rs-grid
  • English

      • Virtual Data (FnDataSource)

      Overview

      FnDataSource generates cell values on demand via a closure. It uses O(1) memory regardless of the row count — ideal for computed data, demos, or very large datasets.

      Creating

      use rs_grid_core::datasource::FnDataSource;

let data = FnDataSource::new(1_000_000, |row, col_key| {
    match col_key {
        "id" => Some(format!("{row}")),
        "name" => Some(format!("User {row}")),
        "value" => Some(format!("{:.2}", row as f64 * 1.23)),
        _ => None,
    }
});

let model = GridModel::with_data_source(
    columns,
    Box::new(data),
    32.0,
    36.0,
);

      Properties

      PropertyValue
      MemoryO(1) — no data stored
      EditableNo — set_cell() is a no-op*
      CloneableNo — clone_box() returns None
      cell_statusAlways Ready or Absent (never Loading)
      Note

      *Editing still works via the patches layer on GridModel. The patch overrides the closure result for that cell. See Data Model.

      Sharing with Rc

      Since FnDataSource cannot be cloned, wrap the GridModel in Rc if you need to share it:

      use std::rc::Rc;
use std::cell::RefCell;

let model = Rc::new(RefCell::new(GridModel::with_data_source(
    columns,
    Box::new(data),
    32.0,
    36.0,
)));

      When to use

      • Very large datasets (millions or billions of rows)
      • Deterministic/computed data (hashes, formulas)
      • Demo and testing scenarios
      • When memory is constrained

      When NOT to use

      • When you need mutable data without patches
      • When you need to clone the data source
      • Server-side data with async loading — use PageCacheDataSource