John W. Krahn wrote:
> Steve Bertrand wrote:
[ snip ]
>> In the third line of the following code, I want the replacement pattern
>> to be (whitespace * $_[1]). I'll deal with the undef issue after I
>> figure out how to use {$num} on essentially nothing. I haven't been able
>> to make it work with the likes of s/blah/\s{3}/g; etc:
>>
>> sub to_word ($;$) {
>> return unless defined $_[0] && $_[0] ne '';
>> my $word = $_[0] =~ s/[^\w\s]/\s{3}/g;
>> return unless defined $word;
>> }
> In the substitution operator, the left part is a regular expression and
> the right part is just a string, so the code above is replacing a single
> character matching '[^\w\s]' with the string 's{3}' (the backslash is
> interpolated away.)
>
> You cannot put "whitespace" in the string half because whitespace is a
> regular expression character class and it has no meaning in a string
> context.
>
> If you want to replace a pattern with the value of $_[1] space
> characters then you could do this:
>
> s/blah/ ' ' x $_[1] /eg;
>
> The /e option treats the string on the right as Perl code and evaluates
> it and returns the result.
Wow. Thanks for such detailed information as to why it was broken, what
needed to be changed, and most importantly why it needed to be changed.
I wasted about two hours of my coding time trying to come up with a
special JAPH, but I just haven't been able to make map() do what I think
it should do ;)
Anyway, for the archive, using a patched version of Data::Type,:
#!/usr/bin/perl
use warnings;
use strict;
use Data::Types qw(:word);
my $string = "hello%world!";
print "orig: $string\n";
to_word($string, 1);
print "new: $string\n";
__END__
pearl% ../data-types/datatypes.pl
orig: hello%world!
new: hello world
...and the patch (wohoo!, my first real attempt at supplying a patch to
a real person!) Please, please give me feedback on the following patch
before I submit it. Does it look ok? Did I update everything
accordingly? Is my "diff" command acceptable for general use?:
http://ipv6canada.com/data-type.patch
Steve
ps. I've errored out on function calls to the module, and haven't had
time to look into whether this module responds with anything other than
a bool. Just look at the patch, and let me know if I'm on my way, capiche?
smime.p7s
Description: S/MIME Cryptographic Signature
