| |
| |
| |
|
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
Comments:
<0> CarlH: egrep -o "test" file
<1> Ok one prob
<1> I want to basically mask --- and show only the mask
<1> for example
<2> CarlH, redduck666 gave a good example. Maybe combine with the -n option and and your expression with /gp if you only want to print lines that match
<1> '^.*?([0-9]+).*?$'
<1> I dont want to print the whole line -- just the masked part
<1> so in above example, anything output would be numbers
<1> but if I tried the above, I would get the whole line
<0> no, you use -o option for that.
<1> explain
<0> man grep
<3> What distro do you people like? I did research, and took the test at http://www.zegeniestudios.net/ldc/ and I don't like any I've seen. My main requirement at this point is that it must have NTFS support because I want to dual boot with windows without changing the to FAT.
<4> CarlH: sed -ie 's/^.*?\([0-9]+\).*?$/\1/g'
<3> Oh, and it must be free.
<0> CarlH: man grep | grep '\--only-matching' -A 2
<4> CarlH: you`ll need <file> at the end of that regex.
<5> KingDillyDilly: you can always dual boot to windows without changing the FS to FAT
<5> I mean, I run a Windows dual boot... Windows running NTFS, Slackware running on reiserfs... support for NTFS read is good
<4> CarlH: and like i told you some time ago and like dimir is telling you right now, grep`s -o option is your friend :)
<5> as for captive NTFS (which gives NTFS write support on Linux)... I haven't figured it out, but at least they work in a dual boot, and that's good enough
<5> most Linux distros are free... there are very few that are not
<3> http://fedora.redhat.com/docs/release-notes/fc4/errata/ "NTFS Support is not included in the Fedora Core kernel because of patent concerns. Using a file system natively supported by Linux such as ext3 is a better option. If you must share files between Windows and Linux on a single system, use FAT/FAT32 (vfat) instead."
<1> thanks redduck666
<5> KingDillyDilly: you can easily recompile the kernel :)
<5> I wouldn't recommend fedora though
<5> that distro, unlike others, is being a real pain by taking way too much unneeded precautions and doing stuff in a weird way
<1> sed: -e expression #1, char 29: Invalid reference \1 on `s' command's RHS
<5> eg. I personally avoid it due to its unwillingness to recognize Taiwan as a country (though by standards it's not... but most other places separate it for sakes of simplicity)
<3> I wanted to choose Mandriva (I don't know Linux) but their IRC channel is a little slow and some stuff isn't free and I found it hard to determine what.
<1> redduck666: any idea why?
<5> in any case... yeah, NTFS support is in the kernel - an OS doesn't have it? recompile the kenrel
<5> KingDillyDilly: I'd recommend ubuntu for you
<5> it's basically an easy distro based on Debian :)
<5> as for free - it doesn't get much more free, they'd even ship you a CD for free
<4> CarlH: well, my sed doesn`t complain.
<1> It doesn't seem to be \1 or \\1 --- when I do sed -ie 's/Test/\1/g' file.txt
<6> back
<1> \\1 doesn't complain -- but doesn't work either :)
<5> mandrivia is mostly free... but that stuff is pure junk IMHO
<7> hmmmmmmmmmmmmmmmmmmmm, time for some peasoup!!!
<4> CarlH: errrr, s/Test/\1/g, you`ll have to create back references first.
<1> explain, ya lost me
<4> CarlH: back references would be the things between \( and \)
<1> ahhhh
<1> \( and \) rather than ()
<5> through my years of experience of different Linuxes... my three top recommendations would be: ArchLinux, Slackware, and Debian
<3> Ubuntu is kind of new and they spend too much time on language support, but I'll reconsider it.
<5> just because a distro is new doesn't mean it ****s... as for language support - that's just an extra feature that you don't have to use
<1> hmm still not working, no error now just no output
<1> sed -ie 's/\(TEST\)/\1/g' file.txt
<5> ubuntu is basically debain with some modifications to make it newbie-friendly
<0> CarlH: when you use \1, that is a reference to the string in parantheses () from the first patter.
<0> s/patter/pattern/
<0> CarlH: in the 's/Test/\1/g' there is no such thing.
<0> CarlH: try this: echo helo | sed 's/.*\(el\).*/m\1on/'
<0> rtfm damn it
<6> I used to use Ubuntu.
<0> :-)
<0> wow, lag!
<5> as for archlinux and slackware... those are really to the point, but you'll be compelled to learn to use them
<4> CarlH: what you wrote would replace 'test' with itself :)
<5> KingDillyDilly: so, what kinda distro are you looking for exactly?
<4> dimir: can you explain why quantify-ing doesn`t work:
<4> yos@yos bak $ echo "akbdfca6777aksb" | sed -e 's/.*\([0-9]+\).*/\1/g'
<4> akbdfca6777aksb
<4> yos@yos bak $ echo "akbdfca6777aksb" | sed -e 's/.*\([0-9]\).*/\1/g'
<4> 7
<8> Try escaping the +.
<1> redduck666: I am trying to output the file such that only 'Test" displays
<1> but I did create \( and \) and still not working:
<1> sed -ie 's/\(TEST\)/\1/g' file.txt
<8> CarlH: Did you try looking at the man page yet?
<9> that is weird..
<4> ignacio: nope, escaped +, outputs simply '7'
<4> CarlH: lemme repeat what i said, you command will do NOTHING.
<4> *your
<1> I get that, what I dont get is how to modify it so that it does ---
<4> it is probably issue with greedy matching :/
<8> I believe so.
<3> I'm going to buy http://www.apricorn.com/product_details.php?ID=333 for backups, and I figured I'll take advantage of the extra space and install Linux. No immediate plans for how I'll use it, but I'll probably install Perl modules that require C, which are a headache in Windows.
<5> do you function well with a command line, and are you a good learner when it comes to computer stuff?
<4> ignacio: any ideas how do i prevent it? .*\? doesn`t seem to do the trick, and neither does .*?
<3> Yes, I'm a good enough learner and patient, but I like good instructions that don't ***ume I know stuff. For a programmer, I don't know much.
<4> CarlH: grep -Eo 'regex' your_file
<8> [^0-9]*
<9> yeah i tried that
<9> echo "akbdfca6777aksb" | sed s/'^[^0-9]*\([0-9]+\)[^0-9]*$'/'\1'/g
<9> <-- didnt match at all..
<5> KingDillyDilly: then something like Arch or Slackware might be perfect for you
<10> W00t, I just broke into google.com
<10> Turns out to be a beowolf cluster.
<10> My uber 1337 skills did it again.
<10> Fear me.
<10> muahahahaha
<3> I heard good things about Slackware a couple of years ago. I'll look at it again.
<11> google.com is huge
<5> since those, while not the easiest, are easy enough (I've taught a few friends how to begin on those)... they don't try to impress you, but they function very well
<12> the moloch must die
<11> probably multiple clusters
<10> Kevin`, it's not that big. It looks big, but it's not.
<3> I heard it's hard, but of course, their website argues against it.
<12> did you deface it?
<10> No
<5> Slackware is constantly improving - main complaint I have is the amount of work you have to go through to install stuff
<10> I'm not a script kiddie.
<1> redduck666: When I do it using grep -Eo --- would I do this: grep -Eo 's/.*?(Test).*?/\1/g' file.txt ?
<5> archlinux is very similar to slackware, except you have a packaging system
<4> yos@yos bak $ echo "akbdfca6777aksb" | sed -e 's/[^0-9]*\([0-9]*\).*/\1/g'
<4> 6777
<4> yos@yos bak $ echo "akbdfca6777aksb" | sed -e 's/[^0-9]*\([0-9]+\).*/\1/g'
<4> akbdfca6777aksb
<4> can anyone explain the above behaviour?
<5> redduck666: would you please use pastebin
<4> CarlH: no, grep -Eo 'test' file
<12> redduck666: http://sitescooper.org/tao_regexps.html
<4> excelblue: maybe i should, but i`m kinda feeling stupid to pastebin 4 lines :)
<12> maybe it doesnt support all regexp stuff
<5> anything over 2 lines warrants a pastebin IMHO
<13> I'm looking for a good family distro that can run on 667Mhz Cpu and 128Mb ram.
<11> family? what do they do
<0> redduck666: * is trying to get as much as possible, and since '[0-9]*' means even zero matches, the '[^0-9]*' '.*' take everything.
<14> bulio|: I've never used it before, but maybe Debian? Fedora?
<11> word processing? ;)
<13> Internet browsing, occasinal word processing, music
<5> bulio|: what do you mean by family distro? something that a family can easily use?
<11> I like debian.. I think it would be fairly easy to learn how to USE it
<3> ..off to eat. Thanks for the help everyone
<14> excelblue: understand that its your window manager that essentially makes it family-usable, not the distro itself
<14> you can run OpenBSD
<0> or...
<14> but if you use KDE, then its all good.
<5> I'd personally go for a slim distro and configure it so others can easily use it
<11> i'd have someone smart do the updates and system management stuff
<5> KDE - not for 667MHz computers...
<13> I use debian with KDE on my PC
<14> fluxbox? fvwm? whatever you can get your family to go along with
<5> for a slower computer, I found out that most people can easily figure out the quirks of fluxbox
<13> 570Mhz cpu 380 mb ram
<13> XFCE is what I'd like for them
<0> redduck666: no, that's not it. :-)
<5> fbdesk + fluxbox - very easy... xfce is very good too
<4> dimir: but [0-9] takes excatly 1 match, even tho it is "0 or more", while + takes all the numbers even tho it is "1 or more"
<5> bulio|: what's the processor? (is it P2)
<4> errr "but [0-9]*
<13> Celeron
<4> "
<5> since if it's a Pentium2 (not VIA Cyrix, AMD K6), an i686 distro such as arch might be good
<13> I didn't like arch
<0> redduck666: there you go: echo "akbdfca6777aksb" | sed -r 's/[^0-9]*([0-9]+).*/\1/g'
<1> redduck666: If I do grep -Eo 'test' file --- then I dont acheive my goal, which is masking a string (per line) and outputting only what is \1 relative to that mask
Return to
#linux
or
Go to some related
logs:
modprobing in ubuntu
ubuntu pt800ce audio
#math
automatically synchronize remote mail locally
#suse
#css
delete mailq deffered queue postfix
xen+mandriva+veth1
#perl
ATI Rage LT Pro AGP xorg dri
|
|