($thing1, $thing2) = ($1, $2) if /^\s{9}(\S+)\s{10}(\S+)/;
[download]

($thing1, $thing2) = split;
[download]

Let us assume that this string is in $_

($meanH1, $meanH2) = /^\s{9}(.*?)\s{10}(.*?)$/;

# Or you may be able to generalize it a bit more with:
($meanH1, $meanH2) = /^\s*(.*?)\s*(.*?)\s*$/;

# Or, if the second option is true, you could event use:
($meanH1, $meanH2) = split;

# Which is a short hand version for
($meanH1, $meanH2) = split /\s+/, $_;
[download]

All of the above are rather basic examples of regex and are well documented in perlre (perldoc or perldoc.com)

If your situation involves fixed length records where each field occupies the same columns on each record, then unpack will work for you.

The Perl Cookbook p.297 has recipe 8.15 titled "Reading Fixed-Length Records" which describes using unpack:

   # $RECORDSIZE is the length of a record, in bytes.
   # $TEMPLATE is teh unpack template for the record
   # FILE is the file to read from
   # @FIELDS is an array, one element per field
   until ( eof(FILE) ) {
      read(FILE, $record, $RECORDSIZE) == $RECORDSIZE
         or die "short read\n";
      @FIELDS = unpack($TEMPLATE, $record);
   }
[download]

   #!perl -w
   use strict;
   my $record = "         none          lt2dpmnt";
   print "\$record = [$record]\n";
   my @FIELDS = unpack('a9a4a10a8', $record);
   foreach (@FIELDS) {
      print "field=[$_]\n";
   }
[download]

   C:\DOCUME~1\hmerrill.000\TEST_P~1>test_unpack.pl
   $record = [         none          lt2dpmnt]
   field=[         ]
   field=[none]
   field=[          ]
   field=[lt2dpmnt]
[download]

HTH.

Greetings all,
Just an FYI you can use an 'x' in your unpack template to remove the spaces ('x'='A null byte.'), that is unless you want the spaces.
so

my @FIELDS = unpack('a9a4a10a8', $record);
[download]


becomes

my @FIELDS = unpack('x9a4x10a8', $record);
[download]


Given your example code above the output would be:


$record = [         none          lt2dpmnt]
field=[none]
field=[lt2dpmnt]
[download]



-injunjoel
"I do not feel obliged to believe that the same God who endowed us with sense, reason and intellect has intended us to forego their use." -Galileo

check out perldoc -f split

HTH

This will not get rid of the leading whitespace on those lines (it returns a null field as the first field). But if you use ' ' instead it should work fine. That is:

my ($meanH1,$meanH2) = split ' ';
[download]

as per the documentation:

A split on /\s+/ is like a split(' ') except that any leading whitespace produces a null first field.

-enlil

($var1, $var2) = ($1,$2) if /(\S+)\s+(\S+)/;
[download]

($var1,$var2) = ($1,$2) if /^\s{9}(\S+)\s{10}(\S+)/;
[download]

-enlil

my ( $meanH1, $meanH2 );
( $meanH1, $meanH2 ) = ( $1, $2 )
    if $line =~ m/^\s{9}(\S+)\s{10}(\S+)/;
[download]

You're correct to be checking the success of your matching. I don't like solutions that skip past this important step.

The preceeding example will look for (and skip past) the leading nine whitespaces. It will then capture all contiguous non-whitespace. It will then look for and skip past the next ten whitespaces. It will then capture all remaining contiguous non-whitespace. If there's anything else on the line (like a trailing newline) it will be ignored.

($meanH1, $meanH2) = ($1, $2) if VALUES FOR$\s+(\S+)\s+(\S+)/;
[download]

I just tried the following:


#!/usr/local/perl

$test = "         foo          bar";
($var1, $var2) = ($1, $2) if ($test =~ /\s+(\S+)\s+(\S+)/);
print "Var1: $var1\nVar2: $var2";
exit;
[download]


...and it grabbed the text out and printed fine, so I'm not sure what you mean when you say none of the suggestions are working for you, can you be more specific?


If you're ever lost and need directions, ask the guy on the motorcycle.

Post your code and a couple of lines of data. The examples given _do_ work. Perhaps we are missing a part of the problem?

#!/usr/bin/perl

while ( <DATA>) {
  ($a,$b) = split ' ';
  print "split '$a','$b'\n";
  my ($a,$b) = $_ =~ /(\S+)\s+(\S+)/;
  print "match '$a','$b'\n";
}

__DATA__
         none          bing 
         some          bong 
         any          bang
[download]

output:

split 'none','bing'
match 'none','bing'
split 'some','bong'
match 'some','bong'
split 'any','bang'
match 'any','bang'
[download]

If you are matching across newlines, are you not reading line by line? If the data is in one big string, perhaps you want something more like:

#!/usr/bin/perl

my $txt = '
         none          bing 
         some          bong 
         any          bang  ';

while ( $txt =~ /^ {9}(\S+) {10}(\S+)\s*$/mg ) {
  print "'$1' '$2'\n";
}
[download]

qq