• xargs
• cp
• wc - Word Count
• grep
• find
• du/df - Directory Size
• Shred
• sed
• Snippets

Chain commands

$ ls -al | grep "file_name"

Use the output of one command as arguments for the next

# Find ruby files in app/ that contain the string user.
#   In the files matched, look for the string "transaction"
$ grep "users" -lr app/**/*.rb | xargs grep "transaction"

$ cp source_file target_file (or target_directory/)
     -r   # Recursive
     -i   # Interactive: ask what to do with existing files in the destination
     -v   # Verbose
     -f   # Force

Count words, lines, characters, etc... in a file

$ wc -[w | l | c] your_file.csv

Search one file

$ grep "redmine" /etc/hosts

// Search filetypes
$ grep -rn "flagged" --include=*.{rb,} .

Flags

  -r      # search recursively
  -n      # display the line number
  -l      # only output filenames of matched files
  -B 10   # display 10 lines before the line containing the match
  -A 10   # display 10 lines after the line containing the match
  -C 10   # display 10 lines before and after the line containing the match

  -i      # case insensitive
  -h      # Omit filenames in the results
  -v      # Invert, i.e. display lines that do not match

find
  -name "*.js"             # Find js files
  -wholename "controller"  # Look through the entire filename, not just extension
  -regex                   # Search with regex pattern
  -type f                  # Files
  -maxdepth X              # Go down at most X levels (1 == current directory only)
  -not
    -path                  # Exclude path
  -delete                  # Delete the files found

# Find all .js files in the app directory minus [[app/assets/.]]
$ find app/. -name "*.js" -type f -not -path "*/assets/*"

# Find files with a specific pattern and exclude hidden files
# 'find' matches the entire path, so append '.*/' to the search
$ find . -maxdepth 1 -wholename ".*/application*.js" -not -name ".*"

# Find files and move htem
$ find . -maxdepth 1 -type f -name "*.sql" | xargs -I '{}' mv '{}' dumps/

du : Display sizes of directories

du
  -s # Display an entry for each file/directory
  -h # Display the sizes in a "human readable" way
  directory_name # Specific directory
  * # Every directory in pwd

# 1: Get directories and files by size
# 2: Reverse sort (-r reverse / -h readable)
# 3: Only display the top 10 results

$ du -sh * | sort -rh | head -10

df : Display space usage at the disk level

df
  -h # Display the sizes in a "human readable" format

$ df -h

Delete in situ lines containing a given string

$ sed -i "/words_in_line/d" ./file_path.log

Replace strings

$ sed -i -- 's/https/http/g' file_path.txt

Safely delete files : rm on steroids! It overwrites the file content

shred
  -f   # Force change permissions to allow writing
  -n   # Number of iterations
  -v   # Verbose (show progress
  -u   # Actually remove file after progress
  -z   # On the last round, overwrite the file with zeros

$ shred -uvzn 3 /Documents/file_to_nuke.txt

###########################################
# == Count the number of lines for  files in a given directory and display the sorted result ==
#
# 1: Find ruby files in the current directory
# 2: Count the number of lines
# 3: Reverse sort them
# 4: Omit the "total" line
# 5: Only display the top 10 results
$ find . -name "*.rb" -type f | xargs wc -l | sort -rn | grep -v ' total$' | head -10
###########################################


###########################################
# 1: Find all files in a directory
# 2: Nuke!
$ find directory_path/ -depth -type f -exec shred -uvzn 2 {} \;
###########################################


###########################################
# == Copy a command output to the clipboard
# Linux
$ sudo apt-get install xclip
$ cat my_file.txt | xclip -selection clipboard

# OSX: pbcopy and pbpaste
$ cat ~/.ssh/id_rsa.pub | pbcopy
###########################################


###########################################
# == grep filetypes
$ grep -rn "string" --include=*.{rb,} ./
$ find . -type f -name "*.rb" | xargs grep -n "string" 
###########################################

Substitute characters in file

Change file encoding

File encoding in Unix

# Encode file to UTF-8
$ vim +"set nobomb | set fenc=utf8 | x" file_to_encode.txt

# Substitute tabs with semi-colons. On OSX, \t doesn't work. ctrl-v then hit tab
$ LC_CTYPE=C sed 's/ /;/g' infile.txt outfile.csv

# Kill all processes from a given application
$ ps aux | grep firefox | awk '{print $2}' | xargs kill -9

# Batch rename files
$ brew install rename
$ find . -type f -name "*GUIDELINES*" -exec rename 's/GUIDELINES.md/README.md/' \{\} \;

# Port already in use?
$ lsof -i :3000 | awk '{print $2}' | grep -v PID | xargs kill -9

# List all available commands
(echo -n $PATH | tr : '\0' | xargs -0 -n 1 ls; alias | sed 's/alias \([^=]*\)=.*/\1/') | sort -u
}}

{{{
# Find duplicate files
$ find app/controllers/ -type f -name "*.rb" -printf '%f\n' | sort | uniq -c | sort -rn
# find app/controllers/ -type f -name "*.rb" -printf '%f\n'
#   Find ruby files and only print their name
# uniq -c
#   Output unique names and prepend the number of occurences
# sort -rn
#   Sort them in reverse order

# Sort folders by size
$ du -cskh * | sort -rh