Shell Programming and Scripting

Hi, I am a programming newbie, I have started learning C++ a couple of months ago. Unfortunately I don't have too much time in my hands to learn it properly. I need to write a simple script for a project I am doing but I am failing quite hard (the fact that I am learning C++ is completely unrelated btw, I am not writing the code with C ).
The script is supposed to change two words, I thought I could easily do it using sed (btw, I use Ubuntu). I tried to use a .sed file and use sed -f option. One problem is that the file name should stay the same. I couldn't figure out how to do this The other huge problem is that there are numerous files in numerous subfolders I need to change and they all have the same .sh extension. I coudn't figure out how to achieve this either. I am sorry to ask such a simple question but I really don't have too much time in my hands to really figure out a solution Help is really appreciated, thanks a lot in advance people

This sounds a lot like home work therefore I will get you started and you
can resarch the rest.

cat /tmp/xx
hello


This command will change the word hello to bye in every occurance
in file /tmp/xx

sed 's/hello/bye/' /tmp/xx

As for the sub-directories try the find command | xargs

find . -name "*.sh" -print | xargs <what ever you want to do?>

Good luck

If it was not for the scale of the task I'd wonder if it was homework.

We gather that Ubundu is the Operating System.
Which Shell are you using? echo $SHELL .
Have you read about the "-i" option to "sed" ? It should be available on your Operating System.

Please post what you have tried, what you expected to happen, and what actually happened.

If you issued this command from the top-level directory in the directory tree does it find all the relevant .sh files?

Code:

find . -type f - name '*.sh' -print

If not, what commands do list all the relevant .sh files?

Well I am going to be honest, I am actually a chemist I am trying to get into computational chemistry and this is NOT homework. I am trying to use a docking software and this is to customize a trillion .sh files for the linux cluster (I am going to write the path of the command in the .sh script).
Thanks for the advice, I have read about the -i option, I am experimenting with it but the main problem is that there is no recursive option for sed. I have found a couple of examples with xargs they didn't work either (and several more with other commands), I will try a couple more things and post results soon.
Thanks again people

---------- Post updated at 12:47 PM ---------- Previous update was at 01:21 AM ----------

Ah I figured it out I now realize this was incredibly easy but I needed more knowledge, thanks a lot for your help
What I did was I used grep as you've said to find the proper files and piped it through xargs to use sed on them. I didn't know both grep and xargs but I've read the manuals and I played around with them. Thanks again The exact code I used;

grep -rl 'string' */ | xargs sed -i 's: oldstring:newstring:'

The reason I used : s in sed was that I was using a string with a lot of / s (the path). I used */ because I am going to make a small .sh file using this and I didn't want to edit the file to set the path every single time. I am a happy man

Also I am going to play around with the find command as methyl said since I think listing .sh files is a better option in my case. But this works for now

-Edit

I have played around with find (thanks again methyl) and I think this is the best code for me;

find ./*/ -type f -name '*.sh' -print | xargs sed -i 's: oldstring : newstring :'

Works perfectly ^^ I will eventually add a backup feature (-i.bak maybe?) into this but I am done for now.

Now I need to actually do the chemistry bit of this