Phase 6 — File I/O & Export

[hypercompetent]

Part of the PyQt → Dash/Plotly migration. See cytoflowweb/PLAN.md for full context.

Estimated effort: 1 week
Depends on: Phase 1 (#1)

Tasks

• dcc.Upload component for FCS file upload
• dcc.Upload component for .flow workflow file upload
• Save workflow via API (keep existing YAML .flow format from cytoflowgui/workflow/serialization.py)
• Load workflow via API
• Export Jupyter notebook (nbformat) via dcc.Download
• Export FCS (ExportFCS view) via dcc.Download

Notes

• The .flow YAML serialization format must remain backward-compatible so existing workflows open in both the Qt GUI and the web UI.
• Notebook export reuses the existing get_notebook_code() implementations from each op/view plugin.
• Large FCS uploads (>50MB): parse server-side and store; only send metadata (channel names, conditions, event count) back to the browser.

Milestone

Round-trip save and load of a .flow workflow file. FCS and notebook export both produce valid downloadable files.