Creating files using CSV delimiters

PCheung · #1 (**permalink**) 05-23-07, 16:14

I have been trying to create a fixed length “Name and Address” file made up of fixed fields. (Typically an old fashion mainframe file) and the input data is via the keyboard. For this example only the name fields maybe between 1 and 15 characters long only and if a longer input stream is entered then it is to be truncated

I do not know how to put the CSV field delimiter in and I not sure if there is anything special when writing this type of record as I haven’t got that far.
I've just got lost

################################################## ##
#!/usr/bin/perl
# Enter Given Name
A05010:
print "Enter Given Name (Maximum 15 Characters),\n";
open ($Customer=<STDIN>);
@Customer=/(.{15})/g;
print $Customer pack("$Customer15", $Customer);
print 'is this correct?';
print "\n";
print 'enter Y or N';
print "\n";
chomp ($Reply = <STDIN>);
goto A05010 if ($Reply eq "N" or $Reply eq "n");
################################################## #
# SHOW ME WHAT HAPPENED
################################################## #
print "\n";
print '$Customer = ', $Customer;
print "\n";
################################################## #
# Enter Family Name
A05020:
print " Enter Family Name (Maximum 15 Characters),\n";
open ($Customer=<STDIN>);
@Customer=/(.{15})/g;
print $Customer pack("$Customer15", $Customer);
print 'is this correct?';
print "\n";
print 'enter Y or N';
print "\n";
chomp ($Reply = <STDIN>);
goto A05020 if ($Reply eq "N" or $Reply eq "n");
################################################## #
# SHOW ME WHAT HAPPENED
################################################## #
print '$Customer = ', $Customer;
print "\n";

KevinADC · #2 (**permalink**) 05-24-07, 04:03

Where are you learning perl from? I haven't seen "goto" in a perl script in more than 10 years. You want to use substr() for the truncating. But your code has other problems.

http://perldoc.perl.org/functions/substr.html

sco08y · #3 (**permalink**) 07-03-07, 00:18

Looking at your code:

Code:

 #!/usr/bin/perl

The "shebang" line (because # is a sharp and ! is a bang) has to be the first line. It's optional; you can always execute your script with perl scriptname.pl. The only reason to have a shebang line is if you want your script to be executable. If you don't know what chmod +x does, don't bother with it.

Code:

open	($Customer=<STDIN>);

Here's what you're telling Perl:

Code:

<STDIN>     Read a line from standard input, STDIN.
$Customer=...  Assign that value to the scalar variable $Customer
open (...)  This is almost certainly not doing what you intended. It's opening a file whose name is the line you read from STDIN and associating that open file with a file handle of the same name.

Here's what you probably mean:

Code:

$Customer = <STDIN>

STDIN is already open for you when your script runs.

Code:

 	@Customer=/(.{15})/g;

Okay, here's a big perl trap beginners run into. $Customer and @Customer are two completely different variables. $Customer is a "scalar", which basically means it takes one simple value like a number or a string. (Perl figures out which automatically.) @Customer is an array, which means it holds multiple values.

It gets worse: $Customer[5] means the sixth value (since perl arrays are zero-based (but this can be changed (but don't do that))) of the array @Customer. There is a certain logic to it, but it's confusing enough for people that they're changing it in perl 6.

Next problem: your regular expression is going to match something that has a *minimum* of 15 characters, not a maximum. What you want is this:

Code:

use strict; # If this complains too much, try use strict 'vars' instead.
my($Customer); # Declares your variable. Do it this way until you learn more.
cust_input: print "Enter family name (15 char max)\n";
$Customer = <STDIN>;
chomp $Customer; # gets rid of the line break at the end.
unless($Customer =~ /^.{0,15}$/) {
    print "Family name too long.\n";
    goto cust_input;
}

Notice a few differences in the match statement:

1. I use =~ instead of =. The = sign means "assign a value." The =~ means "do a match against this value."

2. I added ^ and $. These match the beginning and end of the line. Regular expressions are great but they can be tricky, so it's good to read perlretut and perlre.

sco08y · #4 (**permalink**) 07-03-07, 00:34

It's late so I'm not going to walk through this extensively, but here's a subroutine you can add to help you with CSV files.

Code:

# Formats a single line of CSV data.
sub csv {
    my(@a) = @_; # Work on a copy of data
    $_ = defined ? $_ : '' for @a; # Deal with undefined data
    s/"/""/gx for @a; # Double up quotes, that's all you need for CSV
    join(',', map(qq/"$_"/, @a))."\n"; # Add outside quotes, commas and end of line
}

To use it:

Code:

print csv($givenname, $familyname, $address);

It doesn't care how many variables you give it. Just stick it at the top of your file. This is based on how Excel and other programs I've seen expect CSV files to look. If you have another piece of software, generate some data with troublesome characters (comma, double quotes, spaces, and newlines within a string) to attach and we can see how it handles them.

KevinADC · #5 (**permalink**) 07-03-07, 00:41

Quote:

Originally Posted by sco08y

Looking at your code:

Code:

 #!/usr/bin/perl

The "shebang" line (because # is a sharp and ! is a bang) has to be the first line. It's optional; you can always execute your script with perl scriptname.pl. The only reason to have a shebang line is if you want your script to be executable. If you don't know what chmod +x does, don't bother with it.

Otional on windows but it doesn't hurt anything to have it there. I sure hope they come back and read all that good info you posted.

PCheung · #6 (**permalink**) 07-03-07, 14:15

Thank you very much for your help it is very much appreciated. I did manage
to get my code working and since then have found it easier to use PHP and
MySQL to get the results I was after. However, I must say I do like PERL and
will be using it server side to do the real work of data processing and
thereafter pass the results back to PHP to carry on doing the fetching and
carrying.

Internet Services

Web Development

Digital Marketing

Consumer Tech