Remove some text with shell scripts?

Question

I wanted to get some help on how I can use cygwin, which has bash to shell script the following:

I have a txt file that contains the following:

action "action1"
  reset
  type xformbin
  http-method-limited POST
  http-method-limited2 POST
exit

action "action2"
  reset
  admin-state disabled
  type results
  http-method-limited POST
  http-method-limited2 POST
exit

action "action3"
  reset
  admin-state disabled
  type setvar
  http-method-limited POST
  http-method-limited2 POST
exit

I was hoping a shell script could be written to remove the block where admin-state = disabled?

So, I'm hoping I can iterate through the txt file and if admin-state = disabled, remove everything between "action" and "exit" from that particular block.

I would expect the following final results from the sample text:

action "action1"
  reset
  type xformbin
  http-method-limited POST
  http-method-limited2 POST
exit

Thank you.

Answer 1

So you want the whole block ignored if it has 'disabled' in the middle, but printed if it does not.

sed -n '
  /action/,/exit/ {
  /action/ { x; d; }
  H;
  /exit/ { x;
    /disabled/ d;
    p;         d;
  }
}' x

This will do nothing unless in a block from action to exit. In those -

If a line has action, store it and delete the pattern space to trigger the next read.

Otherwise append the line to the stored hold space.

If the line had exit,

• swap the hold space into the pattern space
• if the collected pattern has disabled, delete it to trigger reading the next record;
• if not, print it, then delete it to trigger reading the next record.

The output:

$: sed -n '
  /action/,/exit/ {
    /action/ { x; d; }
    H;
    /exit/ { x;
      /disabled/ d;
      p;         d;
    }
  }' infile
action "action2"
reset
admin-state enabled
type xform
http-method GET
http-method-limited POST
http-method-limited2 POST exit

Hope that helps.