Find

Date created: 2024-02-23, Last modified: 2025-10-01

find

find is a powerful Unix command-line utility for searching and locating files and directories within a filesystem based on user-specified criteria. It traverses directory trees recursively, allowing complex queries using expressions for name, type, size, modification time, permissions, and more. It can also execute actions on matched files (e.g., delete, move, run commands), making it highly flexible for automation.

Basic syntax

find [path] [options] [expression] [actions]
  • path: Directory to start searching from (e.g., . for current directory).
  • options: Control search behavior (e.g., -maxdepth, -mindepth).
  • expression: Criteria to match files/directories (e.g., -name "*.txt", -type f).
  • actions: What to do with matches (e.g., -print, -exec command {} \;).

my usage examples

calculate the number of files in current directory

find . -maxdepth 1 -type d ! -path . -exec sh -c ‘echo -n “{}: “; find “{}” -type f wc -l’ \;

printing column values and timestamp in OpenMM output file

Purpose: see if the run would be completed in time for the number of hours the node is requested

find . -name "prod_out.txt" -exec sh -c ' dir=$(dirname "$1") xml_file="$dir/prod_initial_state.xml" echo "File: $1" if [ -f "$xml_file" ]; then mod_time=$(stat -c %y "$xml_file") echo "Modification time of XML file: $mod_time" else echo "Modification time of XML file: Not found" fi tail_line=$(tail -n 1 "$1") second_col=$(echo "$tail_line" | awk "{print \$2}") last_col=$(echo "$tail_line" | awk "{print \$NF}") echo "2nd column: $second_col, Last column: $last_col" ' sh {} \;

sort directories numerically

find . -maxdepth 1 -type d ! -name "." | sed 's|^\./||' | sort -g | sed 's|^|./|'

step-by-step explanation via

finding Directories in the current folder

find . -maxdepth 1 -type d
  • find: A command-line tool for searching files and directories.[^1]
  • .: Start the search from the current directory.
  • -maxdepth 1: Only look one level deep, so results are all directories directly under the current location.
  • -type d: Only list directories, not files.

Example output:

.
./1.000000000e+07
./1.000000000e-06
...

removing the ./ Prefix for numeric sorting

sed 's|^\./||'
  • sed: Stream editor for filtering and transforming text.
  • s|^\./||: Substitute command that removes ./ from the start of each line.
    • ^: Start of the line.
    • \./: Matches the literal ./.

Transforms output to:

1.000000000e+07
1.000000000e-06
...

numeric sorting of floating point values

sort -g
  • sort: Standard command for sorting lines.
  • -g: “General numerical sort,” which sorts lines based on their floating-point value (handles scientific notation, decimals, etc).

This puts directory names like 1.321941148e-01 before 2.009233003e+00, for example.

restoring the path prefix

sed 's|^|./|'
  • sed: Again, used to modify text.
  • s|^|./|: Adds ./ to the start of each line to restore the original directory path structure.

Restores output to:

./1.321941148e-01
./2.009233003e+00
...

excluding current directory (Optional)

find . -maxdepth 1 -type d ! -name "."
  • ! -name ".": Excludes the . (current directory) from the results.
find . -maxdepth 1 -type d ! -name "." | sed 's|^\./||' | sort -g | sed 's|^|./|'
  • Lists all directories, strips the prefix, sorts numerically, and restores the prefix for usability.