Repository navigation
A reliable idiomatic wrapper for the C library very_nauty [$400] #448
Description
Activity
Is this issue still open. I wanted to try this
Thanks for your interest! It is still open. Could you share a bit about your background, have you worked with Julia and with C before? Have you worked on Graph theory more broadly?
I’ve done coursework in C language as well as in graph theory, though I haven’t yet contributed to open source projects or created independent projects in those areas. I don’t have prior experience with Julia either.
I recently completed my GSoC project with DeepForest (under NumFOCUS), and while exploring other NumFOCUS-related projects, I came across this one and found it interesting. I was also looking for gigs and saw that this project had a bounty on it, which made me want to try for the bounty. That’s how I came here.
@naxatra2 , apologies, I thought I had responded. Given your lack of experience with Julia, I would like to go through some onboarding before exclusively assigning the bounty to you. For instance, could you choose whichever of the TODO items in the bounty description seems easiest and submit a PR for it? I would be happy to help with first steps and polish, etc. Once such an initial PR is merged, demonstrating your familiarity with the tools and the language, I would be happy to assign the bounty exclusively to you.
Reacted by NakshatraI intend to tackle this. I see I was just invited to join the org, just about now, though. I suppose it is better if I postpone joining the org until later.
Hi, @nsajko ! The invite was because of your message on the Graphs channel on slack about the libraries you are setting up -- you can go ahead and accept that invite if you would like, no need to wait.
As for tackling this bounty, it is free for the taking at the moment.
Reacted by Neven SajkoHi @Krastanov, I've started working on this bounty. I've submitted two PRs to VNGraphs.jl:
- PR Transfer issues from LightGraphs.jl #18 — Complete Graphs.jl AbstractGraph interface:
edgesiterator,outneighbors/inneighbors,rem_edge!,add_vertex!,copy,has_edgebug fix, and GraphsInterfaceChecker.jl compliance tests. All 266 tests pass. - PR broken documentation links in the README #19 — VNAlgorithm dispatch:
is_connected,connected_components,chromatic_number,edge_chromatic_number, andclique_numberwith auto-conversion from any AbstractGraph. All 307 tests pass.
Remaining items from the checklist:
- Fast conversion/views (the current conversion copies edges; I can explore a zero-copy view wrapper if desired)
- Declare
chromatic_number,clique_numberetc. in Graphs.jl itself (separate PR)
Happy to iterate based on review feedback.
- PR Transfer issues from LightGraphs.jl #18 — Complete Graphs.jl AbstractGraph interface:
I'd like to apply for this bounty. I understand the task is to create a reliable, idiomatic Julia wrapper for the C library very_nauty, enabling seamless integration with Graphs.jl's API and existing algorithms.
My delivery plan:
• Implement fast converters/views between Graphs.jl types and very_nauty C structures, ensuring full Graph API support for VNGraph types.
• Develop dispatch mechanisms to route existing Graphs.jl operations to very_nauty implementations, with comprehensive tests and documentation.
Please provide clear acceptance criteria to ensure alignment. I'm [Your Name], GitHub username [username], with experience in Julia C interfaces and Graphs.jl.Hi maintainers, I would like to take "A reliable idiomatic wrapper for the C library
very_nauty[$400]". I will start with a scoped first PR and keep implementation notes transparent. First PR plan: reproduce the issue against acceptance criteria, ship a minimal fix, and include focused tests in the first PR. First PR deliverable: minimal-risk patch, clear commit scope, and verification notes. I will add a short risk/checklist section so review and merge are faster. You can reach me at lmee_mail@163.com. Best, Jerry Li/attempt #448
Hi @Krastanov, I read the complete issue and comments, including the claim process and payout notes.
I can take a scoped first milestone for this bounty:
- wire core
Graphs.jlAPI dispatch paths intoVNGraphs.jl, - add test coverage for conversion + traversal primitives,
- submit as small PRs with clean history.
If approved, I can post the first reviewable PR update within ~48 hours.
- wire core
hey, id like to take this on - ive been looking at the VNGraphs.jl code and the remaining Graphs.jl API gaps look manageable. happy to start with edges/neighbors then work through the interface checker.
Hi @Krastanov, I would like to take this one if it is still free.
Background, since you have asked others for it: I work mainly in SciML, with 86 merged PRs in OrdinaryDiffEq.jl covering solver internals, error control and the MRI-GARK multirate family. More relevant here, I recently built PETScDiffEq.jl, a Julia wrapper over PETSc's C API, so the Julia to C side is what I have been living in lately. Graphs.jl itself I know as a user rather than a contributor.
Rather than claim it cold, I would start where you pointed naxatra2: take the smallest TODO item as its own PR so you can see the code before the bounty is committed to me. From the list that looks like completing the AbstractGraph API on VNGraph types together with the GraphsInterfaceChecker.jl test wiring, as one small PR with clean history. If that reads right to you I will open it this week.
Following up on the above with what I have actually opened, and with one thing addressed directly, since it is the fair worry when a new name turns up with several PRs at once.
VNGraphs.jl#33 fixes
has_edge, #34 addsoutneighborsandinneighbors, #35 addsedges, #36 completes the mutating methods, #37 addsVNAlgorithmwithclique_numberdispatched to the C routine, and Graphs.jl#529 declareschromatic_numberandedge_chromatic_numberas private method-less functions, following your review on #507 point by point. They are one concern per PR rather than one branch because that is what you asked for on VNGraphs.jl#26, and each is independent except #35, which needs #34.Every claim in them is measured rather than asserted. Each has a negative control showing the new test fails without the change, agreement is checked against the
SimpleGraphconversion rather than against my own expectations, and where I claim a performance difference I give the numbers, including the case wherevery_nautyloses to Graphs.jl and I say so. Two of them turned up real defects I had not gone looking for:has_edge(g, e)recursing into aStackOverflowError, andadd_edge!writing past the end of the adjacency array for an out-of-range vertex and reporting success. Where my work overlaps an open PR of someone else's, #14 and #26, I said so in my own PR rather than leave you to find it.On the review burden specifically: I am a GSoC 2026 student with SciML under the NumFOCUS umbrella, with 125 merged PRs across the SciML organisation, 104 of them in OrdinaryDiffEq.jl, so responding to review is the normal shape of my week rather than something I disappear from. If any of these need changes I will turn them around in hours, and I am equally happy for you to close mine in favour of #14 wherever the two overlap. I would rather the package end up right than the PR be mine.
A reliable idiomatic
very_nautywrapper [$400]The JuliaGraphs/VNGraphs.jl package already exists and provides a simple low-level wrapper for the C library. A lot of this work probably will not be directly in the Graphs.jl repository.
Graphs.some_interesting_property(::AbstractGraph)there should now be a new method defined in VNGraphssome_interesting_property(g::AbstractGraph, ::VNAlgorithm)that dispatches to the C implementation. It should convert thegargument to VNGraph if necessary.Required skills: familiarity with the Graphs.jl API and understanding of the Julia-C interface
Reviewer: any Graphs.jl member with merge rights
Duration: 3 months
Payout procedure:
The Funding for these bounties comes from the National Science Foundation and from the NSF Center for Quantum Networks. The payouts are managed by the NumFOCUS foundation and processed in bulk once every two months. If you live in a country in which NumFOCUS can make payments, you can participate in this bounty program.
Click here for more details about the bug bounty program.
Bug bounty logistic details (click to expand)
To claim exclusive time to work on this bounty either post a comment here or message skrastanov@umass.edu with:
If you want to, you can work on this project without making a claim, however claims are encouraged to give you and other contributors peace of mind. Whoever has made a claim takes precedence when solutions are considered.
You can always propose your own funded project, if you would like to contribute something of value that is not yet covered by an official bounty.