Some Scripting Help

[RAGEAngel9]

I'm trying to create a script that will read in each item in a directory and write them to a text file without the .au extension. The problem I'm having is i forgot the proper way to read in the items one at a time.

Assuming the directory is constant, it should just be a while loop right?

[RAGEAngel9]

This is what I have so far. I know I need to use ls on the directory and then use that output in the loop but I'm not sure how to use ls and it's data.

#! /bin/ksh

IFS="\n"


direct="./whateveritscalled"



while read
do
${Data%%.au}

done < word.txt



PS- for those wondering. This technically is for a school project but it's not graded so I don't mind asking for help.

[z0mbix]

I would have thought a for loop would be more appropriate for iterating through each file in a directory.

[Rüpel]

hmm. i have no shell here (it's time to install cygwin!!!) ;)

but something like

ls Data* | awk '{ printf("%s.au\n",$1); }' > files.txt

might work.

but i'm not sure, if i got your problem right... :rolleyes:

========= edit ==========
bummer!

mixed up sed and awk again... :rolleyes:

btw: i did install cygwin and the 1-liner above does list all files from the current directory starting with Data, appends .au at each and prints them one per line into files.txt

is this want you want?

[RAGEAngel9]

That's it!?!
Geez! I really need to learn those cl commands and stuff.
Thank you very muc hfor your help.



========EDIT=======
Oh actually there is one thing. I wanted to get rid of the .au's. But that can't be that hard to change fro mthis right?

[bwkaz]

Pipe the output of that through sed -e 's/.au//g' to get rid of all occurrences of the string ".au".

Or just ls Data* | sed -e 's/.au//g' to begin with, rather than adding a .au and then deleting more than one...

[RAGEAngel9]

Ok so far I'm using this:

ls | sed -e 's/.au//g' > word.txt

and it works mainly.
1 problem and 1 thing I'd like to change is:
1. when I run it it adds word.txt to word.txt is there a way to block that?

2. How can I run this from the parent directory?

[goon12]

ls | grep -v word.txt | sed -e 's/.au//g' > word.txt


I dont have access to a shell right now, but that might do it.

[RAGEAngel9]

Ok ignore my last post.
Upon actually trying things on my own, I now have:

ls ./Words | sed -e 's/.au//g' > word.txt


Which does what I want.
The only thing remaining is I need to make it a script. What is the best way to do this? I was thinking the following:

#! /usr/sh

ls ./Words | sed -e 's/.au//g' > word.txt

But for some reason it won't run.
What's the correct way torun a script?

[dubbac]

did you make the file executable?
if so, try running it like ./scriptname.
If you aren't sure if you made it executable you probably didn't.
If you don't want anyone else to have access to the file, do this:
chmod 700 scriptname
the numbers come from the access type you want to give, 4 for read, 2 for write, 1 for execute. Add up the number, 4 + 2 + 1 = 7.
The remaining zeroes are actually how others will access the file.
So think of it like this.
Owner|Group|World
rwx|rwx|rwx
421|000|000
Result is then:
7|0|0

I think thats it in a nutshell.
Hope it isn't too unclear.

[bwkaz]

Originally posted by RAGEAngel9
#! /usr/sh

ls ./Words | sed -e 's/.au//g' > word.txt Err, that should be like this (I also made it accept the directory to search as the first command line argument -- you wouldn't have to do this, but it helps with reusability):

#!/bin/sh

ls $1 | sed -e 's/.au//g' You can then chmod 700 (or whatever) it, and do a ./scriptname >word.txt to run it. You could move the >word.txt inside the script, but that again reduces reusability. Whatever.

[RAGEAngel9]

Well I figure it's better to have it as simple as possible.
This whole thing is a service project and the script onyl really needs to be run once a year. The main program just needs the text file.