Shell Programming and Scripting

My main thing to do is to separate occurrences of 0 1 2 or 3 and when there has been a change from that value to another.
So for the case

Code:

10 bla bla 3
15 bla bla 3
71 bla bla 3
75 bla bla 3
79 bla bla 1

3 has occurred at 10, there are some duplicates in between which still show it at 3, the actual change from 3 is at 79 to 1. So I want it to be

Code:

10 bla bla 3
79 bla bla 1

and similarly for other cases as well.

Is this ok?
Thanks!

---------- Post updated at 10:50 AM ---------- Previous update was at 10:37 AM ----------

For this code, it misses the initial start of 3 which is at 10.

if I were to run for 0 1 2 and 3 together, the expected output is something like this-

Code:

1 bla bla 0
6 bla bla 2
10 bla bla 3
79 bla bla 1
83 bla bla 2
86 bla bla 3
90 bla bla 2
135 bla bla 1
187 bla bla 0

From this I think I can take out what I need.

Do you mean like this, then?

Code:

awk 'p!=$2 FS $3 FS $4 {print; p=$2 FS $3 FS $4}' infile

or

Code:

awk 'p!=$2FS$3FS$4 && p=$2FS$3FS$4' infile

These two work perfectly. I tried it on with all the other scripts. Perfect!

Thanks all for your patience and support Thanks a lot. u guys rock!