Use a different separator.
Eg.

sed -e "s@DATEST\@$DATESTART@g"

I can't figure out where the real separator is in the middle of the expression.
But you get the idea.

sed -e "s/DATEST\/08/25/2009:07:41:41/g"
One or more of the / in your dates needs escaping.

Step 1

MPUSED=`lsof | grep $MOUNT | awk '{print $2}' | sort -u`

This makes sure that each PID which might have open files is captured only once.

Step 2
Then do

for i in $MPUSED; do

Use a while loop and the 'shift' method to process each argument in turn.

Use http://software.frodo.looijaard.name/getopt/

So change the name in the find command.

Well you could read the intro guides on how to post code, perhaps format the script so it's easier to read.

As opposed to one massive 1-liner which no-one wants to look at at all.

http://www.rt.com/man/chmod.1.html

Piping works, if you can do this

open -e

/path/to/file

That is, if you just type in the "open -e", it gives you some kind of prompt for a filename.

If it doesn't do that, then piping...

http://bashdb.sourceforge.net/bashdb-man.html
By convention, a -- option turns off all further option processing, and treats the rest of the command line as parameters.

The shell will allow you to create multi-line awk programs, which in this case would allow you to only read the file once, AND be sure that the $4 you see printed is the $4 being summed.

awk...

Strange, that would have worked - post your latest code.

I would have hoped after 40 posts that you would have figured this out.
Or at least read it.
http://www.daniweb.com/forums/announcement113-3.html

Just put
ls
in there for now.

It's not magic, it's just a list of commands you want done every time you either login or create a new shell.

Copy the example dot.bash_profile to your home directory as .bash_profile

Then start uncommenting the bits you would like, and adding any new things you would like.

Yes, you're using tr wrongly.
tr just translates (or deletes) characters. Not lines, or things matching lines.

You could try
grep -v $remEmp $dataFile > temp
mv temp $dataFile

which finds...

> sed -i 's/oldInfo/newInfo/g' "$dataFile"
You need to watch your shell's quoting rules.
Things in single quotes are preserved as is
Things in double quotes allow $substitutions.

Perhaps then...

This is a partial list of GNU text utilities
http://www.rtr.com/win95pak/textutils.htm

Then there is
grep / egrep / fgrep - file searching
sed - simple file manipulation
awk - more complex...

http://www.ss64.com/bash/gawk.html

Eg.

awk -v find="Fred" 'BEGIN{FS=","; RS="\n"}
$1 == find {printf "%s, %s, %s, %s\n", $1, $2, $3, $4}' data.txt


Just replace the green code...

Nothing works by editing the file 'in place', even your vi editor.

It's all write the new file, delete the old file, then rename. It's just vi hides those steps from you. On very rare occasions...

Do you want to delete it because it's line 2, or because it's green?

Read up on the sed command.

For example
sed '2d' file > newfile
or
sed '/green/d' file > newfile

http://www.gnu.org/software/bash/manual/bashref.html#Arrays

> awk '{print $rBytes / 100000}'END
How is that anything like what I posted?

Personally, I would just do
awk '{print $2 / 1000000 }'

and stop worrying about just how good the maths is in the shell.

Also, awk has printf() as well, so you've got really good control over...

I'm confused.
Not all those lines contain "QuarkXP", but apparently, you passed the output of lsof through grep.

> lsof |grep QuarkXP|awk '{ print $9,$10,$11,$12 ]'>/tmp/output1
1. Copy the line you used, not something typed "as you remember it". The ] breaks it. I'm not sure your awk command to print a few...

Run mdadm in the forground, hence sequentially.
Run your progress in the background, then kill it when you're done.

And thanks for the explanations :icon_rolleyes:

If you ignore the "progress indicator" for the moment, does this actually work ?

# mdadm --manage /dev/md0 --add /dev/sdb1
# mdadm --manage /dev/md1 --add /dev/sdb2
# mdadm --manage /dev/md2...

> I've already got rid of the tabs with the tr command so its a lot more cleaner now,
And destroyed any sense of where all the columns are with it no doubt.

If the result is variable width...

list in my example was a file, not a command.

> I want to loop through the dir to cat the last 21 lines of each file and put all the output to another file
for i in *; do tail -21 $i; done

...

First off, which "shell" are you talking about?
There are many varieties (sh, bash, ksh, csh) and syntax varies.

Maybe
for i in `cat list`; do tail -21 $i; done

Try putting the filename variable in quotes in other places as well.

Print what you want on separate lines, then use uniq ?

uniq works on lines, not characters on the same line.

> $result="$result $next"
Remove the first $ perhaps?
Like your other assignment earlier on.

http://www.daniweb.com/forums/announcement113-2.html
Posting them is fine, so long as you show some effort.

You almost made it with
"# i tried, the output is a bit weird...."
but it would have...

Smells like homework - where's your attempt at answers?

Well what way(s) did you figure out so far, so we don't duplicate all the simple ones you know about so far.

1. Look up the > and >> output redirection (and < input redirection)

2. Most ed command can be prefixed with a count, something like say 20s/target/

Avoid csh
http://www.faqs.org/faqs/unix-faq/shell/csh-whynot/

Perhaps
cp -r /home/henry/originalDir/* /home/henry/newDirectory

> by the way, what is the difference between cp and hard link?
A hard link isn't a copy of the file, but two directory entries...