| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
Comments:
<0> Botje: thats the idea, but there is no way to have the regex itself return a true/false?
<1> Why are you so insistent on using a regex for this?
<2> note that y/0-9// doesn't modify the string
<1> I mean, /(?:.*?\d){9}/s is hideous.
<1> s/9/10/
<3> cbrinker: \d\D*\d\D*\d\D*\d\D*... (you get the point)
<2> \d(?:\D*\d){9}
<3> yes. what mauke said.
<1> Yeah, the \D is better.
<3> except the y/// version is much much much much more readable.
<3> but then again, your language (not perl) probably doesn't have y///
<0> My hope was to have a good regex that could be (re)used in my perl/ruby/cf/javascript/etc validation routines.
<4> hahhahahaahhaha
<5> why are you asking about non-perl regex in #perl
<5> that's really being a bit dishonest
<3> write your own damn state machine.
<5> in fact, almost downright mean
<4> uh oh
<4> you've hurt merlyn's feelings
<6> does this happen daily at this time?
<5> has nothing to do with my feelings.
<4> bderrly: flaming? no, it happens round the clock
<5> has to do with common sense. :)
<5> the trouble with common sense is that it's not very common any more. :(
<4> merlyn: i consider it a compliment. everybody knows perl folk know just about everything
<2> (replicate 10 () <=) . map (const ()) . filter isDigit
<3> mauke: O_o
<2> what
<0> merlyn: the people in cf chat rooms are not very regex savvy :) and lets just say I am not surprised that someone in here had an elegant answer
<2> GumbyBRAIN: @run (replicate 10 () <=) . map (const ()) . filter isDigit $ "foo123456789bar0"
<7> Filter it at http://perldoc.perl.org/functions/map.html.
<2> wankit
<8> The Greek pronunciation of Perl 6 is Zeus.
<9> Will someone tell me how (or point me at a perldoc) to convert a string of 1s and 0s representing a binary number into that number?
<10> ParlStalker, perldoc -f pack
<8> pack. To access this perldoc please type, at a command line, 'perldoc -f pack'. You may also find it at http://perldoc.perl.org/functions/pack.html
<11> PerlStalker: print oct("0b$binary_string");
<3> PerlJam: that's cheating :P
<11> Botje: I'm sorry ... I thought this was #perl ;-)
<3> go stand in the corner and wear this dunce cap :P
<11> PerlStalker: note though that my solution is only good for numbers that have less than (on my architecture) 33 bits.
<9> That should be fine. I'm playing with netmasks.
<12> If I have a tied hash, can I bless that?
<12> It's tied in XS.
<11> postwait: you can only bless scalar references.
<12> And I want to bless it in XS as well... you think that will backfire?
<12> You can bless refs to just about anything
<3> PerlStalker: I hope you're using Net::Netmask, Net::CIDR, or likewise
<9> I figured I'd do it the hardway.
<11> postwait: what I mean to say is that you can only bless references (which are scalar thingies). You can't bless a hash.
<3> ah. reinventing the wheel for no particular reason. that's always fun ;)
<9> Exactly
<12> I have a ref to the hash.
<12> It's an RV * (noinc) of an HV * that has hv_magic of PERL_MAGIC_tied
<3> yay magic^WXS!
<11> postwait: why is it that you want to bless this tied thing?
<12> In XS.
<12> I have a tied hash...(tied to a an XS implementation)
<12> And I want to be able to invoke methods on it.
<12> So it would make sense to bless the resulting tied hash's ref into a cl*** of my choice
<12> and then implement whatever methods I need in that cl***.
<12> just not sure if it is going to have an "unexpected" interaction with the tie.
<11> postwait: try it and let us know how it goess :)
<11> postwait: I think it'll be just fine given that bless and tie are orthogonalish
<12> they are supposed to be.. yes.
<12> Will let you know.
<13> "SRCR" at 80.127.87.78 pasted "convert Snort signature .txt files to importable mysql file" (64 lines, 2.7K) at http://sial.org/pbot/17474
<14> postwait: it might be better to use the obect returned by the tie rather than the hash tied.
<14> object too.
<15> I want to get the right side of user=someone;
<15> my $left =~ s/.*?=//g; <--- This I tried
<15> But somehow it didn't work!
<11> Stormbreaker: It did work. It did exactly as you asked of it.
<3> my ($left, $right) = split /=/, $foo, 2
<15> Strange..
<15> (Meant the left side)
<14> eval: $_ = left=right; ($l, $r) = /([^=]+)=(.*)/; join ":", $l, $r
<8> cfedde: Error: Can't modify constant item in scalar ***ignment at (eval 170) line 1, at EOF
<14> eval: $_ = "left=right"; ($l, $r) = /([^=]+)=(.*)/; join ":", $l, $r
<8> cfedde: Return: left:right
<3> SRCR: what's so damn hard about it?
<15> PerlJam, but somehow it doesn't print the left side
<3> just slurp it all in one scalar
<15> my $left =~ s/.*?=//g;
<15> print($left);
<3> and split on "\n--\n"
<15> It's just acting as if nothing would be print.
<3> or set $/ to "\n--\n" and read in chunks ..
<11> Stormbreaker: did left start out as "foo=bar"?
<11> s/left/$left/
<15> PerlJam, got it! My mistake!!
<16> Does IO:Socket come with perl 5.8?
<11> And if so, your s/// is asking perl to remove the left side and keep the right side.
<3> corelist: IO::Socket
<8> This was added to Perl5.00307
<3> whiteghost: it comes with perl 5.5, even :P
<17> Botje: It's kinda difficult if you don't know any perl..
<3> SRCR: I can see that.
<11> corelist++
<3> so why are you doing it in perl?
<11> Botje: because perl is the obvious choice! ;)
<16> Botje,I just download it lol
<3> PerlJam: yes! perl is all about regexes! brilliant!
<17> Botje: I agree with PerlJam
<3> why? you don't know perl
<17> Botje: I know that much about perl and I know regexp (in scripts)
<18> PERL IS GUD
<3> okay.
<3> I told you how to do it
<3> try, and if you can't get it to work, come back.
<11> SRCR: http://learn.perl.org/library/beginning_perl
<3> setting $/ is probably the easiest way, if you've done some perl before.
<17> Botje: i have know idea what $/ is but splitting on \n--\n seems sensable
<19> PerlJam, you asked before if i understood why woggle's fix worked. I don't (I have a notion of the problem space, but would not have thought that this would work)
<19> woggle, I was pestering you folks on ##c 'cause I wanted to know the fundamentals behind a prob i'm working on in perl
<19> woggle, i need to ***ign fd 3 to a particular pipe. i can use $(open FOO, '<&=', 3) to ***ociate a particular file handle with 3, but I need to ***ociate a file *descriptor* with 3. any pointers?
<19> PerlJam, re Bummer, yeah, kinda was. i feel pretty sorry for those involved
<1> Quoting Perl code like that will probably confuse people.
<19> woggle, the $(foo) convention? borrowed from shell, but yeah, not so applicable here
<1> However, you can use open() to dup2 one filehandle to descriptor 3.
<11> ericP: in your first example you had something like for (my $i = 0; $i < $max; $i++) { ... } This uses the same $i for each iteration of the loop. woggle told you to use for my $i (0..$max-1) { ... } because it creates a new $i for each iteration of the loop. The nature of closures is to "close over" the environment when they are created, thus the first closed over the same $i which was 3 at the end of your loop.
<20> ericP: You're not making sense
<20> ericP: 3 _is_ a file descriptor, because file descriptors are numbers :)
<20> ericP: if you want to do something with it, you open a handle to it, like you were showing.
<1> Or more manually, you can use the POSIX's modules dup2().
<14> first you have to make cents, then the dollars will follow.
<19> hobbs, yeah, the prob is that the current fd is like 11, after a fork. i need to move it down to 4
<20> ericP: ... why?
<19> hobbs, in C, i guess i would dup
<1> ericP: Once you have a filehandle that refers to the file descriptor you want, you ought to be able to reopen it.
<19> hobbs, 'cause something later on knows of it as 3, and will try to open '<&=' 3 to get to it
<20> ericP: you can use POSIX::dup in Perl -- but why do you need 4? And why do you ***ume that 4 is available? :)
<1> I presume he's about to exec().
<1> In which case he needs to know about $^F
<20> ericP: try POSIX::dup2 then
<19> am indeed about to exec
<19> hobbs, tx. will look at that as soon as i answer woggle and PerlJam
<19> woggle, i had the impression that open would do it, but had no idea where to put the 3 (of course, it's not always a 3, but good for our discussion)
<20> ericP: no, open always gives you an actual filehandle. POSIX::open returns fds, but it doesn't have any of the magical dup features that are in CORE::open -- because those are in POSIX::dup and POSIX::dup2 instead ;)
<19> PerlJam, is it my imagination, or is the distinction between the lifetime of the for (my $i...) and the foreach my $i sort of arbitrary
<20> ericP: not really
<11> it's your imagination :)
<19> PerlJam, that is, i very much appreciate both the solution and the explaination, but i'm trying to understand how i'd be able to figure that out (maybe from syntax hints?)
<20> ericP: the for unrolls into "my $i = 0; { last unless $i < $max; ... } continue { $i++ }"
<19> or maybe my imagination is just trying to let me off easy
<20> er... that's sill
<20> my $i = 0; while ($i < $max) { ... } continue { $i++ }
<20> notice that $i is by necessity scoped _outside_ of that while loop
Return to
#perl
or
Go to some related
logs:
#perl
install pptpconfig fedora 5
chroot ubuntu Gtk-WARNING **: cannot open display:
jorgepena.be
#bash
#perl
#perl
#debian
ubuntu+nForce 410+ethernet
#perl
|
|