Introduction

For advanced users who spend a lot of time in the Code mode, CDP Studio supports GitHub Copilot - an AI coding assistant that helps you write code faster and with less effort.

Requirements

To use the Copilot plugin, you need:

• An active GitHub Copilot subscription
• The GitHub Copilot Neovim plugin, see the installation instructions below.

How It Works

The Github Copilot offers coding suggestions as you type: sometimes the completion of the current line, sometimes a whole new block of code. You can accept all or a part of a suggestion or you can ignore the suggestion and keep typing.

Useful Shortcuts

• Tab - apply a suggestion as a whole
• Ctrl+Right - apply a suggestion word-by-word
• Esc - reject a suggestion

Additional Copilot shortcuts can be configured under Edit -> Preferences... -> Environment -> Keyboard.

Installation

To use the Copilot plugin, you need an active GitHub Copilot subscription.

The CDP Studio Copilot integration depends on the GitHub Copilot Neovim plugin. Follow the Copilot.vim/Readme.md instructions to install the following dependencies:

• Node.js
• Neovim
• GitHub Copilot Neovim plugin

Finally, start CDP Studio, go to Edit -> Preferences... -> Copilot and configure the following settings:

1. Select the Enable Copilot check box to use Copilot.
2. In the Node.js path field, enter the full path to the Node.js executable.
3. In the Path to language-server.js field, enter the path to language-server.js in the Copilot Neovim plugin installation folder.
4. Select Sign In to sign into your subscription, activate your device, and authorize the GitHub Copilot plugin. When done, the button turns into a Sign Out button.

   Note: This step is not needed if you have already signed in from Neovim with the ":Copilot setup" command.

5. Select the Auto request check box to receive suggestions for the current text cursor position when you make changes.

Tips

• Copilot is very helpful with things like:
  • Adding printouts to code
  • Standard algorithms
  • Generating simple test data for unit tests
  • Writing boilerplate code
• To guide the Copilot suggestions, add comments to your code
• Sometimes it is helpful to temporarily copy some extra context into the file you are working on to help Copilot improve its suggestions - for example, a class definition or some documentation.