Why

The PATH environment variable does not get updated after running the install.sh script,
As a result the user would have to manually source the shell configuration file in order to update the environment variables or open a new terminal session.

Even though the below conditional statement to source the configuration file is present in install.sh at the end of file

if [ -f "$HOME/.bashrc" ]; then
	source "$HOME/.bashrc"
fi
if [ -f "$HOME/.zshrc" ]; then
	zsh
	source "$HOME/.zshrc"
fi

But it won't work (in some cases it will) because shell scripts by default run in a non interactive shell and source command only works when the shell is in interactive mode. Which means the environment variables declared or updated while running the install.sh file won't be reflected in the user's shell.

What is changing

A case statement has been added below the previous command which was used to run install.sh so that the configuration can be sourced in the interactive shell and the environment variables are updated in the same shell instance.

case "$SHELL" in
    *bash*)
        if [ -f "$HOME/.bashrc" ]; then
            # if the shell contains "bash" and .bashrc is present
            source $HOME/.bashrc
        fi
        ;;
    *zsh*)
        if [ -f "$HOME/.zshrc" ]; then
            # if the shell contains "zsh" and .zshrc is present
            source $HOME/.zshrc
        fi
        ;;
    *)
        echo "Unknown shell or required configuration file not found."
        ;;
esac

Why did I use this approach to check shell type?

You might notice I used "$SHELL" in *bash* approach which basically is checking a sub string instead of explicitly doing $SHELL=="/bin/bash"
This approach have some added advantages to it instead of explicit check of shell path'

• It would work even if the user shell is /usr/bin/bash or /usr/bin/zsh
• It would work if the user has set it's customized the shell path and not using the traditional /bin/bash or /usr/bin/bash

How to test

To test you will first need to uninstall the existing install (if present) using
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/AssemblyAI/assemblyai-cli/main/uninstall.sh)" and load the latest shell configuration using the above command we used before.

Verify that the assemblyai is not within the PATH entering it on terminal

Now Install the script again using /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/AssemblyAI/assemblyai-cli/main/install.sh)" and run the command for loading the shell configuration we used before.

As you can see below it updated the environment variable and now the user can access the assemblyai