sed

Search

grep is likely better

$ sed -n "/pattern/p" file.txt

$ cat test.txt 
asdfasdf
fjfjfjfjfj
aaaaa

$ sed -n "/a/p" test.txt 
asdfasdf
aaaaa

Find and Replate

g is for global or replace all

$ sed -i -e "s/pattern_to_find/replace_string/" file.txt

$ sed -i -e "s/pattern_to_find/replace_string/g" file.txt

Flags:

-i[SUFFIX], --in-place[=SUFFIX]

              edit files in place (makes backup if SUFFIX supplied)

-e script, --expression=script

              add the script to the commands to be executed

Ex:

$ cat deleteme.txt 
abc

abc

abc
$ sed -i -e "s/abc/XYZ/g" deleteme.txt
$ cat deleteme.txt 
XYZ

XYZ

XYZ

Find Line

Source

$ cat test.txt 
asdfasdf
fjfjfjfjfj
aaaaa

$ sed -n "1p" test.txt 
asdfasdf
$ sed -n "2p" test.txt 
fjfjfjfjfj
$ sed -n "1,2p" test.txt 
asdfasdf
fjfjfjfjfj
$ sed -n "1~2p" test.txt 
asdfasdf
aaaaa
$ sed -n "2{p;q}" test.txt 
fjfjfjfjfj
$ sed -n "1,2{p;q}" test.txt 
asdfasdf

* -n, --quiet, --silent
              suppress automatic printing of pattern space

* “-n ‘5p’” means print only the fifth line

* The sed has provided a ‘q‘ command that allows to “quit” further processing. We can put the ‘q‘ command in the two one-liners

* {p;q} is just to apply the p and the q command