How to remove blank lines without using tr?

Question

vim-cmd vmsvc/getallvms prints like this:

id  virtual machine name
11  VirtualMachine1
15  Virtual Machine 1
6   Virtu al Machin e 1


9   Virtual Machine one
21  VirtualMa chineone

I get the ids using the following:

vm_ids=$(vim-cmd vmsvc/getallvms | awk 'NR>1 {print $1}' | egrep "[0-9]+")

And next I want to loop through them and execute a command against each one so I do this:

i=2
for id in $vm_ids
do
   vm_name=$(vim-cmd vmsvc/getallvms | awk -v i=$i -F "[[:space:]]{2,}+" 'NR==i {print $2}' | egrep "$[[:alnum:]]")
   echo "Doing something to ID: $id for machine: $vm_name"
   $((i++))
done

My script is returning a null value when it gets to those empty lines.

I made the virtual machine names like this just to demonstrate the variety of ways that a name can be returned. Since ESXi doesn't have tr installed, how can I remove those lines from my output, or is there a better way of doing this?

Answer 1

in vim, execute this command:

:v/./d

EDIT

with grep: grep -o '^[0-9]\+' should work without worrying about empty lines. It gives all the leading numbers (ids).

with your command:

vm_ids=$(vim-cmd vmsvc/getallvms|grep -o '^[0-9]\+')

Answer 2

Using sed:

sed '/^$/d' inputfile

Since you're already piping the output to awk, a better approach would be to replace your command with:

vm_ids=$(vim-cmd vmsvc/getallvms | awk 'NF && NR>1 {print $1}' | egrep "[0-9]+")

Answer 3

With grep:

grep . file

With awk:

awk 'NF' file

---

However you can replace your command with a single call to awk:

vm_ids=$(vim-cmd vmsvc/getallvms | awk '/^[0-9]+/{print $1}')