Repository navigation
Conversation
|
@copilot please re-review |
Re-review finding (P2): the new No files changed. Targeted tests could not run because this environment lacks both |
render_embedding_atlas, render_mosaic_widget, and render_quak didn't accept the new listen_to kwarg (TypeError), and render_mols2grid swallowed it in **kwargs, routing it to widget options instead of the accessor validation. Each now takes listen_to explicitly and forwards it to render_widget_base.
|
@copilot please re-review |
|
@cpsievert Unfortunately I hit an unexpected error while processing your comment. I've automatically reported this to GitHub. You can ask me to try again later by mentioning me in a new comment. If you want to contact GitHub about this error, please mention the following identifier so they can better serve you: Sorry for the inconvenience! |
Re-reviewed all four commits. The constructor forwarding issue is resolved in c668733, with regression tests; no new actionable findings. CI for c668733 passed all nine jobs. The Python 3.11 log confirms 1,154 tests passed (3 skipped, 1 expected failure). Local tests remain unavailable because |
Split out of #480; infrastructure for the plotly PR stacked on top of this one.
AccessorSpec.opt_inmarks accessors whose payloads are unbounded or stream continuously; the renderer skips attaching them at render time, and reading one raises an error pointing atlisten_to, unless declared (@render_*(listen_to=["selected"])).listen_tois additive to the always-on accessors, and unknown names raise at decoration time. No accessor is opt-in yet, so this changes no user-visible behavior on its own.