Matching a pattern between two characters (sed)

Shell Programming and Scripting
1

Hi there,

I have a file with lines like these:
0105:ffff0000:001b:01f4:25:434
0299:ffff0000:0009:01f4:2:319
02d2:ffff0000:000e:01f4:2:507

The above values are split up using ":" characters.
I would like capture each value, no matter what length.

Take for example the first line 0105:ffff0000:001b:01f4:25:434.
First value: 0105 (should be captured from the start of the line until the first ":")
Second value: ffff0000 (everything between the first ":" and the second ":")
Third value: 001b (everything betweed the second ":" and the third ":")
and so on..
There is always 6 values split up between 5 ":"-characters.

I tried the following for the first value:

while read line
do
  first=`echo $line | sed "s/^\(....\).*/\1/"`
  echo $first
  #second=`echo $line | sed ??`
  #third=`echo $line | sed ??`
  #fourth=`echo $line | sed ??`
  #fifth=`echo $line | sed ??`
  #sixth=`echo $line | sed ??`
done<file.txt

This will print the first four characters from the line for the $first var.
The first value is always four characters long, but the rest of the values can differ in length.
That's why I need to do the above approach by only looking between the ":"-characters.
Is it possible by only using sed?

Thanks in advance

2 
first=$(awk 'BEGIN{OFS=FS=":"}$2{print $2}' file)
echo $first

you can modify the field accordingly as well as the echo to bring you one value at a time

3

The script is going to run on a busybox-enviroment.
The awk application is not available, only sed and grep.

4

i don't know why you are using sed command.

you mention that data is already exist.
then you can use awk command like this.

first_val=`cat filename|head -1|awk -F":" '{print $1}'`
sec_val=`cat filename|head -1|awk -F":" '{print $2}'`
thrd_val=`cat filename|head -1|awk -F":" '{print $3}'`

if you want for a file then use head $variable in loop.
if you have any doubt please send a mail.

regards
rajesh

5

I'm afraid the environment doesn't ship with "awk".
So there is no way to do the same thing with only sed?
If not then I'll have to ask the software-maintainer to recompile a new build for our platform including awk.

6 
$ while read line
> do
> echo "$line" | sed 's/:/ /g' | while read a b c d e f
> do
> echo $a
> echo $b
> echo $c
> echo $d
> echo $e
> echo $f
> done
> done <  input_file
7 
while IFS=: read a b c d e f junk
do
    echo "a->[${a}]"
    .....
done < myFile
8

vgersh99,

Thanks , I was wondering there will be a short way and you showed how it is.

9

Thanks allot folks!

You really helped me out, especially the example without using sed!
Thanks a million!