spaday-regular-table

Viewport-virtualized data tables for spaday, powered by regular-table.

Documentation

• Build a 100,000-row table — guided virtual-table example.
• Add rich action cells and hot row updates — task-focused Python configuration.
• API reference — props, events, methods, and metadata.
• Why viewport virtualization matters — performance model and tradeoffs.

Quick example

from spaday import CallEndpoint, event_value, serve
from spaday_regular_table import RegularTable

rows = [{"id": i, "symbol": f"SYM{i}", "price": i / 10} for i in range(100_000)]

table = RegularTable(
    columns=[
        "id",
        {
            "key": "symbol",
            "label": "Symbol",
            "cell": {"tag": "button", "class": "symbol-button", "event": "symbol-click"},
        },
        {"key": "price", "label": "Last price", "cell": {"format": "number", "digits": 2}},
    ],
    rows=rows,
    row_header="id",
    style="height: 32rem",
).on("symbol-click", CallEndpoint("POST", "/api/click", event_value()))

serve(table, packages=["regular-table"])

Only cells requested for the current viewport are copied into regular-table and rendered. rows remains a browser-side record array. columns accepts keys or {key, label, cell} objects and is inferred from the first row when omitted. A cell descriptor can wrap the displayed value in an HTML or custom-element tag, apply classes and attributes, or format numeric values. The wrapper performs that work only for visible cells.

For hot server pushes, set stream_url to an SSE endpoint whose messages are row patches. You can also bind rowPatch to CallEndpoint(result=...) state for request/response updates. Each patch contains revisioned update, insert, and remove operations, so a 100,000-row list is not recopied. The browser methods remain available to integration authors, but applications need no JavaScript callback.

cell-click exposes event.detail with type, virtual x/y, column, value, and the source row, so normal spaday actions can handle interactions. virtual_mode supports upstream both, horizontal, vertical, and none; keep both for large datasets.

Run the local example

python -m pip install -e ".[examples]"
python -m spaday_regular_table.example

Open http://127.0.0.1:8014 to inspect the complete market-blotter example: 100,000 virtualized rows, a server row stream, full-width sizing, rich cells, and server-authoritative update, insert, remove, and click operations. All interaction is authored in Python; there is no companion example.js. It passes the local package descriptor directly, so it does not install or resolve the integration from GitHub.