Use a pipe to generate multi-line input:

# print the 4th line of the reflog
git reflog | sed -n -e "4p"

sed can read a file directly as well:

# print 5th line of a file
sed -n -e "5p" my_log_beautiful_file.txt

While programming, I often search for object names across many projects. When I find the desired files, I often want to perform operations on a subset of the results. Let’s say I want to open 100 files while skipping the first 5.

Let’s start by finding all *.java and *.xml files that contain the strings foo and bar. Typically, my spell looks like this:

find . -iname '*.java' -o -iname '*.xml' | xargs grep -irl 'foo' | xargs grep -irl 'bar'

Then apply filtering:

# filter
find . -iname '*.java' -o -iname '*.xml' | xargs grep -irl 'foo' | xargs grep -irl 'bar' | sed '5,105!d'

# open files with kwrite
find . -iname '*.java' -o -iname '*.xml' | xargs grep -irl 'foo' | xargs grep -irl 'bar' | sed '5,105!d' | xargs kwrite

Meaning of the sed expression:

  • 5 - start at the 5th line
  • 105 - stop at the 105th line
  • ! - negate the address range
  • d - delete matching lines, print what is left

Similarly, if you want to keep everything except 100 lines starting from the 5th, then just remove !.

find . -iname '*.java' -o -iname '*.xml' | xargs grep -irl 'foo' | xargs grep -irl 'bar' | sed '5,105d'
https://linux.die.net/man/1/sed https://www.wyzant.com/resources/answers/646117/bash-tool-to-get-nth-line-from-a-file
⤧  Next post .tar.gz — compress & uncompress ⤧  Previous post GIT — ignore file mode changes