cheetahR 0.4.0: editable Shiny tables get an autocomplete column

cheetahR 0.4.0 adds an autocomplete column for editable Shiny tables: it filters a list of known suggestions as the user types, while still letting them commit a value that isn’t on the list. The release also ships a get_grid_data() Shiny helper and fixes a renderCheetah() re-render bug.

Grab it from CRAN:

install.packages("cheetahR")

The autocomplete column

Editing values in a Shiny grid usually means picking between two extremes. Plain text input is fast to wire up but offers no hint about what the user should enter. A closed dropdown shows the valid options but turns unwieldy past about twenty entries. The new autocomplete column in cheetahR 0.4.0 lands somewhere in between. It filters a list of known suggestions as the user types, and still lets them commit a value that isn’t on the list when they need to.

You declare it on column_def() with action = "autocomplete" and pass the pool of valid suggestions via auto_complete_opts. The underlying data column has to be of class character:

library(cheetahR)

countries <- c("Argentina", "Brazil", "Canada", "France", "Germany",
               "Ghana", "India", "Italy", "Japan", "Mexico", "Nigeria",
               "Spain", "United Kingdom", "United States")

data <- data.frame(
  id = 1:6,
  country = c("France", "Germany", "Ghana", "India", "Italy", "Japan"),
  stringsAsFactors = FALSE
)

cheetah(
  data,
  editable = TRUE,
  rownames = FALSE,
  columns = list(
    country = column_def(
      action = "autocomplete",
      auto_complete_opts = countries
    )
  )
)

Once a user double-clicks a cell:

• Typing filters the suggestion list (case-insensitive, matches anywhere in the word).
• The dropdown shows at most 10 matches at a time so it stays snappy on big pools.
• ↑/↓ moves through the highlights; Enter or Tab commits the highlighted suggestion (or whatever they typed, if nothing is highlighted); Escape cancels.
• Setting keyboard_options = list(moveCellOnTab = TRUE) on cheetah() makes Tab also advance to the next cell.

The cell fires the usual CHANGED_VALUE event when it commits, so the new value flows into input$<id>_changed_value like any other editable column.

Also new in 0.4.0

Two smaller pieces shipped alongside autocomplete:

• get_grid_data() is a new Shiny helper that returns the full grid contents as a reactive data frame. Easier than reassembling the table from individual _changed_value events.
• A render bug (#51) is fixed: renderCheetah() no longer stacks a new grid on top of the old one when its reactive deps change.

Pairing it with other column types

The autocomplete column works alongside cheetahR’s other interactive columns. A natural pattern is to mix it with an inline_menu (for a small, closed pick list) and a numeric column with a currency formatter:

columns = list(
  country = column_def(
    action = "autocomplete",
    auto_complete_opts = countries
  ),
  region = column_def(
    column_type = "menu",
    action = "inline_menu",
    menu_options = list(
      Africa = "Africa", Americas = "Americas",
      Asia = "Asia", Europe = "Europe", Oceania = "Oceania"
    )
  ),
  budget = column_def(
    name = "Budget",
    column_type = number_format(style = "currency", currency = "USD")
  )
)

The currency formatting comes from cheetahR’s number_format() helper, which wraps the browser’s Intl.NumberFormat.

Try it out

install.packages("cheetahR") to grab 0.4.0. Full release notes are in NEWS.md. Bug reports and feature requests are welcome on the issue tracker.