Wiggins d'Anconia wrote:
Kevin Old wrote:
Hello everyone,
I have a regex in a file upload script that I'm trying to make work
with Windows upload paths.
Here's the snippet of code I'm working with (see comments for questions):
# This is what is passed to me from the upload form
my $filename = 'C:\Documents and Settings\ata\My Documents\PDF
files\Kerr Price Protection & Titles March.xls.pdf';
# Here I'm splitting it by backslashes (\) and just popping the last
element
# that contains the filename
my @fn = split /\\/, $filename;
print join("*", @fn);
my $thefilename = pop @fn;
print "filename before mods: $thefilename\n";
Checkout File::Basename, it will handle windows and unix paths, and others.
Also File::Spec has various portable functions for working with path names.
# I want to replace any non-Alpha, non-Numeric characters *except*
periods with underscores
# Currently, I'm just removing all non-Alpha and non-Numeric chars
$thefilename =~ s/\W/_/g;
print "removed funny chars: $thefilename\n";
So you want to keep alpha numeric and dots.
$thefilename =~ s/[^a-zA-Z0-9.]/_/g;
tr/// might be better:
$thefilename =~ tr/a-zA-Z0-9./_/c;
# Here I'm consolidating underscores
$thefilename =~ s/__/_/g;
If you have 4 in a row you will still be left with 2 in a row, etc.
$thefilename =~ s/_+/_/g;
$thefilename =~ tr/_//s;
Or combine them both:
$thefilename =~ tr/a-zA-Z0-9./_/sc;
$ perl -le'$_ = q/ab&c$d%_e__f.j=k/; print; tr/a-zA-Z0-9./_/sc; print'
ab&c$d%_e__f.j=k
ab_c_d_e_f.j_k
John
--
use Perl;
program
fulfillment
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
<http://learn.perl.org/> <http://learn.perl.org/first-response>
