IBM Cloud Docs
Getting code suggestions

Getting code suggestions

watsonx Code Assistant

Generate code suggestions by using chat conversations or completing code in the editor.

Using chat conversations

Enter a prompt that explains the code you need, and watsonx Code Assistant generates something you can choose to use.

  1. To start a new chat conversation, click the New chat icon New chat.
  2. Ask a question or give an instruction.
  3. Click the Send icon Send. Watsonx Code Assistant sends your input to the code model, and shows the response in the chat.

Transferring code between the chat and the editor

When you get a code suggestion that you want to use, you can insert a code block from the chat into your editor.

  1. In Visual Studio Code, position your cursor in the editor.
  2. In the chat, click the Insert at cursor icon Insert at Cursor.

If you have code in the editor that you want to use as context in the chat, you can copy the code into the chat.

  1. Highlight the code in the Visual Studio Code editor.
  2. Right click to show the context menu.
  3. Choose watsonx Code Assistant.
  4. Choose Add selected code to chat.

Using natural language

Each chat message is part of a chat conversation. Keep conversations focused around a specific subject or task. When you switch your context to another programming language, another project in your workspace, or a different programming task, create a new chat conversation. You get more relevant results for your questions.

Examples:

  • Create a Python class Customer. Include attributes for firstName, lastName and age.
  • Create a JavaScript function that fetches data from a REST API endpoint, dynamically populates an HTML table with the results, and updates the UI asynchronously

Using the prompt library

Watsonx Code Assistant includes a prompt library that contains a set of common, natural-language coding requests. You can use them as-is in your chat, or modify them for your use case.

To access the prompt library:

  1. Click the Open menu icon Open menu.
  2. Click Prompt library to open it.
  3. Click the Copy to clipboard icon Copy to clipboard.
  4. Click Prompt library again to close it.
  5. Paste the prompt into the chat.

Writing effective chat messages

For the best results in your conversation:

  • Write your chat messages in English.
  • Start with a simple and clear instruction.
  • Use subsequent chat messages to refine the code output of the model.
  • Be as specific and detailed as you can for each step. You can gradually enhance the created code through subsequent chat messages.
  • Use file and method references in your message to provide relevant context. For example, if you want the model to create a method that is similar to another method, add similar to @<method> to your message.
  • If you find the answers become less relevant, or if you start a new task, create a new chat conversation, and work from there. It is better to have many short chat conversations, each with a specific context, rather than one large conversation that might confuse the model with different and unrelated chat messages.

Referencing code in your workspace

To ask questions or refine a specific file, class, function, or method in your workspace, you can use a code reference. These references provide important context and can help to increase the accuracy of the answer.

  1. Type the @ symbol as part of your chat message to see a list of files, classes, and methods from your workspace.
  2. Filter the list by typing characters of the file, class, or method name you want to reference.
  3. Click to select the reference. Watsonx Code Assistant sends the contents of the reference as part of your message.

When you open a workspace folder, watsonx Code Assistant creates an index of these items in memory so you can reference these files and functions in the chat. Your editor also indexes files that you add or change during your session. The index contains up to 1,000 of the most recent files in these programming languages:

  • C
  • C++
  • Go
  • Java
  • JavaScript
  • Python
  • TypeScript

Use cases with examples that reference code:

Chat messages that reference code
Use case Example message
Generate a function based on an existing function Create a method send_translate_message that is similar to @send_code_explanation_message
Generate a unit test that follows existing unit tests Create a unit test for @getName that is similar to the unit tests in @testLoadTablesChildConnectionReceiverJob.h
Enhance existing functions Add error handling and log statements to @format_documents
Enhance existing functions Update @setEmail with error handling for null strings
Explain code What does @main.py do
Explain code Explain the logic of @send_invoice
Generate documentation for functions and classes Add Javadoc to @Customer

Using chat commands

Chat commands that you can use to perform certain tasks:

Chat commands
Command Description For more information
/document Generate documentation for a referenced file or a method or a class in the active workspace Generating documentation
/docs Ask questions that reference IBM and Red Hat product documentation Getting answers from IBM documentation
/explain Use generative AI to analyze and summarize your code to understand what the code does Explaining code
/help Get quick assistance about tasks Type /help in the chat
/unit-test Generate unit tests for various programming languages Generating unit tests
/translate Translate code from one language to the other Translating code

Managing chat conversations

You can use the Chat sessions menu to manage your existing chat conversations.

To open the menu:

  1. Click the Open menu icon Open menu.
  2. Click Chat sessions.

Completing code in the editor

Get code suggestions by entering code or comments in your editor.

Using the editor in Eclipse

For C, C++, Go, JavaScript, Python, and TypeScript:

  1. Right-click on the file.
  2. Click Open with, then click Text Editor.

For Java, inline code generation with the Java Editor in the Eclipse IDE is supported in watsonx Code Assistant but first needs to be set up.

  1. Go to Content Assist in the Eclipse settings.

    Navigation to Content Assist
    OS Path
    Mac Eclipse > Settings > Java > Editor > Content Assist
    Windows Window > Preference > Java > Editor > Content Assist

  2. In the Insertion section, clear Insert single proposals automatically.

  3. In the Auto Activation section, add ._abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ to Auto activation triggers for Java, and confirm that Enable auto activation is checked.

  4. Click Apply and Close.

Single-line completion

As you write a line of code in your editor, watsonx Code Assistant can suggest code to complete it.

  1. Start typing a line of code, then pause.
  2. Watsonx Code Assistant adds a code suggestion to complete the line that you typed.
  3. In Visual Studio Code, press Tab to accept the suggestion. In Eclipse, press Return or Enter to accept the suggestion. Or press Esc to reject.

Multiline completion

You can get a multiline code suggestion based on a line of code or a comment that you enter.

  1. Start typing a line of code, or enter a comment that describes the code you want.

  2. Use a keyboard shortcut.

    Default keyboard shortcuts
    Mac Windows
    Option+. Alt+.

  3. Watsonx Code Assistant adds a multiline code suggestion.

  4. In Visual Studio Code, press Tab to accept the suggestion. In Eclipse, press Return or Enter to accept the suggestion. Or press Esc to reject.

Changing keyboard shortcuts

You can change the keyboard shortcuts that you use to get a code suggestion, if the default combinations are already assigned to other tasks.

To change keyboard shortcuts in Visual Studio Code:

  1. Click the Code menu.
  2. Select Settings.
  3. Select Keyboard Shortcuts.
  4. Search for wca.inlineSuggest.trigger.
  5. Click the Change Keybinding edit pencil icon and assign a different key combination.

To change keyboard shortcuts in Eclipse:

  1. Click the Eclipse menu.
  2. Select Settings.
  3. Type Keys to filter the settings, then click Keys.
  4. Type WCA to filter the list.
  5. Select the keyboard shortcut that you want to change.
  6. In the Binding field, enter a new key combination.

Disabling code completion

If necessary, you can disable the completion of code in the editor window.

To disable code completion in Visual Studio Code:

  1. In the lower right corner of the editor, click the Disable watsonx Code Assistant Completions icon images.
  2. You can re-enable completions by clicking the Enable watsonx Code Assistant Completions icon again.

To disable code completion in the Eclipse IDE:

  1. In the editor toolbar, click the Disable watsonx Code Assistant Completions icon.
  2. You can re-enable completions by clicking the Enable watsonx Code Assistant Completions icon again.

Best practices for generating code suggestions

Use these tips to get the best results.

Use chat for longer code blocks. Use in-editor code generation for refinement and boilerplate code.

Use chat to:

  • Generate an outline of a class
  • Specify a method in detail
  • Refine the suggested code with more instructions (for example, rename or replace)

Use in-editor code generation to:

  • Complete the line of code that you started to type
  • Automatically create boilerplate code, such as getters and setters
  • Create methods and code that can be described with a single comment line

Comment-to-code: use descriptive comments, not instructions

When you use comment-to-code in the editor, write a comment that describes the intended behavior, similar to how you write comments for code you wrote. For example, use //return even numbers from an arraylist or //method that returns even numbers from an arraylist.

Don't write your comment as an instruction, such as //write a method that returns even numbers from an arraylist. Watsonx Code Assistant uses models that are trained to complete code on data that contains many typical, descriptive comments, so these types of comments yield better results.

Context used for in-editor code generation

The key ingredient for code generation is context, that is, the surrounding code that you pass to the model. For in-editor code generation, watsonx Code Assistant uses:

  1. The 20 lines of code before the line where the generation is triggered.
  2. The 20 lines of code after the line where the generation is triggered.
  3. Up to 200 lines from the beginning of the current file where generation is triggered.
  4. Up to 5 code snippets that are similar to the code that surrounds the line where generation was triggered. These snippets are taken from the last 10 files that you opened in the same programming language as the current file.

To improve the results of in-editor code generation:

  • Open files in your workspace that have code you want watsonx Code Assistant to know about. Watsonx Code Assistant tries to find code in these files that is similar to where you trigger completion, for example, methods that use similar classes and variables, and then sends this code to the model as context.
  • Help watsonx Code Assistant by starting with a class definition or adding import statements, or use the watsonx Code Assistant chat to create the initial outline, copy the result into a file in your workspace, and then continue with in-editor code generation. Don’t use in-editor code generation when you start from zero or almost empty files, as these conditions do not provide enough context.