Fix sluggish ZSH shells with lazy loading for slow scripts

I care immensely for having a snappy terminal startup and rebel against demands to blindly source scripts in my ~/.zshrc.

The most recent challenge was Node Version Manager. NVM commits the mortal sin of automatically adding itself to the ~/.zshrc file in the following incantation:

URL='https://raw.githubusercontent.com/creationix/nvm/v0.34.0/install.sh'
curl -o- "${URL}" | bash

# Added to ~/.zshrc.
[ -s "$NVM_DIR/nvm.sh" ] && \. "$NVM_DIR/nvm.sh"

The NVM script requires 800ms to load, a debt paid every time I startup a shell. Consulting internet canon, we calculate how much time it’s worth spending to reclaim the 800ms of shell startup.

xkcd-1205: Is it worth the time? How long can you work on making a routine task more efficient before you're spending more time than you save? (across five years).

Assuming I open 20 shells a day, and it costs 800ms each time, we can approximate from the table entry of 1 second, 50⁄day for a time of 1 day. Then, scale down by 20⁄50 = 0.4 for shells per day and by 800⁄1000 = 0.8 for the 800ms startup. xkcd allows us 0.32 days or about 8 hours to spend fixing this transgression.

The solution is to lazy-load an nvm function that:

1. Unloads the fake nvm function.
2. Sources the slow NVM script to expose the real functionality.
3. Runs the real NVM.

# Shim NVM that will load the real NVM.
# Saves 800ms of script start-up time.
function nvm() {
  if [[ ! -e "${NVM_DIR}/nvm.sh" ]]; then
    echo "NVM not found in NVM_DIR=${NVM_DIR}" >&2
    return 1
  fi

  unfunction nvm
  source "${NVM_DIR}/nvm.sh"
  nvm "$@"
}