Deleimeter script 

Newsgroups:  comp.unix.shell
Date:        Thu, 12 Oct 2006 07:39:52 -0500
>   Is there any pattern matching command in linux so that I can extract
> couple of line based on starting and end delimeter.
>
> Eg:
>  I want to extract the 2 lines based on the starting delemeter '(' and
> ending delemeter ')' .
sed -n '/(/,/)/p' file
awk '/\(/,/\)/' file

Ed Morton

Pick 8-bit ascii using awk 

Newsgroups: comp.unix.shell
Date: Sun, 03 Jun 2007
 > Is there any way to pick those >x7f ascii chars using awk?
 >
 > I've been trying with
 >
 >  awk '/[\200-\377]/{print}' file
 >  awk '/[^\40-\177]/{print}' file
 >  awk '/[^\040-\0177]/{print}' file
 >
 > but neither seems to be working.
[...]

Try

LC_ALL=C command -p awk '/[\200-\377]/'

or

eval "$(
  printf %b 'min=\0200 max=\0377'
)"
LC_ALL=C grep "[$min-$max]"

or

LC_ALL=C perl -ne 'print if /[\200-\377]/'

Stephane CHAZELAS

Pick 8-bit ascii using awk 

>> Is there any way to pick those >x7f ascii chars using awk?
>>
>> I've been trying with [...] but neither seems to be working.
> [...]
>
> Try
>
> eval "$(
>   printf %b 'min=\0200 max=\0377'
> )"
> LC_ALL=C grep "[$min-$max]"
>
> or
>
> LC_ALL=C perl -ne 'print if /[\200-\377]/'

Thanks for the comprehensive reply, Stephane.

Yep, all above works fine.

> LC_ALL=C command -p awk '/[\200-\377]/'

but except this one:

$ LC_ALL=C command -p awk '/[\200-\377]/' test-file 2>&1 | tee /dev/tty | od -t x1
awk: fatal: Invalid collation character: ...
0000000 61 77 6b 3a 20 66 61 74 61 6c 3a 20 49 6e 76 61
0000020 6c 69 64 20 63 6f 6c 6c 61 74 69 6f 6e 20 63 68
0000040 61 72 61 63 74 65 72 3a 20 2f 5b 80 2d ff 5d 2f
0000060 0a
$ awk --version
GNU Awk 3.1.5

PS. lisf of all awk related packages installed in my Debian Etch:

gawk_3.1.5.dfsg-4
mawk_1.3.3-11

thanks

tong