Search Results

A shell script with only shell builtins:
#!/bin/sh
set -f # unquoted $f2
IFS=";"
while IFS="," read f1 f2
do
printf "$f1, %s\n" $f2
done < file

That globally suppresses all error messages.
You can suppress them for specific commands only, for example
df -kP 2> /dev/null
Suppress the expected.
If unexpected things happen then error...

Here is one with awk:
awk '
{ blevel+=gsub("[{]","&")-gsub("[}]","&") }
blevel<dlevel { del=0 }
del==0
$1~/^type=scythe/ { del=1; dlevel=blevel }
' goob

When it meets type=scythe it deletes...

Even without the shopt -s extglob the folllowing is expanded: echo {*,*/*,*/*/*}.{opt,texi}
A shopt -s nullglob suppresses non-matching patterns.

grep -l prints the file names so you can process them with xargs.
grep --include="*.org" --include="*.texi" -lir "Abstract" . | xargs headBut xargs has a problem with space characters in file names....

Yes 97 works, but not 10.
The problem is that the $( ) (likewise the ` `) strips a trailing newline; this is mostly for convenience.
x="
"; echo ${#x}
1
x=$(echo "
"); echo ${#x}
0


---...

I forgot that ( is special in ERE, that awk uses. You found it: needs to be escaped. An alternative would be [(].

Each code line is a condition { action }.
All three code lines are run for each...

When read indicates an EOF and the loop ends, it still has read the last incomplete line into the variable.
So you can process it after the loop.
do_dirn(){
echo "dirn :$dirn"
}
printf...

Like the following shell command with embedded awk script?
awk '
$2 != "=" { next }
$1 == "(SERVICE_NAME" && $3 ~ s { print a }
$1 !~ /^(/ { a=$1 }
' s="pri"

I am not an AIX expert, but NFS is a standard - I guess IBM even licenced it from SUN.
The NFS version is "negotiated" once, at mount time. The client simply asks the server via rpcinfo what...

Thanks for the update.
These hardware RAIDs ship with their own software. I wonder if Solaris provides a general frontend where they can plug in. (Like fbconfig, it was a frontend command for most...

The
ls -lrth $SOURCEONEACTUAL | awk '{print $9}'
can be simplified and improved (in case of a space in the file name):
ls -rt $SOURCEONEACTUAL

The
cd $SOURCEONEACTUAL
you better do once -...

You can try to embed your remote script in a here document, then for faster processing store it in a variable.
scriptcode=$(
cat << "ENDSCRIPT"
echo "running on `uname -n`"
ps -fp `grep \`tail -1...

... that works in ksh93 and newer.

ksh88 needs
b=$(printf "%.f" $a)

awk can do this during reading the input file - no need to store everything and process in the END section.
awk '$1~search {p=1} NF==0 {p=0}; p' search="NUMADDR" fileThe p variable controls when to...

Each character set [ ] matches one character
find ~/ -name 'bin.000[0-9][0-9][0-9]'

crontab -l list.
crontab -e edit with default editor.
EDITOR=nano crontab -e edit with nano.
(crontab -l; echo '2 18 * * 0,2,4,6 find /tempr/DSC/PH[123]* \! -type d -mtime +3 -exec rm -f {} +') |...

Run on "even" weekdays at 18:02
2 18 * * 0,2,4,6 find /tempr/DSC/PH[123]* \! -type d -mtime +3 -exec rm -f {} \;

1. Could you please clarify?
injest => ingest
normalize => sort ?

2. Could you provide a short example of your input data?

3. Could you provide your script trial?
If you run it on the...

That explains it; and your photo shows a non-ascii character.
It's not sufficient to sort out [a-zA-Z].

With an ERE you can ensure the input is a number:
[[ $NUM =~ ^[0-9]+$ ]] || continue
echo...

&2>/dev/null is still odd, the & is too many. The & before a descriptor number is needed on the RHS only, to distinguish it from a file name.
And, the only errors that a [[ ]] can produce are...

With two capture groups and back-references one can insert a newline in the middle:
sed 's/\(.\)\(.:\)/\1\
\2/g'
Here I used the standard backslash-newline to represent the newline.

Consequently you should replace the other grep -E -q with a [[ =~ ]].

And &2>/dev/zero looks odd. Perhaps you mean 2>/dev/null but this is not appropriate after a [[ ]] compound.

If it arranged like this, two fields per line, then you can sort by field 2 numerically
sort -k2n records. txt
Pipe this to a tool that displays the first (head-1) or the last line (tail-1).

Check the process state (S column) with
ps -flu xpinvoice
Are they in D state?