Code Analysis for Jupyter Python notebooks

Notebook
,
1

Hi all,

We (Code Inspector aka Codiga) worked on a Chrome extension to analyze Python code in Jupyter notebooks. We want to provide a tool that plugs on top of Jupyter notebooks and surfaces bugs directly in the notebook so that developers can see (and fix) potential bugs before they ship their code.

We just released a first beta version and we are looking for feedback from Jupyter users. I would love to know if this plugin can help developers using Jupyter notebooks.

Thanks,

Note: This forum is targeted at the ecosystems of tools around Jupyter notebooks. If this message does not belong to this forum, please remove it (or tell me where to post it).

2

Hi @juli1, It’s nice that you are working on improving the code quality in notebooks. Please note that the current development focuses on JupyterLab and RetroLab (which is targeted at users who would like to have the classic Notebook layout but with new features enabled by JupyterLab components). Do you plan to support JupyterLab too?

While I personally won’t grant browser extensions access to my websites, I understand that there may be users who will like that approach. Have you considered offering your service as a remote Language Server Protocol server? Over at GitHub - jupyter-lsp/jupyterlab-lsp: Coding assistance for JupyterLab (code navigation + hover suggestions + linters + autocompletion + rename) using Language Server Protocol we are working on making LSP a prime citizen in Jupyter and while we ship with open-source tooling, I always had the remote proprietary LSP servers such as sourcegraph in mind as a something that ought to be in scope (if anyone wants to use it).

I could not find much about the company or privacy and processing terms. Are you analyzing the code locally in the extension or sending it for processing somewhere else?

3

Thanks for your message! Our plugin already works with JupyterLab (and we will implement RetroLab in the future). Noted that you focus on JupyterLab now, which is also helping us to focus our development efforts.

We do not have yet an LSP implementation and that is something we are considering. I was not aware of any jupyter and LSP integration, that is really awesome to see and is a strong indication we might need to implement it.

About processing, the code is processed on a remote server and we never store the source code or process information (we do not store or keep personal information). All our code analysis platform is available through our API and we offer the same capabilities for VSCode, IntelliJ or Chrome. Our privacy policy and terms of services are really clear about this.

2 Likes
4

Oh, so I was basically looking at a promo website (https://codiga.io/) and not the website where the useful info is (https://www.code-inspector.com/). Thanks for the extra details!

5

We are rebranding soon - the codiga.io is an attempt and will soon be replaced with a brand new website. Sorry for not having it with all the information - it’s coming :slight_smile:

1 Like