Question Read a text file on a user's machine and write to the screen (HTML)

[Toby Corballis]

Hi folks, this is driving me nuts so I do hope someone can help. It seems to me that this should be simple but everything I try I get: No such file or directory

What I want to do is read a text file from a user's machine and write the contents to a new page that I display. Here is (my not so elegant) initial script (though this will be tidied up once it works):

# Print the file upload form
print qq{<FORM ENCTYPE="multipart/form-data" ACTION="xxxxxx.pl"\n
METHOD="POST"> <p>${sfont}Please select the log file you want to process:${efont}<BR>\n
<INPUT TYPE="FILE" NAME="file"> <p> <INPUT TYPE="submit"> </FORM>} ;

Now the relevant bits from the next script (xxxxxx.pl):

use strict ;
use CGI::Carp qw/warningsToBrowser fatalsToBrowser/ ;
use DBI;
use File::Basename;

my $file = param('file') ;
$file=~m/^.*(\\|\/)(.*)/;
my $infile = upload('file');
open(FILE, "<$infile") or die "$!" ;

At which point is dies with the following message on screen:

No such file or directory

I cannot fathom it for the life of me . Before I go completely bonkers and pull the one remaining hair out, can anyone help?

Many thanks

[malcomosx]

Hello Toby

I hope this helps, its a bit of code i use my self

<form action='../cgi-bin/update/getupdate.pl' method='post' enctype='multipart/form-data' target='updateFrame'>
<b>Update Firmware </b><hr>
<input type='file' name='getthisfile'><input type='submit' name='buttona' value='Update'>
</form>

and the actual perl bits to save the file that go into the perl script

use CGI ':standard';
my $query = new CGI;
my $upload_filehandle = $query->upload("getthisfile");
$ps=$ps.$upload_filehandle;
open ( UPLOADFILE, "> /mnt/img/update.pl" ) or print "\nERROR: Unable to create new file (Not uploaded) <br>\n";
binmode UPLOADFILE;
while (<$upload_filehandle>){
	print UPLOADFILE;
}
close UPLOADFILE;

Chris

[Toby Corballis]

Thanks Chris. For some unknown reason I couldn't get anything to work until I stopped trying to open the file. So what ended up working was to get the name of the file from CGI, then treat it as a bytestream:

my $infile = upload('file'); # Get the special file handle from CGI
my $filesize = -s $infile ; # Get the file size
# File handling varaiables
my ($bytesread, $buffer, $str, $hex, $rec_size) ;
$bytesread = read($infile, $buffer, $filesize) ; # Have to read the whole file at once
# The loop round until the end of file is hit
while ($i <= $filesize) {
$chk = substr $buffer, $i, 1 ;
$hex = unpack('H*', $chk) ; # Get the hex representation of the current character
if ($hex eq '0a') { # We've hit the CRLF character so put rec into an array
$end = $i - $start ; # Set the length of the record
.... do stuff
$start = $i+1 ; # Increment the start parameter
#push (@records, $rec) ; # Get the record into the array
}

Odd but it worked.