Extracting Unique Patterns with grep or ripgrep

Note#110

notes
Published

2023-07-30

Extract and sort unique instances of a pattern (ERROR:\w*) from a file:

grep -o 'ERROR:\w*' logfile | sort -u | less

Or to save to a file:

grep -o 'ERROR:\w*' logfile | sort -u > sorted_errors.txt

For ripgrep (rg), add -N to -o to prevent line numbers in output:

rg -o -N 'ERROR:\w*' logfile | sort -u | less

Breakdown

  • grep -o 'ERROR:\w*' logfile or rg -o -N 'ERROR:\w*' logfile: Extracts ERROR: followed by any number of word characters ([a-zA-Z0-9_]).
  • sort -u: Sorts and removes duplicates.
  • less or > sorted_errors.txt: Displays results or saves to file.

Replace logfile with your file and adjust the pattern as needed.