Link to home
Start Free TrialLog in
Avatar of sunny82
sunny82

asked on

how to extract data from a table into a fixed length file

Hi,

I have to extract data from table A. Now A has about 200 columns. Now when I am writing data out into a file, every field should have a pre-specified length. How should I do it in perl?  Any ideas?

like for example -  column id should be 10 characters, name should be 30 characters and address should be 40. So in the output file, name should start from 11th position and address from 1st position

Regards,
Sayantan
Avatar of sunny82
sunny82

ASKER

I mean address from 41st position
ASKER CERTIFIED SOLUTION
Avatar of tel2
tel2
Flag of New Zealand image

Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
Avatar of sunny82

ASKER

Hi,

Many thanks for your response. I used that and it works fine. But the problem is, I have over 200 columns in the table and for each column, I have to write sprintf("%10s%-30s%-40s", $id, $name, $address), corresponding to their widths, which becomes very unwieldy and difficult to maintin.

Is there a way to use config file such that I only need to change the config file once(if there is any field width change in any column) and I can generate the file accordingly?

The config file should just have column name and its width.
If you put fields in this format in a file called config.txt:
id,10
name,30
address,40

Open in new window

You could then do your processing like this:

open CONFIG, "<config.txt" or die "Can't open config file!";
while (<CONFIG>)
{
        chomp;
        ($name, $length) = split(',');
        $format .= "%${length}s";   # This will left-justify all fields
}

# Populate the @values array with the column values of a
# a row of your table.  I'll just hard code it here, but you
# will probably read a record from the table.
@values = (123, 'My Name', 'My Address');

printf($format, @values);

Open in new window

;
Avatar of sunny82

ASKER

I am getting the output like this

%10 s%30 s123


below is my code

use strict;

my $core_config = "$ENV{'CFG'}/menir-core-config-test.txt";
my $name;
my $length;
my $format;

open CONFIG, "<$core_config" or die "Can't open config file!";
while (<CONFIG>)
{
        chomp;
        ($name, $length) = split(',');
        $format .= "%${length}s";   # This will left-justify all fields
}

my $outfile = "$ENV{'DATA'}/CORE.txt";

my @row = (123, 'My Name', 'My Address');
printf($format, @row);
----------------------------------------------

this is my config ->

id,10
name,30
address,10
Hi Sunny,

Your version of the code works for me.  All I changed was:
    my $core_config = "$ENV{'CFG'}/menir-core-config-test.txt";
to:
    my $core_config = 'config.txt';
so it would work in my environment, and I got this output:
       123                       My Name                              My Address

Open in new window


Question:
Does my version of the code work for you (if you just copy/paste it from my previous post)?

tel2
Specifying the field width as part of the csv data is ok, but doesn't scale well.

A better approach would be to use an actual config file for the format specification.  There are a number of config modules to choose from on CPAN.  Here's an example using a YAML module/format.

config.yml
fields:  # list of fields in desired order
    - id
    - name
    - address
    - salary

---  # field format specifications
id:
    width: 10
    justify: '-'
    type: d

name:
    width: 30
    type: s

address:
    width: 30
    type: s

salary:
    width: 12
    type: f
    precision: .2

Open in new window


Test script:
#!/usr/bin/perl

use 5.010;
use strict;
use warnings;
use YAML::Tiny;

my $conf   = YAML::Tiny->read( 'config.yml' );
my $fields = $conf->[0]{fields};
my $format;

foreach my $field ( @$fields ) {
    no warnings 'uninitialized';

    $format .= '%' . $conf->[1]{$field}{justify}
                   . $conf->[1]{$field}{width}
                   . $conf->[1]{$field}{precision}
                   . $conf->[1]{$field}{type};
}

my @row = (123, 'My Name', 'My Address', 12.5);
printf($format, @row);

Open in new window


Output:
D:\test>sunny82.pl
123                              My Name                    My Address       12.50
Avatar of sunny82

ASKER

Hi,

Sorry for replying a bit late.

@tel2: The output I am getting is "%10 s%30 s       123"

The code is exactly the same as your code, word to word.

Is it because of tis statement ?

$format .= "%${length}s";   # This will left-justify all fields

@FishMonger: @INC does not have YAML::Tiny. Can we use some other module?
Hi Sunny,

"> The code is exactly the same as your code, word to word."
Perhaps, but the code you pasted above is not the same as mine.
Please copy/paste my code to your machine (again), run it, then copy/paste my code and the results here for analysis.

"> Is it because of tis statement ?
> $format .= "%${length}s";   # This will left-justify all fields"

Dunno.  That statement works in my Perl 5.10 Linux environment I'm using, in both your version and my version of my code.

I doubt this is an issue, but what version of Perl are you using?  Type "perl -v" and paste the output here.

Thanks.
tel2
What operating system are you running, Sunny?  I'm assuming UNIX/Linux, below.

Further to my post above, here's an alternative way to write that format line (line #6 in my 2nd post):
    $format .= '%' . $length . 's';
If that doesn't work in my script, please run each of the following one-liners from the Linux command line, and post the command and the output here:
    perl -e '$length=12;$format .= "%${length}s";print $format'

    perl -e '$length=12;$format .= "%" . $length . "s";print $format'

When you copy/paste my code back here for analysis (see my last post for details), please use this UNIX/Linux command to display the code before copying it:
    cat -vet the-script-name-of-my-version-of-the-code.pl
Hopefully that will show if there have been any strange characters inserted during the copy/paste to your system.