GitHub Copilot tips and tricks

now that you've been granted access to GitHub copilot let's explore how to get the most out of the tool the first step is to install the appropriate extension or Plugin the exact steps for this will vary depending on the IDE that you're using but in the case of Visual Studio code you'll go to the extensions do a search for copilot and then select install copilot makes suggestions based on the context that it sees this includes the files that you've opened and interacted with the packages and libraries that have been installed the code that you're writing and comments that you add if you notice that copilot is giving you relatively generic or vague suggestions this is typically because it just needs a little bit more context you can use a comment for example to help point it in the right direction and describe exactly what it is that you're looking for copy that really shines when it comes to dealing with repetitive or boilerplate code when we think for example about creating a data model this is relatively repetitive and typically you would find yourself copying and pasting from somewhere else with copilot I can describe with a comment what I need and then get a suggestion for a class that matches the description that I provided always remember that the suggestions from copilot are exactly that they're suggestions this means that even if you accept them you're allowed to modify them to ensure that they are set up and going to run the way that you expect Additionally you want to run any code be it code that you've created or code that copilot has suggested through the normal processes that you would with any code and this includes running it through security checks code quality linters Etc as a developer there's a lot of different Frameworks and bits of syntax that well we don't always quite remember and with copilot I can describe my scenario and get a suggestion so let's take a regular expression I can describe a natural language what it is that I'm looking to validate and copilot will then make a suggestion for me and this allows me to avoid having to leave to go look up what that syntax might be do a search or otherwise I can stay in the zone and in my IDE depending on the scenario copilot will provide multiple suggestions you can cycle through these either by utilizing keyboard shortcuts or the UI and then accept the one that most closely matches what it is that you're trying to do copilot is at its best when working with languages and Frameworks where there's a large publicly available code base you may find that the suggestions aren't necessarily to the Quality that you might want for languages that are more obscure or where that level of a publicly available code base doesn't exist using copilot does not mean that you need to completely change the way that you code you can use comments to describe your scenario and get suggestions and you can also just keep writing code like you normally would that as you add in functions as you add in methods as you add in classes as you write code copilot will be making suggestions and you will find that after a little while after a little bit of practice that you'll be able to interact very naturally with GitHub co-pilot your AI pair programmer

Channel: GitHub

Views: 40,434

Keywords: security, github, devops, owasp, codeql

Id: 1qs6QKk0DVc

Length: 3min 58sec (238 seconds)

Published: Fri Apr 21 2023