Shell Programming and Scripting

Hi ,

I'm so surprised to see my thread closed !

i come here many times and work with some great guys like Perderabo , each time i search for many solutions to write some scripts for my job , at this time i don't see how i can solve this one , so please don't take me as a pupil !

i can test the 7th character then cut and replace but some lines are longer than 7 character and to avoid many tests , i believe that 'AWK' can be my solution for testing the last character !

i have one file like :

APP101A
APP103B
MSG307A
MSG308B
XADM002A
and so on , and i expect to replace the last character A by B and B by A

the result would be :

APP101B
APP103A
MSG307B
MSG308A
XADM002B

how can i do that without doing about ten tests !

thanks in advance

Christian

with 120++ posts, your "newbie" license is already expired.
Also, in your previous posts you were already exposed to awk , sed and unix tools. Therefore, one would expect you to show something you have done, (its your job isn't it??), or else how do you prove you are worth the salary you are taking home?

Don't be surprised. If you "don't have time to try" to solve your issue, you really cannot expect us to have time to do it for you, can you?

Please stick with the technical aspect of your issue and show the effort first.

It seems that relying on the 7-th character replacement is not generic enough. How about dealing with the LAST character on a line?

Well i'm here because i need to modify a script while i'm doing many other admin tasks other than unix.
I'm still a newbie , i'm not an unix admin .
I'm not asking for a complete code , maybe my previous post was a little bit awkward.

"How about dealing with the LAST character on a line?"
i think it's the best way to search , and i don't see how i can do this with awk or sed.
Maybe shifting all the character on the left until the last then replace ?

.........................

i tried this with success :

sed '1,$ s/A$/B/g' oldfile > newfile

can i do in the same line the change of A to B and B to A ?
without doing :

sed '1,$ s/A$/B/g' oldfile > newfile1
sed '1,$ s/B$/A/g' oldfile > newfile2
cat newfile1 newfile2 newfile3

regards
Christian

One does not have to be a 'unix admin' to address these questions.
Once again, please stick with the technical side of your question.
If you "have time" to discuss the merits of your post(s), please do so in the non-technical forums of this site.
Sure:

Code:

sed 's/A$/B/;n;s/B$/A/' oldFile

no, you are definitely not a newbie in unix since you joined this forum 6 years ago. you are exposed to unix since then. BUT i will give you the benefit, here's a Python script, if you have it in your system

Code:

#!/usr/bin/env python
for line in open("file"):
    line=line.strip()
    if line.endswith("A"):
        print line[:-1]+"B" #print from start till the last second character.
    elif line.endswith("B"):
        print line[:-1]+"A"

the algorithm can be adapted to any other language, even awk.

Another approach with awk:

Code:

awk '
substr($0,length)=="A"{sub(/.$/,"B");print;next}
substr($0,length)=="B"{sub(/.$/,"A")}
1' file