Re: [gentoo-user] [OT] One line script for md5sum
On Thursday 18 December 2008 00:31:56 Robert Bridge wrote:
> On Thu, 18 Dec 2008 00:17:18 +0200
>
> Alan McKinnon wrote:
> > On Wednesday 17 December 2008 22:42:34 Robert Bridge wrote:
> > > Or use a wildcard based match.
> > >
> > > namestat.text works, as would name*stat.text
> >
> >
> > name0[01][0-9]{2}stat.text
> >
> >
> > would be better still
>
>
> more typing...
>
>
> Depends on the measure, most times I have seen this kind of thing, the
> shorter pattern I had would have been "good enough". YMMV though, and I
> won't try to argue that yours is more precise.
Agreed :-)
A pattern like that in real life is likely to be something structured and
*highly* unlikely to have odd file names thrown in.
But like all good sysadmin, I can't resist an opportunity to show off a little
bit :-)
--
alan dot mckinnon at gmail dot com
Re: [gentoo-user] [OT] One line script for md5sum
On Thu, 18 Dec 2008 00:17:18 +0200
Alan McKinnon wrote:
> On Wednesday 17 December 2008 22:42:34 Robert Bridge wrote:
>
> > Or use a wildcard based match.
> >
> > namestat.text works, as would name*stat.text
>
>
> name0[01][0-9]{2}stat.text
>
>
> would be better still
>
more typing...
Depends on the measure, most times I have seen this kind of thing, the
shorter pattern I had would have been "good enough". YMMV though, and I
won't try to argue that yours is more precise.
RobbieAB
signature.asc
Description: PGP signature
Re: [gentoo-user] [OT] One line script for md5sum
On Wednesday 17 December 2008 22:42:34 Robert Bridge wrote:
> On Wed, 17 Dec 2008 15:33:35 -0500
>
> Willie Wong wrote:
> > On Wed, Dec 17, 2008 at 08:48:52AM +, Mick wrote:
> > > On Sunday 14 December 2008, Neil Bothwick wrote:
> > > > On Sun, 14 Dec 2008 11:47:51 +0200, Alan McKinnon wrote:
> > > > > That's why I suggested them :-) I use them a lot, especially
> > > > > when I have to run the same set of commands on 15 different
> > > > > hosts, then I do something like:
> > > > >
> > > > > for I in $(seq 1 15) ; do
> > > >
> > > > If you're using bash or zsh,you can speed this up with
> > > >
> > > > for I in {1..15}; do
> > >
> > > Hmm, I tried this with a sequence of files that look like
> > > name0001stat.txt to name0198stat.txt, but when I run {0001..0198}
> > > it fails because it seems to ignore the zeros in 0001 and start
> > > counting from 1. Do I need to use some escape character for this?
> >
> > This is one place bash's brace expansion is sorely lacking compared to
> > zsh. In this case you need to use the seq command from coreutils. See
> > man seq for more info.
> >
> > In your particular case, you can do
> >
> > for I in $(seq -w 198); do ... 0$I ; done
> >
> > seq is more flexible in that it allows arbitrary formatting of the
> > sequence using printf floating-point format.
>
> Or use a wildcard based match.
>
> namestat.text works, as would name*stat.text
name0[01][0-9]{2}stat.text
would be better still
--
alan dot mckinnon at gmail dot com
Re: [gentoo-user] [OT] One line script for md5sum
On Wednesday 17 December 2008, 23:01, Neil Bothwick wrote:
> On Wed, 17 Dec 2008 08:48:52 +, Mick wrote:
> > Hmm, I tried this with a sequence of files that look like
> > name0001stat.txt to name0198stat.txt, but when I run {0001..0198} it
> > fails because it seems to ignore the zeros in 0001 and start
> > counting from 1. Do I need to use some escape character for this?
>
> No,you just need to use a better shell than bash :P
Or just use one of:
seq -f '%04g' 1 198
printf '%04d\n' {1..198}
namestat.txt (although this might match more files than wanted)
Re: [gentoo-user] [OT] One line script for md5sum
On Wed, 17 Dec 2008 08:48:52 +, Mick wrote:
> Hmm, I tried this with a sequence of files that look like
> name0001stat.txt to name0198stat.txt, but when I run {0001..0198} it
> fails because it seems to ignore the zeros in 0001 and start counting
> from 1. Do I need to use some escape character for this?
No,you just need to use a better shell than bash :P
--
Neil Bothwick
"Bother," said Pooh, as he started up DiskSalv
signature.asc
Description: PGP signature
Re: [gentoo-user] [OT] One line script for md5sum
On Wed, 17 Dec 2008 15:33:35 -0500
Willie Wong wrote:
> On Wed, Dec 17, 2008 at 08:48:52AM +, Mick wrote:
> > On Sunday 14 December 2008, Neil Bothwick wrote:
> > > On Sun, 14 Dec 2008 11:47:51 +0200, Alan McKinnon wrote:
> > > > That's why I suggested them :-) I use them a lot, especially
> > > > when I have to run the same set of commands on 15 different
> > > > hosts, then I do something like:
> > > >
> > > > for I in $(seq 1 15) ; do
> > >
> > > If you're using bash or zsh,you can speed this up with
> > >
> > > for I in {1..15}; do
> >
> > Hmm, I tried this with a sequence of files that look like
> > name0001stat.txt to name0198stat.txt, but when I run {0001..0198}
> > it fails because it seems to ignore the zeros in 0001 and start
> > counting from 1. Do I need to use some escape character for this?
>
> This is one place bash's brace expansion is sorely lacking compared to
> zsh. In this case you need to use the seq command from coreutils. See
> man seq for more info.
>
> In your particular case, you can do
>
> for I in $(seq -w 198); do ... 0$I ; done
>
> seq is more flexible in that it allows arbitrary formatting of the
> sequence using printf floating-point format.
Or use a wildcard based match.
namestat.text works, as would name*stat.text
Both are slightly less specific, but if you have other matches which
the seq excludes, you really should look at your nameing patterns.
RobbieAB
signature.asc
Description: PGP signature
Re: [gentoo-user] [OT] One line script for md5sum
On Wed, Dec 17, 2008 at 08:48:52AM +, Mick wrote:
> On Sunday 14 December 2008, Neil Bothwick wrote:
> > On Sun, 14 Dec 2008 11:47:51 +0200, Alan McKinnon wrote:
> > > That's why I suggested them :-) I use them a lot, especially when I
> > > have to run the same set of commands on 15 different hosts, then I do
> > > something like:
> > >
> > > for I in $(seq 1 15) ; do
> >
> > If you're using bash or zsh,you can speed this up with
> >
> > for I in {1..15}; do
>
> Hmm, I tried this with a sequence of files that look like name0001stat.txt to
> name0198stat.txt, but when I run {0001..0198} it fails because it seems to
> ignore the zeros in 0001 and start counting from 1. Do I need to use some
> escape character for this?
This is one place bash's brace expansion is sorely lacking compared to
zsh. In this case you need to use the seq command from coreutils. See
man seq for more info.
In your particular case, you can do
for I in $(seq -w 198); do ... 0$I ; done
seq is more flexible in that it allows arbitrary formatting of the
sequence using printf floating-point format.
W
--
Willie W. Wong ww...@math.princeton.edu
408 Fine Hall, Department of Mathematics, Princeton University, Princeton
A mathematician's reputation rests on the number of bad proofs he has given.
Re: [gentoo-user] [OT] One line script for md5sum
On Sunday 14 December 2008, Neil Bothwick wrote:
> On Sun, 14 Dec 2008 11:47:51 +0200, Alan McKinnon wrote:
> > That's why I suggested them :-) I use them a lot, especially when I
> > have to run the same set of commands on 15 different hosts, then I do
> > something like:
> >
> > for I in $(seq 1 15) ; do
>
> If you're using bash or zsh,you can speed this up with
>
> for I in {1..15}; do
Hmm, I tried this with a sequence of files that look like name0001stat.txt to
name0198stat.txt, but when I run {0001..0198} it fails because it seems to
ignore the zeros in 0001 and start counting from 1. Do I need to use some
escape character for this?
--
Regards,
Mick
signature.asc
Description: This is a digitally signed message part.
Re: [gentoo-user] [OT] One line script for md5sum
Neil Bothwick writes:
> On Sun, 14 Dec 2008 11:47:51 +0200, Alan McKinnon wrote:
> > for I in $(seq 1 15) ; do
>
> If you're using bash or zsh,you can speed this up with
>
> for I in {1..15}; do
You can even use C style: for (( i=1; i <= 15; i++ )); do
Wonko
Re: [gentoo-user] [OT] One line script for md5sum
On Sun, 14 Dec 2008 11:47:51 +0200, Alan McKinnon wrote:
> That's why I suggested them :-) I use them a lot, especially when I
> have to run the same set of commands on 15 different hosts, then I do
> something like:
>
> for I in $(seq 1 15) ; do
If you're using bash or zsh,you can speed this up with
for I in {1..15}; do
--
Neil Bothwick
Rugged: Too heavy to lift.
signature.asc
Description: PGP signature
Re: [gentoo-user] [OT] One line script for md5sum
On Sunday 14 December 2008 11:06:39 Mick wrote:
> md5sum -c token*.md5sum are the easiest on this occasion, although Alan's
> last two commands are indeed "insanely useful"! ha! They will be saving
> me hours of typing in the future. :)
That's why I suggested them :-) I use them a lot, especially when I have to
run the same set of commands on 15 different hosts, then I do something like:
for I in $(seq 1 15) ; do scp package.tar.gz a...@host${i}.domain.com ; done
for I in $(seq 1 15) ; do ssh a...@host${i}.domain.com tar xvzf
package.tar.gz ; done
You can get real cute and put the host name and commands into files (one
command/host) per line and run it all as two nested for loops:
while read -u hostfile HOST ; do while read -u commandfile COMM ; do ssh
a...@${host} ${COMM} ; done ; done
That's off the top of my head, it'll need some parenthesis to make it actually
work otherwise bash gets confused with too many ";"s but you get the idea. A
running ssh-agent in these cases is recommended too :-)
--
alan dot mckinnon at gmail dot com
Re: [gentoo-user] [OT] One line script for md5sum
On Sunday 14 December 2008, Neil Bothwick wrote:
> On Sat, 13 Dec 2008 19:59:44 -0500, Willie Wong wrote:
> > Why not just a simple bash one-liner
> >
> > for i in token{a..z}; do md5sum -c $i; done
>
> What wrong with
>
> md5sum -c token{a..z}
Thank you all for your suggestions. I think think that:
md5sum -c token{a..z}.md5sum and
md5sum -c token*.md5sum are the easiest on this occasion, although Alan's
last two commands are indeed "insanely useful"! ha! They will be saving me
hours of typing in the future. :)
--
Regards,
Mick
signature.asc
Description: This is a digitally signed message part.
Re: [gentoo-user] [OT] One line script for md5sum
On Sat, 13 Dec 2008 19:59:44 -0500, Willie Wong wrote:
> Why not just a simple bash one-liner
>
> for i in token{a..z}; do md5sum -c $i; done
What wrong with
md5sum -c token{a..z}
--
Neil Bothwick
All Scottish food is based on a dare.
signature.asc
Description: PGP signature
Re: [gentoo-user] [OT] One line script for md5sum
token* is the best. I also have this useful script in my bashrc.
its just an extended version of this.
[[ $(md5sum < file1) == $(md5sum < file2) ]]
so you can just give it two files as its argument and it returns 0 if they are
the same. -v for verbose
md5() {
local v x y;
[[ $1 == -v ]] && [[ $# == 3 ]] && { v=true; shift; }
if [[ $# != 2 ]]; then
[[ $v ]] && echo Error: need 2 arguments.
return 2
fi
if [[ $v ]]; then
x=$(md5sum < "$1" ) || return 2
y=$(md5sum < "$2" ) || return 2
else
{ x=$(md5sum < "$1" ) ;} &> /dev/null || return 2
{ y=$(md5sum < "$2" ) ;} &> /dev/null || return 2
fi
if [[ $x == $y ]]; then
[[ $v ]] && echo Same.
return 0
else
[[ $v ]] && echo Different.
return 1
fi
}
- Ian
Re: [gentoo-user] [OT] One line script for md5sum
On Sat, 13 Dec 2008 23:49:50 + Mick wrote: > Hi All, > > Is there a clever way to enter a string (rather than write a script > file) so that md5sum will check a whole series of files in one go and > report success or error; I was thinking along the lines of if $value > is ... then md5sum -c ..., but my non-existent scripting knowledge > won't take me any further. I want it to check a series of files > named tokena, tokenb, tokenc, ... , etc. purely in the spirit of coming up with the simplest one... md5sum -c token* RobbieAB signature.asc Description: PGP signature
Re: [gentoo-user] [OT] One line script for md5sum
On Sat, Dec 13, 2008 at 11:49:50PM +, Penguin Lover Mick squawked:
> Is there a clever way to enter a string (rather than write a script file) so
> that md5sum will check a whole series of files in one go and report success
> or error; I was thinking along the lines of if $value is ... then
> md5sum -c ..., but my non-existent scripting knowledge won't take me any
> further. I want it to check a series of files named tokena, tokenb,
> tokenc, ... , etc.
> --
Why not just a simple bash one-liner
for i in token{a..z}; do md5sum -c $i; done
AFAIK the md5sum will report success or failure for each of the
checksum lines in tokena, tokenb, etc.
If this doesn't suit your needs, can you describe in more detail your
desired behaviour?
W
--
(04:01:59) W: yep
(04:02:02) W: I love linux
(04:02:15) NJYWT: I love penguins
Sortir en Pantoufles: up 736 days, 23:38
Re: [gentoo-user] [OT] One line script for md5sum
On Sunday 14 December 2008 01:49:50 Mick wrote:
> Hi All,
>
> Is there a clever way to enter a string (rather than write a script file)
> so that md5sum will check a whole series of files in one go and report
> success or error; I was thinking along the lines of if $value is ... then
> md5sum -c ..., but my non-existent scripting knowledge won't take me any
> further. I want it to check a series of files named tokena, tokenb,
> tokenc, ... , etc.
You are not being very clear about what you want. When you say "check", what
does that mean? Do you want to test if the files are OK? Then you would
already have the md5sums so you should use md5sum -c
To generate the sums, just run md5sum
In any event, I think you need a fancy for loop. To run the same command on
every file in a directory:
for I in * ; do ; done
for a list of named files:
for I in file1, file2 ; do ; done
for a list of files with a clear name structure:
for I in a b c d ; do token${I} ; done
or
for I in $(seq 1 10) ; do token${I} ; done
The last two are insanely useful. Replace with whatever you need to
be run over and over
--
alan dot mckinnon at gmail dot com
[gentoo-user] [OT] One line script for md5sum
Hi All, Is there a clever way to enter a string (rather than write a script file) so that md5sum will check a whole series of files in one go and report success or error; I was thinking along the lines of if $value is ... then md5sum -c ..., but my non-existent scripting knowledge won't take me any further. I want it to check a series of files named tokena, tokenb, tokenc, ... , etc. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
