Hello Swift Community,
I am a student interested in participating in GSoC 2025 under the Swift organization. Since all official project ideas are taken, I would like to propose a new project:
Project Title: Swift Code Auto-Fix System
Overview:
This project aims to develop an automated tool that can detect common Swift programming errors and automatically fix them. Similar to how Python’s black auto-formats code, this tool will analyze Swift code, find mistakes, and suggest or apply corrections.
Key Features:
Static Code Analysis:Identify syntax errors, unused variables, missing semicolons, and other common mistakes.
Auto-Fix Suggestions:Provide recommendations for fixing errors.
AI-Powered Learning:Improve suggestions using AI models trained on Swift coding patterns.
Swift Compiler Integration: Work alongside SwiftLint and SourceKit-LSP to improve Swift development.
Technologies & Tools:
Swift
SourceKit-LSP
SwiftSyntax (for parsing and analyzing Swift code)
Machine Learning (optional, for AI-based improvements)
Why is this Project Important?
Helps beginners avoid common Swift errors.
Saves time for developers by automating error corrections.
Enhances code readability and maintainability.
I would love to hear feedback from the mentors and the community about this idea. Is this feasible for GSoC? Should any improvements be made before submitting a proposal?
Hi there, I'm the org admin for Swift participating in GSoC.
I'd like to clarify that that's not how GSoC works, no projects are "taken" until organizations announce which project/contributor pairings are accepted. Currently is the proposal submission phase:
March 24 - 18:00 UTC
GSoC contributor application period begins
April 8 - 18:00 UTC
GSoC contributor application deadline
April 29 - 18:00 UTC
GSoC contributor proposal rankings due from Org Admins
Until projects are announced, anyone may submit proposals and it's up to mentors to later on decide which is the best proposals, rank them, and select a number of projects to run (and then requests slots for those from GSoC).
So it's not the case that if someone started a forums thread and they say they'd like to work on something, that they are granted that project -- it all depends on the quality of the final proposals, which are submitted to the http://summerofcode.withgoogle.com/ website.
The project idea you suggest is interesting, but it is also not too well defined and very hard to pull off well. To be successful you would have to find a mentor willing to guide this project, and specify very precisely which outcomes are expected, the timespan of the project, and future steps to have it integrated into the Swift project.
If you'd like to participate in Swift's GSoC this year I think it'd be easier to stick to an existing idea and polish up your understanding of it. Those projects already have mentors assigned, so you don't have to "hunt for" a mentor willing to take on the project, and they're a bit more clearly defined already.
Hope this helps and I look forward to seeing a proposal from you! :-)