Collect and apply the column contract

Like dplyr::collect() on a lazy table, but additionally applies the package's column contract: epoch-millisecond timestamps become POSIXct (UTC), JSON-in-TEXT columns (e.g. transcriptions.segments) become list-columns, and BLOB columns stay raw. Columns the manifest does not describe are returned unchanged.

Usage

entropia_collect(x, n = Inf, ...)

Arguments

x

A lazy table, e.g. from entropia_items().

n

Maximum number of rows to fetch, passed to dplyr::collect().

...

Additional arguments passed to dplyr::collect().

Value

A tibble::tibble() with the column contract applied.

Details

The contract is resolved from the table's base table. For a lazy query that is not a simple single-table read (joins, subqueries) the columns are returned as SQLite produced them – keep filters/selects inside the lazy query for best results.

Examples

con <- entropia_connect(system.file("extdata", "entropia-example.sqlite",
  package = "entropiaR"
))
entropia_collect(entropia_items(con)) # created_at -> POSIXct, metadata -> list-column
#> # A tibble: 3 × 7
#>   id    title collection_id metadata     created_at          updated_at         
#>   <chr> <chr> <chr>         <list>       <dttm>              <dttm>             
#> 1 2222… Mani… 11111111-111… <named list> 2026-01-15 12:01:00 2026-01-15 12:02:00
#> 2 2222… Cart… 11111111-111… <named list> 2026-01-15 12:03:00 2026-01-15 12:04:00
#> 3 2222… Foto… 11111111-111… <chr [1]>    2026-01-15 12:05:00 2026-01-15 12:06:00
#> # ℹ 1 more variable: search_text <chr>
entropia_disconnect(con)