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

So this one bit me today, and I'm wondering if it's documented anywhere or if I just am missing something in my implementation. I have a sub that sorts some lines based on the IP address at the beginning of the line:

my $fp = \&by_ip;
print $fp->(@input);

sub by_ip
{
    return map { unpack('x4A*') }
        sort
        map { pack('A4A*',inet_aton((split /\s/,$_,2)[0]),$_) } @_;
}

[download]

The output is printed all on one line, which I do not want. I am sure that every line in the input ends with "\n", since this sub is swappable (via assignment to $fp) with several others that do not exhibit this behavior. The only unique thing about this sub is that it uses pack. Does anyone have a clue as to what is going on here? Many thanks.

print pack("A25",pack("V*",map{1919242272+$_}(34481450,-49737472,6228,0,-285028276,6979,-1380265972)))

Replies are listed 'Best First'.
Re: pack auto-chomps/chops?
by ikegami (Patriarch) on Oct 07, 2009 at 18:49 UTC

    pack:

    A: A text (ASCII) string, will be space padded

    Emphasis mine. If spaces are added by pack, they'll be removed by unpack. I guess it means \s when it says space. Fix: 

    sub by_ip
{
    return
        map substr($_, 4),
        sort
        map inet_aton( (split /\s/, $_, 2)[0] ) . $_,
        @_
}

    [download]
[reply]
[d/l]
[select]

      Ah, that's tricky. I suppose you're right, though. Do you know if there's a web-browsable version of the Perl source code anywhere I could check into that?

      By the way, your solution works but is slower than just appending a "\n" to the end during the final map.

      print pack("A25",pack("V*",map{1919242272+$_}(34481450,-49737472,6228,0,-285028276,6979,-1380265972)))
[reply]
[d/l]
[select]

        your solution works but is slower than just appending a "\n" to the end during the final map

        If pack is faster — seems odd — you could use Z instead of A.

        Ah, that's tricky. I suppose you're right, though. Do you know if there's a web-browsable version of the Perl source code anywhere I could check into that?

        http://perl5.git.perl.org/perl.git Look for pp_unpack in pp*.c (probably pp.c specifically pp_pack.c).

        The relevant code: 

        1458                     for (ptr = s+len-1; ptr >= s; ptr--)
1459                         if (*ptr != 0 && !isSPACE(*ptr)) break;
1460                     ptr++;

        [download]

        and 

        #define isSPACE(c) \
        ((c) == ' ' || (c) == '\t' || (c) == '\n' || (c) =='\r' || (c)
+ == '\f')

        [download]
        aka \s

        Update: Added relevant code.

[reply]
[d/l]
[select]

Back to Seekers of Perl Wisdom