Parse bibliographic references with anystyle

Usage

parse_ref(
  input,
  output_dir = NULL,
  format = c("json", "bib", "csl", "xml"),
  parser_model = NULL,
  overwrite = FALSE
)

Arguments

input

Either a character vector of references (one reference per element), or paths to text files holding one reference per line. input is taken as files when all its elements are paths of existing files.

output_dir

Directory where anystyle writes the output (one file per input file, named after it; references.<format> when input holds references). If NULL (the default), the output is returned.

format

Output format: "json" (the default), "bib", "csl" or "xml" (anystyle's tagged XML, the format used to train models).

parser_model

Path to a custom anystyle parser model, or NULL for the model shipped with anystyle.

overwrite

Logical. Overwrite existing files in output_dir? Default FALSE.

Value

If output_dir is NULL, a character vector with one element per input file (named after it), or a single string when input holds references. Otherwise, invisibly, the paths of the written files.

Details

Runs anystyle parse, which splits each reference into its fields (authors, date, title, journal, volume, pages...). anystyle reads one reference per line. Use parse_ref_to_df() to get a tidy data frame directly.

Line breaks inside a reference are replaced by spaces and empty references are dropped before parsing: otherwise anystyle would read them as several references, or skip them, and the output would not match the input. Note that anystyle replaces a ditto mark (e.g. "———" for "same author") with the authors of the previous line: keep the references of a document together and in order.

See also

parse_ref_to_df(), find_ref()

Examples

if (FALSE) { # \dontrun{
parse_ref("Doe, J. (2020). Title of the article. Journal Name, 13(2), 123-456.")

txt <- system.file("extdata", "example_doc_1.txt", package = "anystyleR")
parse_ref(txt, format = "bib")
} # }