c:\Users\xxxxxx\OneDrive - BD\Documents\Projects\ShinyPython.venv\Scripts\python.exe -m shiny run --port 51127 --reload "c:\Users\xxxxxx\OneDrive - BD\Documents\Projects\ShinyPython\app.py"

'c:\Users\xxxxxx\OneDrive' is not recognized as an internal or external command,
operable program or batch file.

If you put double quotes on the first part it works.

Currently, debugging a Shiny app requires configuring launch.json. It'd be nice to have a "Debug Shiny App" command in the Run menu instead.

• Options
  • shiny.r.port
  • shiny.r.devmode
• Snippets
• Commands
  • shiny.r.runApp

Thank you for this great extension! 🌻

Currently, the "Run" button defaults to the main interpreter (e.g. /usr/bin/python3). Is there a way to specify a different interpreter in Settings? I'm inside a conda environment and right now I'm just doing shiny run --reload and that works, but would be neat to just click a button.

When I try to run my Shiny python app with the extension via the run button, I get the following:

/usr/bin/python3 -m shiny run --port 45287 --reload "/workspaces/MY_APP/app.py"
/usr/bin/python3: No module named shiny

The extension is defaulting to /usr/bin/python3, but I'm using the /usr/local/bin/ python install.

I don't see any docs on changing the path used by the extension for running shiny run. It would be helpful to have more docs on changing the extension settings.

For reference, my VS Code .devcontainer.json file:

{
    "name": "MY_APP",
    "build": {
        "dockerfile": "Dockerfile",
        "context": ".."
    },
    "extensions": [
        "github.copilot",
        "github.copilot-chat",
        "ms-python.python",
        "ms-azuretools.vscode-docker",
        "codezombiech.gitignore",
        "Gruntfuggly.todo-tree",
        "redhat.vscode-yaml",
        "george-alisson.html-preview-vscode",
        "posit.shiny-python"
    ],
    "settings": {
        "python.linting.enabled": true,
        "python.linting.pylintEnabled": true,
        "python.analysis.typeCheckingMode": "basic",
        "python.analysis.diagnosticSeverityOverrides": {
            "reportUnusedFunction": "none"
        }
    },
    "features": {
        "ghcr.io/devcontainers/features/docker-in-docker:2": {
        "version": "latest"
        }
    }
}

My dockerfile:

FROM python:3.9-slim

WORKDIR /app

RUN apt-get update && apt-get install -y \
    build-essential \
    curl \
    software-properties-common \
    git \
    libpq-dev \
    && apt-get clean \
    && rm -rf /var/lib/apt/lists/*

COPY ../requirements.txt ./

RUN pip3 install -r requirements.txt

The extension never loads the app correctly when working with large data sets. No error message is given, it just stops after the message INFO: Application startup complete.

I'm loading a pandas dataframe with 520 rows and 20 columns. When I take only 10 rows from the original data set, the app loads fine.

I have the shiny extension installed. But when I open an app.R, it does not load shiny extension. The "Run Source Code" button still sends the R code to radian console as any other R programme, and nothing happens. The app is not started. What do I need to do to get the shiny extension to come into play.

In mid-May (when Joe gets back from vacation):

• Publish new extension under Posit.shiny
  • Check version number in package.json
  • Update CHANGELOG.md
  • Ensure CHANGELOG.md version number is correct
  • Smoke test
  • Create release on GitHub with a vX.Y.Z tag, pasting in the changelog entries
  • Add .vsix to release artifacts
  • Blog/slack if appropriate
  • Close version milestone, reallocate issues if needed
  • After publish, increment version number
• Simultaneously, publish new version of Posit.shiny-python that removes all contribution points and depends on Posit.shiny, and rename to "[Deprecated] ..."
  • Check version number in package.json
  • Update CHANGELOG.md
  • Ensure CHANGELOG.md version number is correct
  • Smoke test
  • Create release on GitHub with a vX.Y.Z tag, pasting in the changelog entries
  • Add .vsix to release artifacts
• Unpublish the very old, long-deprecated rstudio.pyshiny so we don't have two deprecated extensions showing up in search
• Review usage of Posit.shiny-python and PR the name change (e.g. Posit.shiny-python is a recommended extension in py-shiny)

Open the ui.R, server.R or global.R file of an Shiny for R app.

1. The Run button modification isn't activated.
2. Manually running the "Run Shiny app" command tries to run runApp("/full/app/path/ui.R"), which fails. It should be runApp("/full/app/path/").

The path escaping for cmd.exe turns

"C:\Program Files (x86)\R\R-4.3.3\bin\Rscript.exe"

into

"C:\Program Files ^(x86^)\R\R-4.3.3\bin\Rscript.exe"
#> The system cannot find the path specified.

Testing with Windows 11 via Parallels

If/when posit-dev/py-shiny#521 happens, this should follow naturally.

Hello, the link to the docs leads to a 404 page, the link appears in the README, and so shows up on the extension's page in VSCode.

This is the link - https://shiny.rstudio.com/py/docs/install.html#configure-visual-studio-code

and this is what I see after following the link -

We're expecting powershell.exe but it's now C:\Program Files\PowerShell\7\pwsh.exe.

It would be able to run Shiny for R apps with the shiny vscode extension.

Using https://code.visualstudio.com/api/references/activation-events#onUri we could receive URLS like

vscode://Posit.shiny/save?url=https%3A%2F%2Fshinylive.io%2Fpy%2Feditor%2F#code=...

which would let us add an "Open in VS Code" link to the ShinyLive share sheet.

Hello, for me it would useful if I could somehow define the path to the R executable on my machine instead of using the RPATH environment variable.

Would it be possible to do something similar to the Python version where it looks for the path definied in the Python Extension? Instead looking for the path definined in the R Extension using the following record in setting.json.
"r.rpath.windows":

when I hit run button,
This print out in terminal
C:\Program Files\Python310\python.exe -m shiny run --port 61780 --reload "c:\\Users\\user.name\\path\\app.py"

bash: C:Program: command not found
So there is a space in path C:\Program Files\ , thats the reason it is not working. But, by default my python exe file is there.
If I add quote manually in the python exe file path and run then that works.

'C:\Program Files\Python310\python.exe' -m shiny run --port 61780 --reload "c:\\Users\\user.name\\path\\app.py"

I am building Shiny Python apps with VSCode 1.74.3 and Python 3.10.9 in Windows 11. I have confirmed that shiny is installed and can be imported. I installed the Shiny extension for VS Code using version 0.0.3. I have created a file called app.py and copied the script for the basic app on your website. When I go to hit run by clicking the run icon in the top right, I am not presented with a choice to run a shiny app. When I try to run the app, the output in the terminal is simply python.exe c:/dir0/dir1/app.py and exits. No app is found at the 127.0.0.1:8000

Any ideas what might be missing? Thanks

Thanks for the VS Code extension. It would be nice if there was a default keyboard shortcut to shiny.python.runApp. The shortcut could be the same as the render or preview Quarto extension shortcut ( Ctrl+Shift+K or Ctrl+Shift+L). Edit: I realized after posting that this issue should belong in the vscode extension repo which doesn't appear to be public at this time. Please close this issue if it is already being addressed.

I receive the following error in Chrome and VS Code Simple browser when trying to use input.textarea with autoresize=True.

Failed to load module script: Expected a JavaScript module script but the server responded with a MIME type of "text/plain". Strict MIME type checking is enforced for module scripts per HTML spec.

Can successfully run shiny from terminal but shiny play/run button in VScode does not seem to acknowledge the virtual environment.

running VScode on Mac Sonoma 14.2.1 with an M2 chip

(base) brendan@Brendans-Air distrodash % conda activate distrodash2
(distrodash2) brendan@Brendans-Air distrodash % /Users/brendan/anaconda3/bin/python -m shiny run --port 50911 --reload "/Users/brendan/Documents/GitHub/distrodash/shinypy/app.py"
/Users/brendan/anaconda3/bin/python: No module named shiny
(distrodash2) brendan@Brendans-Air distrodash % python -c "import shiny; print(shiny.__version__)"
 0.7.1
(distrodash2) brendan@Brendans-Air distrodash % /Users/brendan/anaconda3/bin/python -m shiny run --port 50911 --reload "/Users/brendan/Documents/GitHub/distrodash/shinypy/app.py"
/Users/brendan/anaconda3/bin/python: No module named shiny
(distrodash2) brendan@Brendans-Air distrodash % cd shinypy
(distrodash2) brendan@Brendans-Air shinypy % shiny run app.py
INFO:     Started server process [58637]
INFO:     Waiting for application startup.
INFO:     Application startup complete.

• Sign-off from QA
• Check version number in package.json
• Update CHANGELOG.md
• Ensure CHANGELOG.md version number is correct
• Smoke test
• Create release on GitHub with a vX.Y.Z tag, pasting in the changelog entries
• Add .vsix to release artifacts
• Blog/slack if appropriate
• Close version milestone, reallocate issues if needed
• After publish, increment version number

This could be used to enable Shiny-based gadget apps that could be run as shortcuts. Here's a sketch of a shortcut to run the chattr R app.

    {
        "key": "cmd+k cmd+c",
        "command": "shiny.r.runApp.code",
		"name": "Chat with chattr",
        "args": {
            "langId": "r",
            "code": "chattr::chattr_app()"
        }
    },

The app would start up in a named terminal and the app would open in a simple browser window so that it could be moved around among the editor panes.

We should follow progress in posit-dev/positron#3173, if we can use multi-session support in Positron, that would improve the utility via hooks available in rstudioapi.

https://github.com/posit-dev/pyshiny-vscode/blob/3635a936a35859af82b3a02bf8246142d0631635/package.json#L18-L24

The onCommand events are flagged by linters:

This activation event can be removed for extensions targeting engine version ^1.75 as VS Code will generate these automatically from your package.json contribution declarations.

I think this is safe: VS Code 1.75 was released in January 2023.

Apparently, if you search the posit-dev org for "shiny" repos, you won't find this one. https://github.com/orgs/posit-dev/repositories?language=&q=shiny&sort=&type=all

From that search it also seems that, we've coalesced around the py-shiny or py- prefix.

Edit: actually I think we should call this repo shiny-vscode (see #29).

https://github.com/posit-dev/pyshiny-vscode/blob/505043eb0903696b28cee51aee592d6c3bb7e72e/package.json#L2-L4

this could instead be

  "name": "shiny-python",
  "displayName": "Shiny",
  "description": "Run and debug Shiny apps in Python or R.",

RStudio Workbench VS Code users cannot install extensions from the Microsoft Marketplace. Instead, the extension browser searches the Open VSX Marketplace. We should publish the latest extension to Open VSX, so our pro customers can install the extension.

Two major features were added for rc-v0.2.0:

1. #30
2. #44

Follow-up issues found in this process:

1. #52

Shiny for R apps

This feature brings support for running Shiny for R apps. The extension activates for R files named app.R, ui.R, server.R, global.R, app-*.R or *-app.R. Clicking the play button for such an app should run the app and preview it in the location determined by shiny.previewType (internal, external or none).

We need to check that this part of the extension works in the following environments:

1. macOS
2. Windows
   • For Windows, see the testing notes in #26, using a Shiny for R app in place of a Shiny for Python app.
   • Test in PowerShell, cmd.exe and Bash terminal environents.
3. Linux
4. GitHub Codespaces
5. Posit Workbench

Shinylive

• Test that we can share single-file and multiple-file apps to Shinylive. (See #44 for details).
• Test that we can save single and multiple-file apps from Shinylive.
  • Single-file apps prompt users to pick a path to a file, e.g. examples/app.py. VS Code will prompt the user if they want to overwrite an existing file.
  • Multi-file apps prompt users to pick a non-existent directory.
• Apps should roundtrip: create a shinylive link, choose to copy it, then save the app from the shinylive link.
• Test hosts: both shinylive.io (released shiny) and posit-dev.github.io/shinylive (dev shiny)
• Test output locations: open in browser and copy
• Test modes: app or editor
• The above settings can be fixed by choosing default values in the settings UI.

For the purposes of testing the extension, a Shinylive link is successfully created when opening the app in editor mode contains the same code as seen locally. There are many reasons why an app my fail to actually run in Shinylive; if the code and files you intended to send are indeed included in the editor, then the link was successfully created. Similarly, saving a Shinylive app locally should create files containing the expected code.

If port access is controlled in the environment running VS Code, users can set the port-related options to a fixed port to an allowed, open port. However, if the user has multiple VS Code instances open, they are limited to running one Shiny application at a time at the fixed port address.

We could update shiny.python.port, shiny.r.port and shiny.python.autoreloadPort to accept a range or array of ports:

"shiny.python.port": "8010-8020,8123",

The syntax could be {start}-{end} for start through end (inclusive) or comma separated values. The above would include ports 8010, 8011, ..., 8020, and 8123.

This would give system administrators the ability to open a limited range of ports, while allowing multiple apps to be run simultaneously.