Repository navigation
docs: 📝 replace the MkDocs site with a Fumadocs site - #56
Merged
Merged
Conversation
Contributor
Port the Next.js + Fumadocs setup from monoprop to docs/: MDX pages for getting started, concepts, features, benchmarks, references and the developer guide; an API reference generated from the docstrings with griffe; tutorial notebooks executed at build time; KaTeX math and BibTeX citations. The Documentation workflow builds the static export, runs the Python examples in the pages with pytest --markdown-docs, and deploys to GitHub Pages under /src-method. Closes #32 Assisted-by: pi:claude-opus-5.5
Assisted-by: pi:claude-opus-5.5
Panadestein
force-pushed
the
docs/32-fumadocs-site
branch
from
October 6, 2026 13:32
edd7a74 to
8a454cc
Compare
Panadestein
added a commit
that referenced
this pull request
Oct 7, 2026
Resolve the conflicts with stdlib logging (#52), input-precision sketches (#36), validation (#38), ty enforcement (#37, #55) and the Fumadocs site (#56): - the sweep draws its sketches with gaussian_sketch and logs its plan, pass times and stalls at DEBUG with %-style arguments; - the working dtype is promoted over every site, not just the first; - docs/large-problems.md moves to features/large-problems.mdx and the out-of-core testing notes to contributing/testing.mdx; - bench_large.py logs through the stdlib and gains --debug. Type lazily read sites: the entry points accept any array-like with shape, dtype, ndim and np.asarray support but were typed Sequence[NDArray]. Add the SiteLike protocol and the Site alias, use them from src/apply/compress down to the site source, and export SiteLike. Also fix the other ty findings: the Tier list in the planner, known_kind() for validated trains, padded_shape without a type: ignore, cupyx as an allowed unresolved import, and typed fakes in the tests. Assisted-by: pi:claude-opus-5.5
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
🤖 AI text below 🤖
Closes #32.
What
This ports the monoprop Next.js + Fumadocs setup to
docs/and removes MkDocs:mkdocs.yml,gen_ref_pages.py, the MathJax/CSS assets and themkdocs*packages in thedocsgroup.docs/src, configs,package.json/package-lock.json): copied from monoprop. It keeps the Algorithmiq branding: the GT-Planar font, the teal accent and the logo. Changes for this repo:basePathandtrailingSlashare set, because the site is served atalgorithmiq.github.io/src-method/.basePathcomes fromNEXT_PUBLIC_BASE_PATHand is empty fornpm run dev. The static search index and the "Copy Markdown" URLs get the prefix too.metadataBaseis set for the Open Graph images.docs/content/docs/):cutoff, precision and seeding, GPU, logging.benches/.docs/developer-guide/*.mdpages were template leftovers that mentioned "aurora", setuptools-scm and pytest plugins we don't use. I rewrote them for this repo.gen_api_dump.py(griffe → JSON) andgenerate-api.mjs(JSON → MDX) generate it from the docstrings ofstack,applyandcompress. Compared with monoprop's scripts:Raises:sections are rendered instead of dropped.*argsno longer show= ()and pick up their*namedocstring entry./after*args.<and{are no longer escaped inside code spans, where the escape showed up as\<psi|.logger/constants are hidden.notebooks_to_mdx.pyexecutes eachdocs/notebooks/<name>/<name>.ipynb; any cell error fails the build. There is one tutorial, Accuracy versus bond dimension, which compares SRC against contract-then-SVD aschi_outgrows and runs in about 10 s.[@key]links pointed at a#bib-…anchor on the current page, which only exists on the references page (monoprop has the same bug). A small rehype pass now points them at/references#bib-….docpages.yml): Python + Node 22 → generate the API → run every Python block in the prose pages withpytest --markdown-docs(blocks that can't run standalone are markedpython notest) → execute the notebooks →npm run build→ deploydocs/outto Pages onmain.docsgroup is nowgriffelib,nbconvertandpytest-markdown-docs, plustestandinteractive.README.md,AGENTS.mdanddocs/README.mdexplain the local preview.Checks
NEXT_PUBLIC_BASE_PATH=/src-method.next buildexports 81 static pages.tsc --noEmitis clean.out/under/src-method/and checked in headless Chromium:/api/searchand the markdown routes all resolve under the prefix.uv run prek run --all-filesis clean.Notes and follow-ups
docs/developer-guide/dependencies.md, which this PR moves todocs/content/docs/contributing/dependencies.mdx. Whichever merges second should port that section, plus thebenchgroup from chore(deps): 🔧 move benchmark-only dependencies to a bench group #53.lycheelink check or the Cloudflare PR previews. The base path breaks lychee's root-relative resolution unless it's remapped, and the previews need Cloudflare secrets. I can add either if wanted.