Hello all

I am trying to delete any blank lines within two patterns e.g.

Address: 53 HIGH STREET Cred Id :
MYTOWN
MYCOUNTY



MM12 6MM
Pay Method : Crossed Cheque

The start of my pattern is "Cred Id" and the end is "Pay Method" and I want to delete the blank lines between county and post code.

I did find the code below but it doesn't seem to change anything:

sed -ne '/Cred Id/,/Pay Method/!bp' -e '/^$/b' -e -e p ll.out

I can get it to print just the range I'm interested in by doing sed -ne '/Cred Id/,/Pay Method/p' so I think that bit is right but despite doing some research I'm struggling a bit to understand the rest of the code.

Can anybody help or point me in the right direction?

Many thanks
Paul

Hi,

Is this what you are looking for:

sed '/Cred Id/,/Pay Method/{/^$/d}' infile

Hope this helps.

1 members found this post helpful.

That's great. That does work thank you although I've since realised that they weren't blank lines in between the county and post code. The contains contained about 20 spaces instead!

Hi,

sed '/Cred Id/,/Pay Method/{/^[[:blank:]]*$/d}'

The above command will remove all lines that are empty, have spaces and/or tabs.

Hope this helps.

1 members found this post helpful.

I got the "command garbled" message when I tried the above.

I managed to get it working by just removing spaces and also found that the space in the patterns needed to be escaped. In case anyone's interested I got it working by:

sed -e /Cred\ Id/,/Pay\ Method/!b -e '/^[ ]*$/d'

Thank you for putting me in the right direction druuna. Much appreciated!

Paul

Hi,

The command posted in post #4 works on this side (copied command form post #4):
The infile has a mix of spaces, tabs and empty lines.

Instead of using [[:blank:]] you can also use [ <tab>] (that's a space and a tab). You can create the <tab> by pressing ctrl-v and then the tab key.

Hope this helps.