Fix: block pipeline node switch during active edit session + Perf: edit session remap cache (B3, B4)

CLAUDE.md

@@ -143,12 +143,12 @@ Before opening a PR, check [TODO.md](TODO.md) for related open items to update o
 - `runtime_setup.py`: `RuntimeContext` dataclass and `create_runtime_context()` — ParaView objects created before Trame is available. `ParaViewRuntime` is constructed in `factory.py` once `state` and `view_controls` exist.
 - `handler_registration.py`: wires controllers and state handlers.
 - `state_setup.py`: initializes all Trame state. If adding UI controls, add defaults here.
-- `state_handlers.py`: shared `@state.change(...)` callbacks for selected file, color-by, representation, active pipeline node, interaction quality, edit mode.
+- `state_handlers.py`: shared `@state.change(...)` callbacks for selected file, color-by, representation, active pipeline node, interaction quality, edit mode. Pipeline node switches are blocked during an active edit session — the handler emits a warning and resets `state.active_pipeline_item` to snap the UI selection back.
 - `view_controls.py`: central wrapper for Trame view update callbacks and optional ParaView diagnostics.
 
 ### ParaView Backend Path
 
-- `paraview_backend.py`: owns ParaView sources/displays, pipeline nodes, filters, coloring, display controls, picking, selection overlays, saving/export.
+- `paraview_backend.py`: owns ParaView sources/displays, pipeline nodes, filters, coloring, display controls, picking, selection overlays, saving/export. `set_edit_target_dataset()` builds four remap caches at session begin (`_edit_source_dataset`, `_edit_source_point_indexes`, `_edit_target_cell_map`, `_edit_point_id_map`) so picks avoid per-pick Fetch/rebuild — see `docs/logic_flows.md §5b`.
 - `paraview_runtime.py`: synchronizes backend state into Trame state, handles render pushes, edit-session overlay sync. Forwards event normalization calls to `paraview_event_utils.py`.
 - `paraview_event_utils.py`: pure helpers for normalizing ParaView picking event payloads (`normalize_edit_selection_ids`, `summarize_edit_event`, coordinate mapping). No Trame dependency.
 - `paraview_controllers.py`: user actions from the UI: pipeline actions, filters, edit sessions, selection, field creation, display/color controls.
@@ -201,6 +201,12 @@ Do not implement replace as a toggle. ParaView native payloads can contain toggl
 
 Selection overlays can interfere with native picking. `paraview_backend.py` clears transient edit-selection overlays before pick queries.
 
+### Pipeline Lock During Edit Session
+
+`state_handlers.py` blocks `active_pipeline_item` changes while an edit session is active. When blocked, it emits a warning notification and resets `state.active_pipeline_item` to `pv_backend.active_node_id`, which snaps the pipeline tree UI back to the current node. The reset triggers a second `@state.change` event that is a no-op (Trame suppresses events when the value is unchanged), so there is no infinite loop.
+
+No pre-change hook exists in Vuetify/Trame — `VListItemGroup v_model` writes state immediately on click — so interception must happen inside the callback after the write.
+
 ### EditSession Geometry Modes
 
 `edit_session.py` supports three modes selected per-field:

TODO.md

@@ -2,9 +2,17 @@
 
 ## Priorita alta
 
+- **Performance selezione su dataset grandi** — bottleneck documentati in
+  `docs/logic_flows.md` §5a e §6. Piano a step progressivi in [issue #34](https://github.com/2listic/coral-visualizer/issues/34); ogni step
+  è indipendente e lascia i test esistenti verdi. Step 1 può essere fatto direttamente
+  su `paraview_backend.py` senza refactor preventivo.
+
 - Separare `paraview_backend.py` in servizi di dominio più piccoli (3900 righe).
-  Iniziare con `backend/selection.py` + `backend/selection_geometry.py`.
-  Vedi [docs/backend_refactor.md](docs/backend_refactor.md) per strategia, layout moduli e firme.
+  Iniziare da moduli con dipendenze unidirezionali (`backend/display.py`,
+  `backend/coloring.py`, `backend/pipeline.py`) — più self-contained e a minor rischio
+  di regressione rispetto alla selezione, che dipende trasversalmente dallo stato del
+  backend. La selezione segue dopo. Vedi [docs/backend_refactor.md](docs/backend_refactor.md)
+  per strategia, layout moduli e firme.
 
 ## Priorita media
 
@@ -20,10 +28,7 @@
 
 ## Miglioramenti futuri
 
-- **Performance selezione su dataset grandi**: identificare il bottleneck (pick
-  query, overlay rebuild, state sync). Da affrontare dopo aver estratto e
-  compreso `backend/selection.py`.
-- **Riorganizzazione UI/UX**: migliorare l'esperienza utente dopo aver spezzato
+- **Riorganizzazione UI/UX**: migliorare l'esperienza utente eventualmente dopo aver spezzato
   `ui.py` in moduli navigabili.
 
 ## Riorganizzazione cartelle