--- Ronald J Kimball <[EMAIL PROTECTED]> wrote:
> On Tue, Jan 16, 2007 at 04:59:51PM -0800, Phil Carmody wrote:
> > --- Juho Snellman <[EMAIL PROTECTED]> wrote:
> > > Phil Carmody <[EMAIL PROTECTED]> writes:
> > > > -p0 s/^((.*)(.*)
> > > > (?=\2.\3
> > > > |$))*$//
> > >
> > > Assuming the last l
In article <[EMAIL PROTECTED]>,
Phil Carmody <[EMAIL PROTECTED]> writes:
> $ echo -e "1\n12\n173\n" | perl -n0 -e 'a//^((.*)(.*
> )(?=\2.\3|$))*$/'
> Illegal division by zero at -e line 1, <> chunk 1.
Yes, it was the expected failure outcome
>
> Unfortunately it looks like it fails to ev
Phil Carmody <[EMAIL PROTECTED]> writes:
> 1//regexp/ would crash on the same thing that s/regexp// fails to blank.
> So I thought that might be the intended (not space-saving) technique with
> a rather bizarre unnecessary use of an unquoted literal to add confusion!
> But a typo seems to make more
--- Ton Hospel <[EMAIL PROTECTED]> wrote:
> -p0 s/\G(.*)(.*
> )(?=\1.\2|$)//g
>
> This should be about as efficient as a simple regex
> approach gets. It's also the shortest yet :-)
> It outputs starting from the first line that can't
> be extended. No output means all lines work.
Excellent! That
Ton Hospel wrote:
> -p0 s/\G(.*)(.* )(?=\1.\2|$)//g
>
> This should be about as efficient as a simple regex approach gets.
> It's also the shortest yet :-) It outputs starting from the first
> line that can't be extended. No output means all lines work.
If you're free to output whatever you want
In article <[EMAIL PROTECTED]>,
Phil Carmody <[EMAIL PROTECTED]> writes:
> There seemed to be a bug in previous versions (though to be honest not a
> serious one, one that could be considered as inviting a loosening of the
> specification) - if you added an extra blank line between two real
In article <[EMAIL PROTECTED]>,
Daniel Tiefnig <[EMAIL PROTECTED]> writes:
> If you're free to output whatever you want on mismatch, you can also
> just omit the "\G".
>
> -p0 s/(.*)(.*
> )(?=\1.\2|$)//g
>
> Will print all lines that preceed invalid changes. (As requested!) One
> can crea
--- Ton Hospel <[EMAIL PROTECTED]> wrote:
> By the way, what these programs don't check is if the first line is a
> single character. The challenge statement is unclear on if the first
> line should be seen as an extension of a ghost empty line or not.
Not a requirement. This lets me check slices
On Wed, Jan 17, 2007 at 01:25:20PM +0200, Juho Snellman wrote:
> Phil Carmody <[EMAIL PROTECTED]> writes:
> > 1//regexp/ would crash on the same thing that s/regexp// fails to blank.
> > So I thought that might be the intended (not space-saving) technique with
> > a rather bizarre unnecessary use o
I was just wondering, reading all your talk about your solutions to the
fonality golf, I would assume that everyone converted in arabic,
calculated and then converted back to roman. When I first had a look at
the problem I thought that that would be the obvious way to do it. At
the time I thoug
On 1/17/07, Daniel Cutter <[EMAIL PROTECTED]> wrote:
I was just wondering, reading all your talk about your solutions to the
fonality golf, I would assume that everyone converted in arabic,
calculated and then converted back to roman. When I first had a look at
the problem I thought that that wou
--- Jasper wrote:
> p.s.
> for($f=12;$f-->0;){
> surely
> for($f=12;--$f;){
> ?
p.p.s. Mr Magoo. I think you meant for($f=12;$f--;)
For cheap thrills, I fixed a few more Magoos, whittling dcutter's 247 to 237:
#!perl -aln
@a=(IV,IX,XL,XC,CD,CM,M,D,C,L,X,V);@b=(I,VI,X,LX,C,DC,DD,CC,LL,XX,VV,II);s
In article <[EMAIL PROTECTED]>,
Daniel Cutter <[EMAIL PROTECTED]> writes:
> I was just wondering, reading all your talk about your solutions to the
> fonality golf, I would assume that everyone converted in arabic,
> calculated and then converted back to roman. When I first had a look at
>
13 matches
Mail list logo
