R Extension for Visual Studio Code

The R extension provides a comprehensive environment for R development, including:

• Language Support: R Language Service (completion, signatures, diagnostics, etc.) and R Markdown support (chunk navigation, execution, and preview).
• Interactive Viewers:
  • Data Viewer: View data.frame or matrix in grid or treeview structures.
  • Plot Viewer: Interactive viewing with jgd or httpgd support.
  • Webpage/Browser Viewers: View htmlwidgets and interactive shiny apps.
• Workspace Management: Sidebar tools for viewing global variables (Workspace viewer), searching help topics (Help pages viewer), and managing/installing packages.
• Development Tools: Package development commands (build, test, install, load) via devtools integration and RStudio add-in support.
• Remote Development: Full support for working via SSH, Containers, or WSL.

Introduced in version 3.0.0-rc, sess is a modern R package that implements a new IPC (Inter-Process Communication) architecture for the extension.

Key improvements over the legacy file-based IPC:

• Architecture: Uses an in-memory WebSocket architecture based on JSON-RPC 2.0.
• Performance: Eliminates the need for OS-level file watchers, resulting in faster and more robust communication.
• Installation: The extension will automatically prompt you to install the sess package if it is missing when you start an R session.

In the R help viewer (v2.5.2+), you can interact with code sections:

• Hover: Highlights code sections.
• Click: Copies the code to the clipboard.
• Ctrl+Click (Win/Linux) / Cmd+Click (macOS): Sends the code directly to the R terminal.

You can customize this behavior using the r.helpPanel.clickCodeExamples setting.

The R session watcher (introduced in v1.2.0) allows the extension to interact deeply with your R session.

Key capabilities:

• Attach Active Terminal: You can attach the extension to an R session via a command or by clicking the status bar item.
• Auto-attach: If you source init.R in your .Rprofile, starting an R session will automatically notify vscode-R to attach.
• Global Symbol Hover: Provides hover information for global symbols in the attached session.
• Plotting: Shows plot files on the fly.
• WebViews: Uses WebViews to present htmlwidgets, Shiny apps, and data structures (like data.frame or list) when calling View().

The sess protocol is a modern IPC layer designed for R-to-client communication. It relies on two core principles:

1. JSON-RPC 2.0: All communication uses JSON-RPC 2.0 message formats, supporting Notifications (one-way), Requests (expecting a response), and Responses.
2. JSON Lines (JSONL) Framing: Messages are delimited by newlines (\n). This allows the receiver to buffer stream chunks and dispatch complete JSON objects safely, even when the stream is fragmented.

Unlike previous WebSocket-based transports, sess uses UDS/Named Pipes and has removed the authentication token exchange requirement.

To improve the R development experience in VS Code, the following tools are recommended:

Plotting Backends

Install one of these for an improved interactive plotting experience:

• jgd: A lightweight JSON graphics device with native vscode-R integration.
• httpgd: An SVG-based graphics device served via HTTP and WebSockets.

Enhanced Consoles

• arf: A modern R console featuring syntax highlighting, fuzzy history search, multiline editing, and vi/emacs keybindings. It is the successor to radian.

Debugging

• VSCode-R-Debugger: A separate VS Code extension required to enable R debugging capabilities.

As of version 2.1.0, the R language service is integrated directly into vscode-R.

Migration Steps:

1. Search for the r-lsp extension in VS Code.
2. Uninstall the r-lsp extension.
3. Ensure the R package languageserver is installed in your R environment.
4. vscode-R will now automatically start the language service.

To work with existing self-managed, persistent R sessions when the extension is upgraded, you must source the init.R file again before attaching. This ensures the session watcher and other extension features are correctly linked to your R process.

Run the following command in your R terminal:

source(file.path(Sys.getenv(if (.Platform$OS.type == "windows") "USERPROFILE" else "HOME"), ".vscode-R", "init.R"))

To establish an IPC connection between an R session and a client (like the VS Code R extension), use the sess::connect() function.

By default, if pipe_path is NULL, the function attempts to resolve the connection path using the following priority:

1. The SESS_PIPE environment variable.
2. A session file located at ~/.vscode-R/sessions/{PID}.json (specifically the pipe field).

Transport mechanisms used:

• Unix domain sockets (macOS/Linux)
• Windows named pipes

sess::connect(
  pipe_path = NULL,      # Character: pipe/socket path. NULL -> SESS_PIPE or session file fallback
  use_rstudioapi = TRUE, # Logical: enable rstudioapi emulation
  use_httpgd = TRUE      # Logical: use httpgd for plotting if available
)

The R extension provides a dedicated Help Tree View (rHelpPages) in VS Code to browse, search, and manage R documentation. The tree view is organized into several functional areas:

• Navigation & Search: Quick access to the R help Home, searching by alias (?), searching by text (??), and opening help for the currently selected text in the editor.
• Package Management: A section titled "Help Topics by Package" allows you to browse installed packages, filter them, and manage favorites. You can also install new CRAN packages or update existing ones directly from the tree.
• Local Previews: If you are developing an R package locally, a "Local Preview" node appears, allowing you to browse the documentation of your local package.
• Maintenance: Options to clear the help cache and restart the help server to ensure documentation is up to date.