Query documents from a krpoltext corpus

Loads a dataset if needed, applies named filters, and optionally selects a subset of columns.

Usage

get_docs(
  dataset = c("campaign_booklet", "party_statements"),
  ...,
  .data = NULL,
  .select = NULL,
  .strict = FALSE
)

Arguments

dataset

Character; one of "campaign_booklet" or "party_statements".

...

Named filtering arguments. Values can be scalars or vectors.

.data

Optional pre-loaded data. If NULL, the requested dataset is loaded with the package defaults.

.select

Optional character vector of columns to keep.

.strict

Logical; if TRUE, invalid filter or selection columns raise an error. If FALSE (default), invalid names are ignored with a message.

Value

A data.table subset of the requested corpus.

Examples

dt <- data.table::data.table(
  year = c(2020L, 2021L),
  title = c("A", "B"),
  text = c("first", "second"),
  conservative = c(1L, 0L),
  id = c("ps-1", "ps-2")
)

get_docs("party_statements", year = 2020, .data = dt)
#>     year  title   text conservative     id
#>    <int> <char> <char>        <int> <char>
#> 1:  2020      A  first            1   ps-1
get_docs(
  "party_statements",
  conservative = 1,
  .select = c("year", "title"),
  .data = dt
)
#>     year  title
#>    <int> <char>
#> 1:  2020      A