ENH: wire the JupyterLite build into the docs (JupyterLite split 5/5) - #14157
Open
natinew77-creator wants to merge 31 commits into
Open
ENH: wire the JupyterLite build into the docs (JupyterLite split 5/5)#14157natinew77-creator wants to merge 31 commits into
natinew77-creator wants to merge 31 commits into
Conversation
Enables jupyterlite_sphinx, stages the data the browser notebooks read, prepends the setup cell to each notebook copy, and leaves the launch badge off the pages that cannot run in the browser.
It sits with "Get data and triage examples to run" rather than after test-doc. Same effect, since all it has to do is precede make html, and it keeps the step away from the wheel build's insertion point.
natinew77-creator
marked this pull request as ready for review
August 17, 2026 20:18
natinew77-creator
requested review from
agramfort,
drammock and
larsoner
as code owners
August 17, 2026 20:18
find_wheels() returns Path objects since the wheel build moved to pathlib,
so the log line was rendering as [PosixPath('...')].
# Conflicts: # doc/documentation/datasets.rst
The tutorial crops the white margins off brain.screenshot(), and vtk.js cannot hand a framebuffer back to Python, so there is no screenshot to crop.
Two entries had gone stale: upstream renamed decoding_rsa_sgskip.py and montage_sgskip.py, so both examples were badged again despite still needing the datasets they were excluded for. Also excludes the new 21_interactive_dipole_fit, which is a tour of mne.gui.dipolefit and needs a picker vtk.js does not provide.
9 tasks
Stage the served data from a sphinxext module only when MNE_DOC_BUILD_JUPYTERLITE=1, which make html sets, so pattern and noplot builds skip the copy and the lite build; skip files already staged; drop the exclusions for the pages that now run through the backend; fix stale text.
teonbrooks
reviewed
Sep 11, 2026
Comment on lines
+279
to
+285
| # Ensure the data the JupyterLite notebooks need is on disk so conf.py can | ||
| # copy the required subset for the build. lite_data fetches only the | ||
| # curated files the browser notebooks use (from the MNE-lite-data OSF | ||
| # project) instead of the full sample/kiloword/erp_core/mtrf/eegbci | ||
| # datasets, which slims the build. The curated files share hashes with the | ||
| # full datasets, so on a full build that already downloaded them nothing | ||
| # extra is fetched. |
Member
There was a problem hiding this comment.
I would shorten this to say Download MNE-lite-data to run JupyterLite notebooks or something more concise.
teonbrooks
reviewed
Sep 11, 2026
Comment on lines
+45
to
+59
| "forward/20_source_alignment.ipynb", | ||
| "mne.gui.coregistration", | ||
| "**This cell does not run in the browser.**\n" | ||
| "\n" | ||
| "`mne.gui.coregistration` sets the fiducials by clicking on the scalp\n" | ||
| "surface, and the vtk.js renderer used here draws scenes without a\n" | ||
| "picker, so there is nothing for those clicks to hit. Run it from a\n" | ||
| "local MNE install instead:\n" | ||
| "\n" | ||
| "```python\n" | ||
| 'mne.gui.coregistration(subject="sample", subjects_dir=subjects_dir)\n' | ||
| "```\n" | ||
| "\n" | ||
| "The video above walks through the same steps, and the rest of this\n" | ||
| "notebook runs normally.\n", |
Member
There was a problem hiding this comment.
you can use """ to avoid needing linebreaks
teonbrooks
reviewed
Sep 11, 2026
Comment on lines
+497
to
+504
| # Inject the required subset of MNE-sample-data for JupyterLite. The data is | ||
| # placed under doc/lite_extra/mne_data and served at the docs root via | ||
| # html_extra_path (added below). The JupyterLite setup cell fetches these | ||
| # files over HTTP into the Pyodide kernel: the /drive virtual-filesystem | ||
| # bridge needs cross-origin-isolation (COOP/COEP) headers that static | ||
| # artifact servers (e.g. CircleCI) do not send, so it is unusable there. | ||
| # lite_data (mne.datasets.lite_data) extracts the curated subset here, with the | ||
| # files under their original dataset folders (MNE-sample-data/, ...). |
teonbrooks
reviewed
Sep 11, 2026
| f"{(mne_data_base / 'MNE-sample-data').exists()}, " | ||
| f"curated archive={src_sample_data.exists()}" | ||
| ) | ||
| if (mne_data_base / "MNE-sample-data").exists() or src_sample_data.exists(): |
Member
There was a problem hiding this comment.
is there a way to get a list of files from mne-lite-data instead of listing them below in the required_files? it would make this DRYer if we didn't need to list them again here
teonbrooks
reviewed
Sep 11, 2026
| lite_data_base.mkdir(parents=True, exist_ok=True) | ||
|
|
||
| src_ssvep = mne_data_base / "ssvep-example-data" | ||
| dst_ssvep = lite_data_base / "ssvep-example-data" |
Member
There was a problem hiding this comment.
is there a reason that the lite_data doesn't ship with the ssvep or eeglab data?
* upstream/main: (25 commits) ENH: add the JupyterLite notebook setup cell (JupyterLite split 4/5) (mne-tools#14150) FIX: Ver [circle deploy] [skip azp] [skip actions] Hotfix fix EDF round-trip (mne-tools#14296) Brain GUI modernization (Phase 7) (mne-tools#14270) Fix bug with SciPy 1.18.0 EEGLAB reading (mne-tools#14293) remove plotting abs as the defauls for plotting volumetric (mne-tools#13989) Coregistration GUI modernisation (phase 4) (mne-tools#14285) Dipole GUI interactivity and visual enhancements (phase 2) (mne-tools#14291) Width [circle deploy] [skip azp] [skip actions] Post-release deprecations (mne-tools#14290) FIX: Come on [circle deploy] [skip azp] [skip actions] PyVista try again [circle deploy] [skip azp] [skip actions] FIX: PyVista [circle deploy] [skip azp] [skip actions] FIX: Really [circle deploy] [skip azp] [skip actions] FIX: Doc Release commit [circle deploy] MAINT: Update code credit (mne-tools#14289) MAINT: Update pre-commit hook versions (mne-tools#14288) BUG: use fixed normalized CTPS cutoff for automatic ECG detection (mne-tools#14279) ...
Member
|
I am about to push changes now that #14150 was merged. I want to slim down the files that are used since it's currently 1.2 GB, which is a lot. |
* upstream/main: No PyQt5 [circle deploy] [skip azp] [skip actions]
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
3 participants
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.
Part 5 of the split of #13925, and the last one. Parts 1 to 4 are #14128, #14135, #14144 and #14150.
Wires the previous four together: enables
jupyterlite_sphinx, serves the data subset the browser notebooks read, and prepends the setup cell to each notebook copy. It also carries the list of pages that do not get an "Open in JupyterLite" badge, since sphinx-gallery adds one to every example. Only the badge is dropped, the notebook source is untouched.build_docsis red until #14144 and #14150 land, anddefaultjust mirrors it. Parts 1 and 2 have merged, so the data and wheel steps now pass and it stops inconf.py:jupyterlite_cell_notes.pyimportsjupyterlite_setup_cell, which #14150 adds, and that in turn reads the renderer from #14144. Every other check is green, since nothing but Sphinx readsconf.py.Checked on #13925, where the same wiring builds with every page executed and no sphinx warnings.