I have done a bit more research and I find that I was wrong in my assumptions.

The file and path in my example was NOT "O:\TestPath\*.FXC". It was actually "O:\TestPath\????????.FXC".

The strange thing is that on a 64bit file system this will not find any files which have names not exactly 8 bytes long (excluding the suffix which MUST be "FXC"). By contrast on a 32bit file system all files are found which only have the "FXC" suffix.

This can be demonstrated outside the system.

Create three text files 12900508.FXC, LONGFILE12900510.FXC and LONG FILE 12900509.FXC.

Place the files in "O:\TestPath\" where "O:" is on a 64bit file system. Place the same files in "C:\TestPath\" where "C:" is on a 32bit file system.

Open a CMD session and DIR "O:\TestPath\????????.FXC". One file will get listed (12900508.FXC).

Now DIR "C:\TestPath\????????.FXC". ALL the files will get listed.

Sorry, for posting the query. There is no real problem in my code. I need to re-examine the logic of using the "?" wildcard.

However, any comments on the issue are welcome.

pfc_n_cst_filsrvwin32 ist not the right class for unicode systems! You have to use pfc_n_cst_filesrvunicode. This uses the unicode functions FindFirstFileW and FindNextFileW.

Use should use the global function f_setfilesrv for instantiation. It selects the right class.

example:

n_cst_filesrv lnv_filesrv

f_SetFileSrv (lnv_filesrv, TRUE)  // create

// call functions of lnv_filesrv

f_SetFileSrv (lnv_filesrv, FALSE)  // destroy

Try using * for the wildcard, it will pick up file names with any length.