GetprivateProfi leString() and WritePrivatePro fileString() problem with UTF8 ini files

[xoinki]

hi all,
I have a ini file in UTF-8, I am using GetPrivateProfi leStringA() for reading and WriteprivatePro fileStringA() for writing into the ini file. I am writing multibyte data into the file (using normal char *). so ini file is saved in UTF-8. And ini file is maintained in App data folder in windows systems.

Now the problem is.
if a user has unicode username, then SHgetFolderPath () -- with CSIDL_APPDATA option would return a path with that unicode username as part of it. I am handling all these in utf8. now GetPrivateProfi leStringA() is failing with this path since it can only handle ANSI paths.

I am Not using GetprivateProfi leStringW() and WritePrivatePro fileStringW() since it wont save ini files in unicode! it is saving ini files inj ANSI eventhough I supply unicode data.

Waiting for your Reply,
Xoinki

[weaknessforcats]

You do not ever call GetPrivateProfi leStringA() or GetProfileStrin gW().

Instead, you include tchar.h and call GetPrivatePrile String(). Read this.

Notice the arguments s LPCTSTR, which are long pointers to constant TCHAR strings.

Whether ASCII or Unicode is used, is determined by the character set you specify in your project settings. GetPrivateProfi leString() will resolve to GetPrivateProfi leStringA() or GetProfileStrin gW() based on that setting. And the argiments will resolve to LPCSTR or LPCWSTR.

Be sure to use a _tmain() and observe all _t and _T functions plus use the TEXT macro. If you don't things won't work.

[xoinki]

hi,
Thnx for the reply.
I am compiling My application with UNICODE and _UNICODE defined.
so if i use Getprivateprofi lestring() then it would internally use GetPrivateProfi leStringW().. My problem is if i use this function to write unicode data into ini.. It should save ini file in Unicode format.. but it doesnt..
http://blogs.msdn.com/michkap/archive/2006/09/15/754992.aspx
So what I am doining is I am using WritePrivatePro fileStringA() and writing UTF-8 data so forcing the file to be in UTF-8. but this works in all cases except when the path parameter has some utf-8 characters. It fails this case because WritePrivatePro fileStringA() doesnt understand Utf-8 path.

So I am in dilemma of what to do so that ini file is in Utf8.

Is there any way to maintain ini file in utf8 without using WritePrivatePro fileStringA() ?

Regards,
Xoinki

[xoinki]

hi,
I found out that opening a ini file and writing Unicode BOM explicitly would write unicode data properly when WriteProfileStr ingW() is used.. but this doesnt help since I want ini to be in Utf-8 any idea how i can maintain ini file in Utf8 and use utf8 paths??

Regards
Xoinki

[xoinki]

hi,
I got the problem solved by using a customatized ini-handler. Not using any MS API's, It just creates a separate Linked list for storing section info, key info, value info.. and a simple parser for fetching desired values.. It is funny that we are replacing MS API's coz of one limitation or other.. Earlier it was Codepage detection API's which are flawed(although quick).

Thnx
Xoinki