Python lab

by Cristian Vasquez
5
4
3
2
1
Score: 44/100

Description

Category: Coding & Technical Tools

The Python lab plugin provides an experimental interface for Python developers to integrate and test Python scripts within Obsidian. It allows users to implement and interact with scripts for various tasks, such as finding similar notes, translating text, or summarizing notes, through a simple GUI. The plugin communicates with a user-defined Python server to handle operations like inserting new text, replacing existing text, displaying elements in a panel, and facilitating chat-based interactions. Designed for rapid prototyping, it enables developers to explore Natural Language Processing and other Python-based functionalities before transitioning to JavaScript if needed.

Reviews

No reviews yet.

Stats

95
stars
11,423
downloads
3
forks
1,844
days
1,632
days
1,632
days
0
total PRs
0
open PRs
0
closed PRs
0
merged PRs
5
total issues
2
open issues
3
closed issues
0
commits

Latest Version

4 years ago

Changelog

Fixed a bug in the chat GUI

README file from

Github

Obsidian Python lab plugin

The Obsidian Python lab is just a collection of dumb interfaces to enable python developers to use/test functionality within Obsidian.

Motivation

Say you have a terrific script to:

  • Find similar notes to the current one.
  • Translate a text.
  • Write the summary of a note.
  • ....

And you want to quickly see if it's helpful in Obsidian, just using some python.

That is the purpose of this plugin!.

Why this plugin? Why not program all in Javascript?

Sometimes is quicker to experiment in Python and later, if you want, make it work in Javascript. This is especially true when using Natural Language Processing libraries.

How it works?

The plugin is just a GUI to make calls to a server of your choice. Currently, the plugin has implemented the following operations:

  1. Insert new text.
  2. Replace the current text.
  3. Show elements in a panel.
  4. Have a chat conversation

Use it

Plugin-Server interaction

  1. When the user runs one of the plugin's commands, for example, 'replacing fancy text,' the plugin makes a POST call to your server with some context data, such as the current note, what was selected, etc.

Say your terrific script returns a list of random notes of your vault. Then the plugin does:

POST: http://127.0.0.1:5000/scripts/some_list

With some context data

{
  "vaultPath": "/home/cvasquez/obsidian/development",
  "notePath": "snippets-plugin/Test1.md"
}

It then returns a JSON response, which the plugin uses to show something in a Widget

Response

{
  "contents": [
    {
      "info": { "score": "0.9820077811564822" },
      "path": { "path": "/path/to/the/note 1.md" }
    },
    {
      "info": { "score": "0.9365154046414078" },
      "path": { "path": "/path/to/the/note 2.md" }
    }
  ]
}

Quickstart

  1. Install the plugin. (Maybe you already did)

  2. Write a script somewhere

def hello():
    return {
        'contents': f'Hello world!'
    }
  1. Run the application to expose the script.

Server

  1. The script should be now be detected by the plugin. Then the operation needs to be configured to specify how it interacts with the user. Any change in the options will persist in the plugin configuration.

Options

  1. Use it!

Example widget

Example python server

I wrote a minimal server, obsidian-lab-py, that exposes some scripts. It might be useful to look at. I use something different each time, like this example.

Build the plugin

This is not necessary if the plugin is installed from the store. However, it is built like all the other plugins,

  1. Clone this repo into

/{vault}/.obsidian/plugins

  1. Install the dependencies

    yarn install

  2. build the app

    yarn build

This will build the main file; that Obsidian should detect. Activate the plugin from inside Obsidian, in community plugins

Forum

This repo has github discussions enabled.

Status

This is still a proof of concept; please send any feedback :)

Contributing

Pull requests are both welcome and appreciated.

Similar Plugins

info
• Similar plugins are suggested based on the common tags between the plugins.