http://www.perlmonks.org?node_id=916006


in reply to Attempt to parse a line with odd characters is unsuccessful.

\cI is the tab character and \cM\cJ is the dos newline. So ... splitting on whitespace and taking the third value should work. 
  my $value = (split( /\s+/, $a1[0] ))[2]

[download]
-derby

Replies are listed 'Best First'.
Re^2: Attempt to parse a line with odd characters is unsuccessful.
by jwkrahn (Abbot) on Jul 22, 2011 at 01:41 UTC
    splitting on whitespace and taking the third value should work.

      my $value = (split( /\s+/, $a1[0] ))[2]

    Unless there is leading whitespce.    This will work in either case: 

      my $value = ( split ' ', $a1[ 0 ] )[ 2 ];

    [download]
[reply]
[d/l]
[select]
      true ... true .. but there's no leading whitespace in the OP.
      -derby
[reply]
Re^2: Attempt to parse a line with odd characters is unsuccessful.
by trailhounds (Initiate) on Jul 21, 2011 at 22:27 UTC
    Derby,
    Thanks for your answer.
    Will the linux perl interpret the whitespace from Windows correctly?
    My test did not work.
    It keeps coming back when I show the value in the interpreter as "undef".
[reply]
      Whitespace is whitespace. Windows nor linux should matter (granted the eol delimiter is different but they're both whitespace). I'm guessing there's more to the code and/or data than the snippet you've shown. If you could give more details for 'It keeps coming back when I show the value in the interpreter as "undef"', we could track down the issue.
      -derby
[reply]
      Check the input file really is what you expect with something like od -xc file.txt
[reply]
[d/l]

In Section Seekers of Perl Wisdom