rimvydazas has asked for the wisdom of the Perl Monks concerning the following question:

Have anyone experienced the problem like this one? Look at the script and string I have in $msg: 
#!/usr/bin/perl -w
use strict;
my $msg = '1324^@^@^@';
my $lng = length($msg);
print "The length of the message is: $lng\n";
print "This is the message itself: [$msg]\n";

[download]
When I execute it, I get the following: 
cdruser1@cdr1:~/cdr/scripts> ./test1.pl 
The length of the message is: 7
This is the message itself: [1324]

[download]
The problem is with those characters ^@^@^@. When I print those characters, I can't see nothing! However, the length, as you see, includes them. What I really want to do is to get rid of them from my string, but I am not sure what to use in substitution command. By the way, those characters in Windows are printed as spaces.

Replies are listed 'Best First'.
Re: Weird characters' issue
by ikegami (Patriarch) on Dec 06, 2007 at 19:16 UTC

    ^@ is how your editor represents the NUL character.
    In the file, each of those ^@ is the single byte chr(0). 

    >perl -e"printf qq{^%s %2d\n}, chr($_+64), $_ for 0..31"
^@  0
^A  1
^B  2
..
^Y 25
^Z 26
^[ 27
^\ 28
^] 29
^^ 30
^_ 31

    [download]
[reply]
[d/l]
[select]
Re: Weird characters' issue
by FunkyMonk (Bishop) on Dec 06, 2007 at 19:21 UTC
    They look like chr(0)'s to me. You can get rid of them using $msg =~ s/\0//g;.

[reply]
[d/l]
[select]
      That sounds like an aweful idea. Adding s/\0//g might make the program *appear* to work, but would almost certainly add a bug. It would be best to find out why they're there in the first place. For example, if they're present after every other character or so, it because text encoded as UTF-16 or UCS-2 is being treated as the text encoded in another encoding (probably iso-latin-1).
[reply]
[d/l]
        I use s/\0+//g right now and it seems to work fine. Where do I get those strings? They come on the socket I am listening to. String are generated by a phone switch. NULL characters are always added whenever the date is inserted in a string like in the example bellow: 
        120507 1324 01300            6417                  13639820545  801<CR
+><LF>13:24 12/05<CR><LF>^@^@^@13:24 12/05<CR><LF>^@^@^@120507 1324 00
+456      2867416007  808             14972681971  801<CR><LF>13:24 12
+/05<CR><LF>^@^@^@

        [download]
        I don't see any other places with those characters (not in the data I am capturing, or anywhere else), only when the timestamp is added in string. I guess it is safe to use s/\0+//g in this case, am I right?
[reply]
[d/l]

Back to Seekers of Perl Wisdom