Search Results

Not to my knowledge. shells need this detour.

Not bad an approach. Try to escape the / inside the string.

Try
tail -n+2 file.dat | tac | tail -n+2 | tac

Maybe something more like:
#!/bin/ksh
FILE_NAME_WO_EXTN="whatever"
YrDoy=`date '+%Y%j'`
filename="${FILE_NAME_WO_EXTN}_$((YrDoy - 1)).ZIP"
printf 'filename has been set to "%s"\n' "$filename"
...

I believe that the output of that script would always be:
doy
and never:
36
If you change:
echo doyto:
printf '%03d\n' $doy
you should get something closer to what you want.

how about this ?

same code of @Rudic.. but just adding exit to read only the first line on all files

for file in *.txt; do awk -F'|' '$1=="\"HD\"" {gsub(/"/,"",$3); print FILENAME, $3;exit}'...

How about awk -F'|' '$1=="\"HD\"" {gsub(/"/,"",$3); print FILENAME, $3}' *.txt | sort -k2

You'll need to experiment a bit. Try
find "$INBOUND_PATH" \! -path "$INBOUND_PATH" -prune

Thanks for the corrections MiG! I should have test the script I posted ;-)

There is a bug: -name . only matches if the start directory is .
cd "$INBOUND_PATH" && find . -type d ! -name . -prune -o -type f -size 0 \
-exec sh -c 'for i do echo mv "$i"...

Try to escape the & char with a \ char.

---------- Post updated at 12:20 ---------- Previous update was at 12:16 ----------

With an unescaped & the shell will try to put the command line...

Something similar...

if pgrep "some_pid" >/dev/null
then
echo "do some stuff"
fi


pgrep = process id grep

--ahamed

one simple thing is before reading the file, replace the variable in the file.
sed "s:\$JAVA_TOP:$JAVA_TOP:" DATA.csv > /tmp/DATA.csv

read the value from "/tmp/DATA.csv" instead of "DATA.csv"