Load one A2AJ dataset as a data.frame — morie_ingest_a2aj

Downloads (or reuses) the dataset's Parquet file and decodes it. The unofficial_text_* columns hold the full texts and dominate the file, so pass columns to skip them when only metadata or the citation network is needed.

Usage

morie_ingest_a2aj_load(
  dataset,
  doc_type = c("cases", "laws"),
  columns = NULL,
  engine = c("auto", "native", "arrow"),
  cache_dir = NULL,
  refresh = FALSE
)

Arguments

dataset: Dataset code, for example "SCC".
doc_type: "cases" (default) or "laws".
columns: Optional character vector of columns to decode.
engine: "auto" (default: arrow when installed, otherwise the native reader), "native" or "arrow". The native reader is pure R and decodes the LIST-typed citation columns; it is fine for the smaller courts and slow for the largest files, which is where arrow earns its place.
cache_dir, refresh: Passed to morie_ingest_a2aj_download.

Value

A data.frame; cases_cited_* and cases_citing_* are list columns with one character vector per row.

Examples

# \donttest{
sct <- try(morie_ingest_a2aj_load("SCT",
  columns = c("citation_en", "document_date_en", "cases_cited_en")
))
if (is.data.frame(sct)) head(sct$citation_en)
#> [1] "2016 SCTC 11" "2014 SCTC 8"  "2016 SCTC 15" "2015 SCTC 3"  "2017 SCTC 1" 
#> [6] "2016 SCTC 10"
# }