Extracts scripts from an html page

Usage

cas_extract_script(
  html_document,
  script_type = NULL,
  match = NULL,
  accessors = NULL,
  remove_from_script = NULL
)

Arguments

html_document: An html document parsed with xml2::read_html() or rvest::read_html().
script_type: Defaults to NULL. Type of script. Common script types include application/ld+json, text/template, etc.
match: Default to NULL. If given, used to filter extracted scripts. Must be a named vector in the format c(@type = "NewsArticle") for a script of type "NewsArticle".
accessors: Defaults to NULL. If given, a vector of accessors passed to purrr::pluck in order to extract sub-components of the list resulting from reading the with jsonlite the result of the previous steps and filter.
remove_from_script: Defaults to NULL. If given, removed after the script has been extracted but before processing the json.

Value

May return a list or a character vector. If no match is found, returns NA_character_

Examples

if (FALSE) { # \dontrun{
if (interactive()) {
  url <- "https://www.digi24.ro/stiri/externe/casa-alba-pune-capat-isteriei-globale-nu-exista-indicii-ca-obiectele-zburatoare-doborate-de-rachetele-sua-ar-fi-extraterestre-2250863"

  html_document <- rvest::read_html(x = url)

  cas_extract_script(
    html_document = html_document,
    script_type = "application/ld+json"
  )

  # get date published
  cas_extract_script(
    html_document = html_document,
    script_type = "application/ld+json",
    match = c(`@type` = "NewsArticle"),
    accessors = "datePublished"
  )

  # get title
  cas_extract_script(
    html_document = html_document,
    script_type = "application/ld+json",
    match = c(`@type` = "NewsArticle"),
    accessors = "headline"
  )

  # get nested element, e.g. url of the logo of the publisher

  cas_extract_script(
    html_document = html_document,
    script_type = "application/ld+json",
    match = c(`@type` = "NewsArticle"),
    accessors = c("publisher", "logo", "url")
  )
}
} # }