GSoC 2019 Ideas

Create a Spyder-VCS plugin

This purpose of this project is to create a Spyder plugin that allows users to work comfortably with Git repositories in Spyder and submit their work to Github. In particular, this plugin should allow users to:

• Browse current active branches and change between them (from git-cola).
• View git diff, git log and git blame for each file, and launch the git mergetool if a merge conflict is detected (from git-cola).
• See an indicator of lines changed, added and removed in an Editor gutter panel (like VSCode offers)
• Commit work to the git repo with a commit message, either in total or by hunks (from git-cola).
• Explore git history (with qgit).
• Be able to easily define or create a remote on Github.

Motivation

Better Git/Github integration is perhaps the most-requested Spyder feature by our users that has not yet been implmeneted or scheduled for implementation. This would allow scientists, engineers and data analysts to much more easily employ powerful VCS tools to greatly aid reproducibility and reusability of their code and share them with the world, all within a familiar graphical user interface.

Expected Skills

Familiarity with Python, PyQt/PySide, Git and Github.

Available Mentors

• Carlos Cordoba (@ccordoba12)
• C.A.M. Gerlach (@CAM-Gerlach)

Add refactoring/formatting/code snippit support to Spyder

Many other popular IDEs have a variety of code cleanup and automated refactoring tools built in, and thus this would be a valuable addition to Spyder as well, and one that a number of users have requested. This would be implementing by providing a GUI (menu and/or panel) for and employing the existing refactoring tools offered by Rope within Spyder. This would all be accessed through the standardized Language Server Protocol client already integrated within Spyder to provide all code-related services. Additionally, support for code snippits and auto-formatting could be added.

Motivation

Many users have requested better refactoring and auto-formatting tools like those offered by other IDEs, along with code snippit support, in order to aid their workflows. Refactoring and auto-formatting would help Spyder users right better quality, more re-usable code, and all three would greatly speed up many algorithm, scientific package, and engineering toolkit development workflows that Spyder is built for, while giving it the power of a large, software-development focused IDE.

Expected Skills

Familiarity with Python, PyQt/PySide, and basic refactoring practices. Experience using rope and/or the Language Server Protocol a plus.

Available Mentors

• Carlos Cordoba (@ccordoba12)
• C.A.M. Gerlach (@CAM-Gerlach)

Update and improve Spyder-Reports

Spyder-Reports is a Spyder plugin developed to allow users to create dynamic, rich text documents incorporating code blocks, Markdown prose, and dynamically updated output and figures/visualizations, all within Spyder. The goal is to bring the power and functionality of a tool equivalent to Sweave/Knitr and R Markdown to a Python IDE.

This project will update spyder-reports to work with the latest versions of Spyder and pweave (our backend renderer). In addition, it'll add the ability to view reports in PDF format inside Spyder, improve markdown rendering in Spyder (e.g. code fences are not syntax highlighted right now), extend the syntax in Pweave to add features like citations and cross-references, and add support for LaTeX as well as Markdown for scientific output, as this is already fully implemented in pweave.

Motivation

Although Jupyter notebooks can be a good reporting facility for some applications, they offer critical shortcomings for many others: they require Jupyter Notebook to view them, along with all the dependencies of the original work; they are often unworkable for projects with large datasets or long compute times as they most be executed dynamically, they are not self-contained, cells can be executed in any order and have other unexpected quirks harming reproducibility. Meanwhile, presenting raw Python code requires either sharing the source files with all their extra cruft and inaccessibility to non-experts, and without contextual prose, output or plots; or manually copying and pasting code, output and figures into a separate document, and having to update everything by hand when anything changes.

Therefore, integrating the capabilities of pweave with Spyder allows researchers and professionals to easily write up and share their work and results with colleagues, managers and the public at large, and also enables the creation of informative guides, tutorials and even whole textbooks, all within Spyder.

Expected Skills

Familiarity with Python, PyQt/PySide and Markdown or LaTeX. Experience using pweave, Sweave, or R Markdown a plus.

Available Mentors

• Carlos Cordoba (@ccordoba12)
• C.A.M. Gerlach (@CAM-Gerlach)

Update and improve Spyder-Notebook

Spyder-Notebook is a plugin developed to allow users to open, edit, execute and interact with Jupyter Notebooks all within the Spyder IDE, bridging the popularity and interactivity of the former with the powerful tools and re-usable workflow of the latter.

This project will update spyder-notebook to work with the latest versions of Spyder and Jupyter. In addition, it will add several features requested by the user community employing this tool (see the Spyder Notebook Github for specific examples).

Motivation

Although Jupyter notebooks offer some advantages for particular niche applications, they have many downsides if used on their own for serious research. They require Jupyter Notebook to view them, rather than being usable with any standard, interoperable Python interpreter; they have limitations when working with large datasets or multiple threads, cells can be executed in any order and have other unexpected quirks harming reproducibility, and the native UI offers extremely limited tools for developing proper scientific workflows.

However, many scientists have grown accustomed to Juptyer-Notebook despite its many limitations due to it being a familiar and accessible tool, and others used to the Spyder workflow may have the need to work with notebooks from time to time. Therefore, integrating the capabilities of Jupyter within Spyder offers considerable value to both large communities.

Expected Skills

Familiarity with Python, PyQt/PySide and Jupyter Notebooks.

Available Mentors

• Carlos Cordoba (@ccordoba12)
• C.A.M. Gerlach (@CAM-Gerlach)