«

I have a string like this:
-rwx------ 1 user group 0 Jan 06 15:07 a file with spaces

and im placing that string in an array with this command:
my @dline = split(' ', $dirline);

now notice that there are three spaces between 'with spaces'
so when i re-stringify this array using this command:
my $filename = join(' ', @dline);

$filename now contains 'a file with spaces'
notice only one space between 'with spaces'

did split eat my extra spaces? i doubt it would have eaten other chars.
any suggestions or explanations? perhaps another way of doing this?

thanks, and i hope my post hasnt upset you very much, have a nice day

"$_"@_.% wrote:
>
> I have a string like this:
> -rwx------ 1 user group 0 Jan 06 15:07 a file with spaces
>
> and im placing that string in an array with this command:
> my @dline = split(' ', $dirline);
>
> now notice that there are three spaces between 'with spaces'
> so when i re-stringify this array using this command:
> my $filename = join(' ', @dline);
>
> $filename now contains 'a file with spaces'
> notice only one space between 'with spaces'
>
> did split eat my extra spaces? i doubt it would have eaten other chars.
> any suggestions or explanations? perhaps another way of doing this?
>
> thanks, and i hope my post hasnt upset you very much, have a nice day

split usually splits on a pattern. As a special case, a space (' ') will
split on the pattern /\s+/, to emulate awk's default behaviour.

see perldoc -f split and scroll almost to the end.

--
Josef Möllers (Pinguinpfleger bei FSC)
If failure had no penalty success would not be a prize
-- T. Pratchett

> Josef M=F6llers (Pinguinpfleger bei FSC)
> If failure had no penalty success would not be a prize
> -- T. Pratchett

wrote:

> split usually splits on a pattern. As a special case, a space (' ') will
> split on the pattern /\s+/, to emulate awk's default behaviour.
>
> see perldoc -f split and scroll almost to the end.
>

Thank you very much that is very helpfull.

What i ended up doing was replacing the spaces with *'s which should work out
since a * is illegal in a filename (i think).

#example raw dir lines
#need to distinguish btwn dirs and files and reformat for display
#desired format is: file size or <DIR> date
#-rwx------ 1 user group 217752 Nov 05 1995 file.exe
#drwx------ 1 user group 0 Jan 07 11:57 dir3
#-rwx------ 1 user group 0 Jan 06 15:07 a file with spaces
#-rwx------ 1 user group 13 Jan 06 15:08 file.txt

$dirline =~ s/\s/*/g;

>>>>> "J" == <_@_.%> writes:

J> Thank you very much that is very helpfull.

J> What i ended up doing was replacing the spaces with *'s which should work out
J> since a * is illegal in a filename (i think).

no it isn't.

on unix flavors, the only char not allowed in a filename is /

instead of split, use a regex so you get the fields you want. and there
is at least one module on cpan that parses out ls -l listings.

uri

--
Uri Guttman ------ -------- http://www.stemsystems.com
--Perl Consulting, Stem Development, Systems Architecture, Design and Coding-
Search or Offer Perl Jobs ---------------------------- http://jobs.perl.org

> >>>>> "J" == <_@_.%> writes:
>
> J> Thank you very much that is very helpfull.
>
> J> What i ended up doing was replacing the spaces with *'s which should work out
> J> since a * is illegal in a filename (i think).
>
> no it isn't.
>
> on unix flavors, the only char not allowed in a filename is /
>
> instead of split, use a regex so you get the fields you want. and there
> is at least one module on cpan that parses out ls -l listings.
>
> uri
>
> --
> Uri Guttman ------ -------- http://www.stemsystems.com
> --Perl Consulting, Stem Development, Systems Architecture, Design and Coding-
> Search or Offer Perl Jobs ---------------------------- http://jobs.perl.org
>
Thanks! yea this is soo difficult, some ftp site use an arbitrary number of spaces
in the time format.

Uri Guttman <> wrote:
> >>>>> "J" == <_@_.%> writes:
> J> What i ended up doing was replacing the spaces with *'s which
> J> should work out since a * is illegal in a filename (i think).
>
> no it isn't.
>
> on unix flavors, the only char not allowed in a filename is /

Just for correctness: and "\000", ASCII NUL.

Ben

--
Although few may originate a policy, we are all able to judge it.
- Pericles of Athens, c.430 B.C.

In article <btn3hn$31u$>,
Ben Morrow <> wrote:

:Uri Guttman <> wrote:
:> on unix flavors, the only char not allowed in a filename is /

:Just for correctness: and "\000", ASCII NUL.

To be even more pedantic: a pathname that starts with exactly two
slashes may be interpreted in an implimentation-defined manner.
In theory, that implimentation-defined manner could treat later slashes
as filename components (or as anything else it wanted). [IEEE 1003.1
section 2.2.2.57] Just not as POSIX standard filenames, which -are-
defined to not contain slash or null [2.2.2.32].

To avoid confusion here: POSIX.1 is distinguishing between "pathnames"
and "filenames", and to POSIX.1, a "filename" is the part in-between
slashes, so it is perfectly correct to say that POSIX.1 "filenames"
cannot contain / or null -- but POSIX.1 leaves an loophole for
local non-POSIX structures in which slash might be just another
character.
--
The Knights Of The Lambda Calculus aren't dead --this is their normal form!

On Fri, 09 Jan 2004 15:03:32 +0000, $_ wrote:
> I have a string like this:
> -rwx------ 1 user group 0 Jan 06 15:07 a file with spaces
>
> and im placing that string in an array with this command:
> my @dline = split(' ', $dirline);
>
> now notice that there are three spaces between 'with spaces'
> so when i re-stringify this array using this command:
> my $filename = join(' ', @dline);
>
> $filename now contains 'a file with spaces'
> notice only one space between 'with spaces'

split() normally takes a pattern as first argument; If you use ' ' as
argument, Perl will read this as '\s+', ie. matching one or more spaces.

You should use '/\s/' as pattern, ie.:

my @dline = split(/\s/, $dirline);


--
Tore Aursand <>
"Out of missiles. Out of bullets. Down to harsh language." -- Unknown