Bash script: ws

For the past year or so I had a little function in my .bash_profile file which made my interaction with the terminal roughly 9001% better - yeah, over 9000! This little function is called ws and makes switching between projects inside my workspace a lot easier. Both my private and work projects are inside the ~/Development directory. If I want to get to a project I would normally navigate like cd ~/Development/private/github/kevingimbel/kevingimbel.com - even with tab completion I am too lazy to type this every time I want to work on my website, so I wrote a script for it!

The script: ws

The script is actually a shell function and it looks like this.

function ws() {
  local workspace="$HOME/Development"
	# $1 is the target directory
  local target_dir="$1"
  # We start searching from the workspace dir
  cd $workspace
  # -maxdepth is the max directories to go "down"
  # head -n1 gives back the first result
  workspace_path=$(find . -maxdepth 3 -type d -name "*$target_dir*" | head -n1)
  if [  -z "$workspace_path" ]; then
    echo "Directory '$target_dir' not found. Changing to $workspace"
  else
    cd "$workspace_path"
  fi
}

This function is placed inside my ~/.bash_profile or ~/.zshrc so the function is available on every terminal start. It might look odd if your not too farmiliar with shell scripts but it is really rather simple and basic - let’s explore it line by line.

With function ws() we declare the function name. In shell scripts a function name has parentheses after it tho it does not take arguments this way (If you know why this is, open an issue and explain it please or tweet me!). The next lines declare a local variable, workspace and then target_dir=$1 were I assign whatever is passed to the function as being the target directory the user is trying to find. Then, I execute cd $workspace to go to the workspace directory and start a search with workspace_path=$(find -maxdepth 3 -type d -name "*$target_dir*" | head -n1) - now that might look weird. What happens here is the following:

• We declare a variable, workspace_path
• We execute a function with $()
• We use find . -maxdepth 3 -type d -name "*$target_dor*" to search for a directory with a name similar to $target_dir (the variable from earlier)
• We then pipe (|) the output from find to head and read in the first line -n1 - this is the first match
• The return value from $(find ... | head -n1) is then available in the $workspace_path variable

$workspace_path now holds whatever directory was found - or nothing. This is why we check if it is empty (-z "$workspace_path") in the if block.

if [  -z "$workspace_path" ]; then
  echo "Directory '$target_dir' not found. Changing to $workspace"
else
  cd "$workspace_path"
fi

This reads: If $workspace_path is empty, then echo the directory was not found and notify the user we changed to $workspace with the cd $workspace earlier. Else, cd into the found directory. And that’s it! Some examples can be seen in the screenshot below.

I am sure the ws function can be improved more but it has served my needs well so far. One optimisation might be to use the second argument, $2, as parameter for -maxdepth.