xxx yyy zzz aaaaa bbb cccccccccccc...
[download]

foreach (@infile) {
   /(\S+\s+\S+\s+\S+)\s+(\S+)\s+(\S+)\s+(.*)/;
   my @fields = ($1,$2,$3,$4);
   push @output, \@fields;
}
[download]

A short remark, you suggest to

populate an array (@infile) with the lines from the log file.

This is absolutely not necessary in this case, it's just a waste of memory. Why not do it like this:

open(LOG,'<','logfile') or die "Couldn't open logfile: $!";
while (<LOG>) {
  # do something
}
close LOG;
[download]

-- Hofmator

I assume you know how to use the function split. Type "perldoc -f split" if you don't. Using split, you can get all fields and join the fields you want to produce the strings required.

Now from looking at your logs, it seems like all columns have fixed widths. The first 21 chars go to the first cell in the table. Skip one field, then grab the next 13 chars, skip one and get the rest.

To match these fields, one can use a regular expression to do the job. /^(.{21}) (.{13}) (.*)/ should match 21 chars, followed by a space, followed by 13 chars, followed by a space, then the rest. In list context, the matching operator returns a list of all things between brackets.

# example:
my $str = "Sat Jul 21 22:30 2001 144.02.26.399 www.myserver.com";
my ($date, $stuff, $servers) = $str =~ /^(.{21}) (.{13}) (.*)/;

print "<table>";
for(@lines){
    print "<tr>";
    print "<td>$_</td>" for /^(.{21}) (.{13}) (.*)/;
    print "</tr>";
}
print "</table>";

Hope this helps,,,

Aziz,,,

print "<table border='3'>\n";
while (<DATA>) {
  next if m/^\s*$/;
  my @elements = split /\s+/, $_;
  print <<"  LINE";
  <tr>
    <td>@elements[0..4]</td>
    <td>$elements[5]</td>
    <td>$elements[6]</td>
  </tr>
  LINE
}
print "</table>";



__DATA__
Sat Jul 21 22:30 2001 144.02.26.399 www.myserver.com ...
Sat Jul 21 23:01 2001 352.34.12.345 www.myserver.net
[download]

I read from the DATA filehandle for this demo. You will need to open your logfile to a filehandle to read from it. You do that like this:

print "<table border='3'>\n";
open FILE "</path/to/logfile.txt" of die "Oops $!\n";
while (<FILE>) {
  next if m/^\s*$/;
  my @elements = split /\s+/, $_;
  print <<"  LINE";
  <tr>
    <td>@elements[0..4]</td>
    <td>$elements[5]</td>
    <td>$elements[6]</td>
  </tr>
  LINE
}
print "</table>";
close FILE;
[download]

Normally I would also post some explanation but as you have not posted any code I will leave it up to you to research anything in this code that is unfamiliar. You will need to add the usual HTML HEAD and BODY tags to complete a valid page.

cheers

tachyon

s&&rsenoyhcatreve&&&s&n.+t&"$'$`$\"$\&"&ee&&y&srve&&d&&print

Others have suggested various methods for parsing your logfile, to which I would add a suggestion to try looking at substr, especially if you're not yet totally comfortable with regular expressions and the uses thereof (and assuming that your records are fixed-width).

You don't say if you're planning to generate this table dynamically or to store it: if you're doing it on the fly, then I hope you're using CGI.pm, and if you are I suggest outputting the table the easy way, using the Tr and td functions.

If you have a row of data in an array variable (say, @foo), you can generate table cells for that row with the statement

my $html = td (\@foo);
[download]

So the relevant section of your code could look something like this highly off-the-cuff and untested fragment:

my @rows;
while (<LOGFILE>) {
    my @data;
    #your routine of choice to split the input into fields 
    #and stash it in @data

    push @rows, td(\@data);
}

print table (Tr (\@rows));
[download]

Which, as you see, makes the whole formatting question pretty much go away.

If you're not doing this dynamically, you could still use the CGI module (use CGI qw (:all -nodebug), but there are probably simpler solutions available on CPAN (try looking under the WWW section for HTML building solutions).