MjZ-5dQn_EvaqDVCWO9_HyYEQJyRQYPtRCL9Wm-ajmzVoAFUlE7c_fAt
>> http://btload.googlegroups.com/web/ntfs.tar.bz2?gda=zErXED4AAABs_gmy4a1S9lRiXjEy-V5O7RN7t-m4MjZ-5dQn_EvaqG3K0t6fVz8SMYStF_2dqCPjsKXVs-X7bdXZc5buSfmx
>>
>> The Chinese characters in the fat32 partition can be displayed correctly now.
>>
>> when mount windows partitions, you should do like this:
>> mount_ntfs -C UTF-8 /dev/ad?s? /path/to/mount
>> mount_msdosfs -L zh_CN.UTF-8 /dev/ad?s? /path/to/mount
>
> I've also looked at it recently but my idea was to update kiconv to
> handle ucs-2 (or whatever it is) -> utf-8 conversion in kernel. kiconv
> supports different translation methods. Current implementation just
> creates a table for ucs2->utf-8 conversion and passes it too kernel. It
> seems it can't even handle all of unicode chars.
> Fixing kiconv would be a better solution, imho.
>
>