Hadoop Shell Command Auto-completion

Usage

Similar to the Bash auto-completion you use every day, press the TAB key twice: TABTAB.

Installation without Bash Completion

• Get the script, assuming you save it to ~/bin/hadoop-completion.sh:

    curl https://raw.githubusercontent.com/guozheng/hadoop-completion/master/hadoop-completion.sh -so ~/bin/hadoop-completion.sh && chmod +x ~/bin/hadoop-completion.sh

• Add to your ~/.bash_profile (for Mac) or ~/.bashrc for (other platform) to call the script:

    if [ -f ~/hadoop-completion.sh ]; then
      . ~/hadoop-completion.sh
    fi

• Either source your ~/.bash_profile (for Mac) or ~/.bashrc for (other platform), or open a new terminal and start using hadoop-completion script.

Installation with Bash Completion

On Mac OS X

• Install Homebrew, see instructions on its website.

• Install Bash Completion using Homebrew: brew install bash-completion.

• Add to your ~/.bash_profile to use Bash Completion:

    if [ -f $(brew --prefix)/etc/bash_completion ]; then
      . $(brew --prefix)/etc/bash_completion
    fi

• Install hadoop-completion.sh to the Bash-Completion script directory /usr/local/etc/bash_completion.d.

    curl https://raw.githubusercontent.com/guozheng/hadoop-completion/master/hadoop-completion.sh -so ~/bin/hadoop-completion.sh && chmod +x ~/bin/hadoop-completion.sh

• Either source your ~/.bash_profile or open a new terminal and start using hadoop-completion script.

On other platforms

• Install Bash Completion according to the instructions on its website.

• Add hadoop-completion script to one of the following directories:

  • /etc/bash_completion.d
  • /usr/local/etc/bash_completion.d
  • ~/bash_completion.d

For example:

    curl https://raw.githubusercontent.com/guozheng/hadoop-completion/master/hadoop-completion.sh -so ~/bash_completion.d/hadoop-completion.sh && chmod +x ~/bash_completion.d/hadoop-completion.sh

• Either source your ~/.bashrc, or open a new terminal and start using hadoop-completion script.

Credits

Inspired and modified based on the hadoop completion script from Facebook/hadoop-20, which does not work with my hadoop installed using Homebrew, so I modified it and also added a couple more commands.