I have == Changelog == in a textfile and I want to create a new file where this line and everything after it will be deleted.

I am sure the answer is here somewhere, I have tried to search, now my head is spinning. sed, awk, csplit, ... ?

  • 2
    awk '/Changelog/{exit}1' file > newfile
    – jasonwryan
    Mar 15, 2015 at 6:34
  • Thanks. I better use '/== Changelog/{exit}1' or I get the wrong line by accident or not? Mar 15, 2015 at 7:23
  • That all depends on what else is in your input file... Does Changelog appear on another line?
    – jasonwryan
    Mar 15, 2015 at 7:25
  • Maybe, maybe later so I will be specific in the first place. Mar 15, 2015 at 8:07

3 Answers 3


With awk:

awk '/Changelog/{exit}1' file > newfile

or with sed:

sed -n '/Changelog/q;p' file > newfile
  • Doesn't this answer ignore the == signs?
    – o0'.
    Mar 15, 2015 at 11:09
  • @Lohoris Yes, see the comments under the question...
    – jasonwryan
    Mar 15, 2015 at 16:11

I would use sed:

sed '/== Changelog ==/,$d' file > newfile

This syntax is less efficient than the one proposed above by jasonwryan, but it is more readable by most people. There are two famous use for sed: the basic search and replace s/// and range based action /beginpattern/,/endpattern/X where X is a sed action like delete, search-replace... Any construction with a semicolon (;) is more programmatic and less understood.

  • Accepted for putting the == in there. And because sed ,$d is quite easy to remember for me. Its actually almost impossible that there is another capital case Changelog in there but still I like it this way more. But for to include just the == before the word so I can forget the == after the words. I did that once. Its about wordpress.org/plugins/about/readme.txt btw. In this annoying light markdown hears closing are actually not optional like in real markdown. Mar 18, 2015 at 10:04
csplit -f newfile -n1 infile /==\ Changelog\ ==/ 

That will create ./newfile0 and ./newfile1 (w/ GNU csplit, or POSIXly ./newfile[12]). The first will contain all lines up to the first occurring line that matches == Changelog == and the second will contain all lines following that.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .