Thanks for your input MizzGail, but I'm afraid oombera is right... I'm not sure which script you refer to when you ask if it changes when it changes from one location to another... The idea is that the script stays the same and the only input the user should have, are the new pathnames in the config (.ini) files, which are entered when oombera's script asks for them. 
Oombera, I tried your latest script and seems to have wiped the ini files and now they only contain LOC=/u/[DIR]/main... hehehe
Oops!! That's no good!
In the last script, the one line says echo $LINE >> $TMP_00 ... that should be TMP_00 without the dollar sign ..
lol! 
well it has worked for the most part! except for:
rm: TMP_00 non-existent
rm: TMP_00 non-existent
rm: TMP_00 non-existent
rm: TMP_00 non-existent
rm: TMP_00 non-existent
rm: TMP_00 non-existent
rm: TMP_00 non-existent
rm: TMP_00 non-existent
[project4] #
and it erases everything else on the line too... 
but other than that, it rocks
oombera, you are using commands of the form:
sed 's/this/that/'
and you have problems if the "this" ot "that" actually is a string that contains slashes. The sed s command does not require slashes. You can pick another character. Like this:
sed 's=this=that='
So just pick a delimiter character that is not used in the strings and now you have no backslash problem.
OMG! That's great Perderabo - I didn't know that! Thanks 
So if you wanna just request one directory, you can just use this Ypnos:
echo -n "Enter a directory: "
read DIR
for FILE in `ls`
do
sed -e 's=/u/.*/main/=/u/'$DIR'/main/=g' < $FILE > TMP_00
mv TMP_00 $FILE
done
Now the user can enter any symbol except the equal sign, which should be fine..
ah ha!! it works a treat! (except for erasing the rest of the line)
thank you loads, oombera and everyone else
aaight
if I take the star out, so it looks like:
sed -e 's=/u/.=/u/'$DIR'=g' < $FILE > TMP_00
then it no longer erases the rest of the line, however it also inserts the user defined directory between the / and the pre-existing directory, instead of over-writing the existing....
tried a couple of . and * combos to no avail...
Do you mean there's other stuff on the line before and/or after the directory path? Even with other stuff on the line, that script should still work..
Can you post the whole line that contains the path name?
Try inserting */ after the . (<= period)
I think you're omitting the forward slash .. that seems to hack off the rest of the line..
sDataPath=/u/ttc/data:/u/ttc/ptm:/u/ttc/pfx
sTempPath=/u/ttc/data
Those are the lines that need changing... I tried what you said and now it doesn't seem to change the files at all...
Is ttc literally what you have in the file? Cause then you could just replace ttc wherever it occurs with the directory:
...
...
sed -e 's=ttc='$DIR'=g' < $FILE > TMP_00
mv TMP_00 $FILE
...
...
Nope, ttc is just one of 150+ different names there can be. Sometimes it's ttc, sometimes it's sky, mor, hai, and so on...
this is turning out to be the never ending script 
small line would do the job
cd dir-of-configs
perl -pi -e "s:/your/first/dir:/your/second/dir:g" *
or perl -pi -e "s:\/your\/first\/dir:\/your\/second\/dir:g"
not sure if it works wothout scaping the slashes
Arg .. this is the script that never ends.. it goes on and on my friends...
Okay.. let's see if this works:
echo -n "Enter a directory: "
read DIR
for FILE in `ls`
do
sed -e 's=/u/.../=/u/'$DIR'/=g' < $FILE > TMP_00
mv TMP_00 $FILE
done
rm TMP_00
Heh, when we started this thread, you said you'd been using Unix a couple days.. now it's been a couple weeks. How do you like it so far?
Quirky but amusing, oombera
Quite enjoying it actually
Hello again
Here's the deal, I have a file and it is of the following format:
HEAD SOME DATA 123456789
LINE DATA DATA DATA
LINE DATA DATA DATA
LINE DATA DATA DATA
HEAD SOME DATA 999999999
LINE DATA DATA DATA
LINE DATA DATA DATA
LINE DATA DATA DATA
HEAD SOME DATA 555555555
LINE DATA DATA DATA
LINE DATA DATA DATA
LINE DATA DATA DATA
Is there a script that will edit a file (which will be generated every night), take the number following 'HEAD' (see example) and add it to the start of every line which begins with the word LINE to make it look like this?
HEAD SOME DATA 123456789
123456789 LINE DATA DATA DATA
123456789 LINE DATA DATA DATA
123456789 LINE DATA DATA DATA
HEAD SOME DATA 999999999
999999999 LINE DATA DATA DATA
999999999 LINE DATA DATA DATA
999999999 LINE DATA DATA DATA
HEAD SOME DATA 555555555
555555555 LINE DATA DATA DATA
555555555 LINE DATA DATA DATA
555555555 LINE DATA DATA DATA
Thank you kindly
PS Everything in this file is fixed length and there are ALOT more fields that the example I use.
If you have more fields, you'll have to change $4 to reflect the field that contains the number in it.
number=`head -n 1 text3 | awk '{print $4}'`
while read LINE
do
echo "$LINE" | grep -q "^LINE"
if [ $? -eq 0 ]; then
echo "$number $LINE"
else
echo "$LINE"
number=`echo "$LINE" | awk '{print $4}'`
fi
done < yourFile
hmm...
[project4] # ./import
head: cannot open text3: No such file or directory (error 2)
is it a problem that LINE is also a shell command?
I sorted out that error, but it doesn't seem to have written anything to the file
That's because the code doesn't tell it to write to a file.. this does:
number=`head -n 1 yourFile | awk '{print $4}'`
while read LINE
do
echo "$LINE" | grep -q "^LINE"
if [ $? -eq 0 ]; then
echo "$number $LINE" >> resultFile
else
echo "$LINE" >> resultFile
number=`echo "$LINE" | awk '{print $4}'`
fi
done < yourFile
mv resultFile yourFile
omg oombera!!! you completely and utterly ROCK! 
Thank you thank you thank you!!!!