Learning GitHub Copilot

Chapter 4. Advanced Editing and Autonomous Workflows in the IDE

In Chapters 2 and 3, we covered the basics of creating code and using the chat interface to get things done with Copilot in the IDE. The items we’ve discussed will help handle basic use cases. But for more intensive coding efforts like refactoring, making significant edits, or automatically handling more complex changes, Copilot provides advanced features.

In this chapter, we’ll look at a powerful set of features that leverage the AI to handle larger sets of changes and code generation for you. These include the following:

• Predictive edits with Next Edit Suggestions
• Batch changes with Copilot Edits
• Automated end-to-end updates with Copilot Agent mode
• Producing code based on images with Copilot Vision
• Debugging with Copilot

Let’s start off looking at a feature that can predict where your next edit should be.

Predictive Edits with Next Edit Suggestions

Have you ever been trying to refactor your code in some way and found yourself searching through your file to locate all the places you need to make changes? Co­pilot’s Next Edit Suggestions (NES) is designed to eliminate the need for that and simplify the process of identifying and making the next change.

NES predicts and suggests edits based on ongoing changes in your code, typos, or logic errors. The changes it suggests can range from a single symbol to multiple lines of code. In the refactoring use case, NES can analyze the edits you’re making and locate the next ...