I have a Win32 Perl coding challenge to search a directory (WinXP, NTFS) with a file specification pattern where the directory may contain files with unicode / wide filenames in addition to ANSI filenames. Through trial, error, and searches against Perl mailing list archives, it seems apparent that Win32 Perl's builtin directory functions do not support / return Win32 unicode / wide filenames. More specifically, the builtin functions return filenames like "??????_HostID_2006-01-19_213218.xls" when the filename contains unicode / wide characters (the same as the DOS "dir" command). The problem is that you can't pass these filenames to something like stat or via OLE to ask Excel to open it.
I saw other postings that referenced the Win32 API directory search functions FindFirstFile, FindNextFile, and FindClose. I also noticed that these functions are not packaged in "Win32API::File" -- the title of which is "Low-level access to Win32 system API calls for files/dirs" (hmm, why are these low level Win32 API directory functions not included in Win32API::File?). I was able to code a test script that used the ANSI versions of these Win32API calls but it produced the same results as Perl's builtin directory functions: it returned filenames like "??????_HostID_2006-01-19_213218.xls" when the filename contains unicode / wide characters. Which is to be expected.
I then found a posting from Jan Dubois (9 Dec 2005, perl5-porters) that suggested the solution to this problem was to use Win32::OLE and the Scripting.FileSystemObject. I was able to successfully implement this method with one shortcoming: the Scripting.FileSystemObject does not support directory searches -- only directory listings (as best I can tell.) THANK YOU JAN DUBOIS!
Being stubborn and curious, I went back to fiddling with the Win32 API directory search functions to see if I could make the "wide" version of these calls work (starting with example code posted to these lists by others -- whom I thank). I think I'm close to getting these functions to work... but I'm a novice at implementing Win32 API functions in Perl -- and using Perl's Unicode functions. The following example script runs (and sometimes crashes) but for filenames with only ANSI characters, $FileInfo->{cFileName}, seems to contain only the first character. And for filenames that start with unicode, wide characthers, $FileInfo->{cFileName}, seems to contain only the leading unicode/wide characters. In the data dump output, the buffer seems to show the full unicode file name (e.g., "t^!k i<eQR_ H o s t I D _ 2 0 0 6 - 0 1 - 1 9 _ 2 1 4 3 5 8 . x l s") -- but this is not getting into $FileInfo->{cFileName}, at least, not that I can tell. I've tried multiple different decode values none of which seem to work. I'm also not sure of the proper array dimension for cFileName in the WIN32_FIND_DATAW struct. In the ANSI version of this structure, it's 260 (MAX_PATH). In some Python source code, I've seen it set to "MAX_PATH*2+5" -- which I assume is 525 which is what I've set it to below. Though if this is a null terminated C string, I'm not sure that the dimension matters.
Any ideas on how to make these functions work correctly would be greatly appreciated.
Regards,
... Dewey
use strict;
use Win32::API;
use Data::Dumper; $Data::Dumper::Indent=1; $Data::Dumper::Sortkeys=1;
use Encode qw(encode decode);
use English;
$OUTPUT_AUTOFLUSH=1;
$Win32::API::DEBUG = 0;
#binmode( STDOUT, ':encoding(cp1252)' );
binmode(STDOUT, ":utf8");
use constant ERROR_NO_MORE_FILES => 18;
use constant INVALID_HANDLE_VALUE => -1;
Win32::API::Struct-> typedef('FILETIME', qw(
DWORD dwLowDateTime;
DWORD dwHighDateTime;
)); # 8 bytes
use constant FILE_ATTRIBUTE_READONLY => 0x00000001;
use constant FILE_ATTRIBUTE_HIDDEN => 0x00000002;
use constant FILE_ATTRIBUTE_SYSTEM => 0x00000004;
use constant FILE_ATTRIBUTE_DIRECTORY => 0x00000010;
use constant FILE_ATTRIBUTE_ARCHIVE => 0x00000020;
use constant FILE_ATTRIBUTE_NORMAL => 0x00000080;
use constant FILE_ATTRIBUTE_TEMPORARY => 0x00000100;
use constant FILE_ATTRIBUTE_COMPRESSED => 0x00000800;
use constant MAX_PATH => 260;
Win32::API::Struct-> typedef('WIN32_FIND_DATAW', qw(
DWORD dwFileAttributes;
FILETIME ftCreationTime;
FILETIME ftLastAccessTime;
FILETIME ftLastWriteTime;
DWORD nFileSizeHigh;
DWORD nFileSizeLow;
DWORD dwReserved0;
DWORD dwReserved1;
WCHAR cFileName[525];
WCHAR cAlternateFileName[14];
));
my $FindFirstFile = Win32::API->new('kernel32.dll', 'FindFirstFileW', 'PS', 'N') or die "FindFirstFile: $^E";
my $FindNextFile = Win32::API->new('kernel32.dll', 'FindNextFileW', 'NS', 'I') or die "FindNextFile $^E";
my $FindClose = Win32::API->new('kernel32.dll', 'FindClose', 'N', 'I') or die "FileClose $^E";
my $FileName = "//?/C:/My Documents/Tool/*.xls\0";
my $FileInfo = Win32::API::Struct-> new('WIN32_FIND_DATAW');
#print Data::Dumper-> Dump([$FileName, $FileInfo], [qw($FileName $FileInfo)]);
my $handle = $FindFirstFile-> Call(encode("UTF-16LE", $FileName), $FileInfo);
if ($handle == INVALID_HANDLE_VALUE) {
printf "Error is %d - %s\n", Win32::GetLastError (),
Win32::FormatMessage (Win32::GetLastError ());
exit(1);
} else {
#print Data::Dumper-> Dump([$FileInfo], [qw($FileInfo)]);
print "first filename is ", decode("cp932", $FileInfo->{cFileName}), " length = ", length($FileInfo->{cFileName}), "\n";
while (my $result = $FindNextFile->Call($handle,$FileInfo)) {
print "next filename is ", decode("cp932", $FileInfo->{cFileName} ), " length = ", length($FileInfo->{cFileName}), "\n";
print Data::Dumper-> Dump([$FileInfo], [qw($FileInfo)]);
}
}
$FindClose->Call($handle) or die "FindClose $^E";
_______________________________________________ Perl-Win32-Users mailing list [email protected] To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
