Skip to content

Latest commit

 

History

History
125 lines (87 loc) · 2.8 KB

sed and awk.md

File metadata and controls

125 lines (87 loc) · 2.8 KB

sed

gereral notes

  • add -i file_name to edit a file and save the changes and don't do that when piping into sed to edit a stream
  • don't forget to add g in the end of the sed expression if there is more than a single substitution per line

sed tricks

  • remove the first 3 lines from file:
sed '1,3d'
  • remove leading whitespaces from file:
sed 's/^[ \t]*//'
  • shorten all the words in a line to 3 letters:
sed -r 's/([a-z]{3})[a-z]*/\1/g'
  • comment out the entire file (with #):
sed 's/^\(.*\)/#\1/'
  • uncomment all comments from a file:
sed 's/^#*//'
  • remove blank lines from output
sed '/^ *$/d'
sed '/regex/{N;s/\n//;}'
  • adding text to the end of a line
#add hello to the end of all lines in a file
sed 's/$/hello/'
#add hi to the end of each line matching a regex
sed '/h+[123]/s/$/hi/'
  • add line before the last occurrence of an expression in a file
sed -nE '1h;1!H;$g;$s|(.*)(^regex$)|\1new line before the last occurrence of regex\n\2|;$p'
#    ^ #-n suppresses automatic printing of line, so only prints on p command at the end ^
#        ^^^^^^^^^ 1h;1!H;$g copy all lines to hold space, and then at the last line, copy them back to pattern space
#                  ^^ $s at the last line (meaning, after all lines were copied to pattern space), perform substition
#                     ^^^ (.*) first capture group, greedy capture of all characters, until what comes at the second capture group
#                         ^^^^^^^ (^regex$) second capture group, in this case a line containing only the word regex
# then, at the end, \1something\n\2 replaces the part that was captured, with the first capture group, then the "something", a new line character, and the second capture group
# finally, $p, prints the pattern space, since -n suppressed printing lines

awk

  • print only lines shorter than 80 notes
awk "{if (length(\$0) < 80){print \$0}}"
  • use awk to match only the command part of history list
history | awk 'match($2, /^ssh$/) {print $0}'

or

history | awk '$2~/^ssh$/ {print $0}'
  • use awk with variables from shell
awk -v var=$var '$1 == var {print $2}';
  • change the default delimeter of awk
#read csv files
awk -F',' '$1 == "something" {print $0}';
  • start printing only after a specific regex was found
awk 'BEGIN{f=0}; /regex/{f=1} {if(f) print}'
  • print from line x to line y:
awk 'BEGIN{f=0}; NR == 118 {f=1} {if(f) print} NR == 128 {f=0}' < /path/to/file
  • print column x of line y:
# print column 1 of line 4
awk 'FNR==4 {print $1}'