As an old user who often uses IntelliJ IDEA to write code, I pay close attention to every update of JetBrains software.
This is not the most recent time, JetBrains has issued multiple EAP versions of software:
At the same time, JetBrains' official blog also announced an important new feature, that is:
An AI assistant called [AI Assistant] will be integrated in all JetBrains IDEs and .NET Tools.
In other words, whether it is JetBrains IDE like IDEA, Clion, PyCharm:
Or .NET tools like ReSharper, Rider, dotPeek, etc.:
All will integrate the new feature [AI Assistant].
As we all know, in recent times, generative AI and large language models are rapidly changing the landscape of software development tools, so JetBrains also said that it will take it for granted that these AI functions will be integrated and integrated into its own products.
According to the official introduction, these AI functions will be supported by JetBrains AI service modules.
The service is transparent to users and can connect users to different large language models and enable specific AI functions in JetBrains software products. The service also supports OpenAI, including hosting many of the smaller models created by JetBrains.
In the future, JetBrains also plans to cooperate with more AI model suppliers, so that users can obtain the best options and models according to their needs. In addition, the official also plans to support local (including locally deployable) models in the future to expand usage scenarios and functions.
So let's take a look at some of the changes brought about by this new feature in combination with the official examples.
AI dialogue
Users can directly use the AI Assistant tool window to conduct large-scale dialogues, such as directly asking questions or some question-and-answer dialogues based on the current task.
At this time, the IDE will give some results based on the project context. If the user accepts it, the code generated based on AI can be directly inserted into the current content, which is very convenient.
code understanding
In addition to generating content, the newly added AI Assistant also supports code understanding and content suggestions.
The user can directly select a specific code in the current IDE, and then invoke the relevant AI auxiliary functions by calling out the [AI Actions] menu, typically such as:
- code snippet understanding
- Code Refactoring Suggestions
- Potential problem discovery
I have to say that this is very helpful for improving the efficiency in the process of writing code or reading code.
document generation
The newly added AI assistant also supports automatic generation of documents or code comments, but currently supports Java, Kotlin and Python code.
Taking Java and Kotlin codes as examples, you can directly use standard methods /**to trigger the automatic generation of code comment documents.
Naming suggestions
When users need to rename certain variables, functions or declarations in code such as Java, Kotlin or Python, the AI assistant will give some corresponding content suggestions based on the context.
Commit information generation
The [Code Information Submission] window of the new version now also has a [Generate Commit Message with AI Assistant] button.
After clicking this function, AI Assistant will automatically generate Commit information according to the content difference of code changes, which can be said to be very convenient.
postscript
All in all, judging from some features and examples given in the new version, it feels pretty good.
Although some third-party plug-ins also support similar functions, this time it is an official new feature after all. I believe that there should be a good degree of completion in terms of functionality, stability, and user experience.
However, JetBrains officially stated that the current JetBrains AI service will not be open to all users immediately, but will invite a certain number of users to try it first, and then it will gradually roll out later.
What do you think of this new feature launched by JetBrains this time? Are you looking forward to it?
Note: This article has been included in the GitHub open source warehouse "Road to Programming" https://github.com/rd2coding/Road2Coding , which contains the self-study routes of the 6 major programming directions (posts) + knowledge points sorting out and interview test points I compiled , my resume, a few hardcore pdf notes, and the life and perception of programmers, welcome to star.