Merge pull request #6 from chrbknudsen/main

gammelt materiale bevaret

.Rbuildignore

@@ -6,3 +6,4 @@
 ^_pkgdown\.yml$
 ^docs$
 ^pkgdown$
+^skrammel$

skrammel/DESCRIPTION

@@ -0,0 +1,19 @@
+Package: KBAPI
+Type: Package
+Title: What the Package Does (Title Case)
+Version: 0.1.0
+Author: Who wrote it
+Maintainer: The package maintainer <yourself@somewhere.net>
+Description: More about what it does (maybe more than one line)
+    Use four spaces when indenting paragraphs within the Description.
+License: What license is it under?
+Encoding: UTF-8
+LazyData: true
+RoxygenNote: 7.2.0
+Suggests: 
+    testthat (>= 3.0.0)
+Config/testthat/edition: 3
+Imports: 
+    dplyr,
+    magrittr,
+    rvest

skrammel/NAMESPACE

@@ -0,0 +1,10 @@
+# Generated by roxygen2: do not edit by hand
+
+export("%>%")
+export(get_authors)
+export(get_categories)
+export(get_collections)
+export(get_editions)
+export(get_genres)
+export(get_subcollections)
+importFrom(magrittr,"%>%")

skrammel/R/get_authors.R

@@ -0,0 +1,104 @@
+#' Get authors
+#'
+#' Retrieves a list of available authors in Text Collections
+#'
+#' @param live logical value indicating if list of authors should be retrieved live, or from internal data.
+#'
+#' @return None
+#'
+#' @examples
+#' get_authors()
+#'
+#' @export
+
+get_authors <- function(){
+        test <- jsonlite::read_json("https://api.kb.dk/data/rest/api/text?q=cat_ssi:author+AND+type_ssi:work&wt=json&start=0&rows=75&defType=edismax",
+                            simplifyVector = T)
+        test$response$docs$inverted_name_title_ssi
+}
+
+# det er ikke givet at det er den format der skal bruges til at søge.
+# Så her er nogle alternativer
+# test$response$docs[[2]]$work_title_tesim
+# test$response$docs[[2]]$sort_title_ssi
+# test$response$docs[[2]]$inverted_name_title_ssi
+
+
+
+# Nedenstående skal placeres i et internt datasæt, og hentes som alternativ
+# til at trække data fra apien. (for det tager tid!)
+fall_back <- c("Palladius, Peder"                 ,
+"Paludan-Müller, Frederik",
+"Ploug, Carl",
+"Møller, Poul Martin",
+"Richardt, Christian",
+"Saxo",
+"Schack, Hans Egede",
+"Schandorph, Sophus",
+"Sibbern, Frederik Christian",
+"Skjoldborg, Johan",
+ "Skram, Erik",
+ "Michaëlis, Sophus",
+ "Staffeldt, Schack",
+ "Steffens, Henrich",
+ "Sthen, Hans Christensen",
+ "Stub, Ambrosius",
+ "Stuckenberg, Viggo",
+ "Thomissøn, Hans",
+ "Topsøe, Vilhelm",
+ "Wessel, Johan Herman",
+ "Wied, Gustav",
+ "Winther, Christian",
+ "Worm, Jacob",
+ "Aakjær, Jeppe",
+ "Aarestrup, Emil",
+ "Ravnkilde, Adda",
+ "Arrebo, Anders",
+ "Bagger, Carl",
+ "Baggesen, Jens",
+ "Bang, Herman",
+ "Bergsøe, Vilhelm",
+ "Blicher, Steen Steensen",
+ "Bødtcher, Ludvig",
+ "Bording, Anders",
+ "Brahe, Tycho",
+ "Brandes, Georg",
+ "Brorson, Hans Adolph",
+ "Chievitz, Poul",
+ "Claussen, Sophus",
+ "Dalgas, Ernesto",
+ "Drachmann, Holger",
+ "Brandes, Edvard",
+ "Ewald, Johannes",
+ "Falster, Christian",
+ "Fibiger, Mathilde",
+ "Heiberg, Johanne Luise",
+ "Gjellerup, Karl",
+ "Goldschmidt, M. A.",
+ "Grundtvig, N. F. S.",
+ "Gyllembourg, Thomasine",
+ "Hauch, Carsten",
+ "Andersen, Hans Christian",
+ "Helie, Paulus",
+ "Hertz, Henrik",
+ "Hjortø, Knud",
+ "Holberg, Ludvig",
+ "Hostrup, Jens Christian",
+ "Rode, Helge",
+ "Ingemann, B. S.",
+ "Knudsen, Jakob",
+ "Jacobsen, Jørgen-Frantz",
+ "Heiberg, Johan Ludvig",
+ "Jacobsen, J. P.",
+ "Juel-Hansen, Erna",
+ "Kidde, Harald Henrik Sager",
+ "Kierkegaard, Søren",
+ "Kingo, Thomas",
+ "Kaalund, Hans Vilhelm",
+ "Larsen, Thøger",
+ "Larsen, Karl",
+ "Leonora Christina",
+ "Lyschander, Claus Christoffersen",
+ "Nansen, Peter",
+ "Oehlenschläger, Adam",
+ "Heiberg, Peter Andreas")

skrammel/R/get_categories.R

@@ -0,0 +1,30 @@
+#' Get categories
+#'
+#' FIXME
+#'
+#' @param live logical value indicating if list of authors should be retrieved live, or from internal data.
+#'
+#' @return None
+#'
+#' @examples
+#' get_categories()
+#'
+#' @export
+
+get_categories <- function(){
+  base_url <- "https://api.kb.dk/data/text"
+
+  test <- rvest::read_html(base_url)
+
+  res <- test |>
+    rvest::html_element("#cat_ssi") %>%
+    rvest::html_elements("option") %>%
+    rvest::html_attrs() %>%
+    unlist()
+  unname(res[nchar(res)>0])
+}
+
+# den kan nok også med fordel have en "non-live" option
+# og så kommer den ud som en named character vector.
+# det skal der nok gøres noget ved.
+

skrammel/R/get_collections.R

@@ -0,0 +1,32 @@
+#' Get collections
+#'
+#' FIXME
+#'
+#' @param live logical value indicating if list of authors should be retrieved live, or from internal data.
+#'
+#' @return None
+#'
+#' @examples
+#' get_collections()
+#'
+#' @export
+
+get_collections <- function(){
+  navn <- NULL
+  base_url <- "https://api.kb.dk/data/"
+
+  rvest::read_html(base_url) %>%
+  rvest::html_elements("body .main-nav a") %>%
+  purrr::map_dfr(~data.frame(
+    "navn" = rvest::html_attr(., "href"),
+    "beskrivelse" = rvest::html_text(.)
+  )) %>%
+  dplyr::mutate(link = paste0(base_url, navn))
+}
+# Hvis ovenstående fejler - returner denne:
+# navn         beskrivelse                        link
+# 1 dsfl  Aerial photographs https://api.kb.dk/data/dsfl
+# 2  cop Digital collections  https://api.kb.dk/data/cop
+# 3 text    Text collections https://api.kb.dk/data/text
+
+

skrammel/R/get_cop.R

@@ -0,0 +1,18 @@
+hent_have_data <- function(){
+  url <- "http://www5.kb.dk/images/billed/2010/okt/billeder/subject37894/da/?page="
+  data <- httr::GET(paste0(url,1), add_headers(Accept = "application/json")) %>%
+    content(type = "text", encoding = "UTF-8") %>%
+    fromJSON()
+  data <- data$response
+  sider <- data$pages$total_pages
+  res <- list()
+  res[[1]] <- data$docs %>% as.data.frame()
+  for(i in 2:sider){
+    res[[i]] <- httr::GET(paste0(url,i), add_headers(Accept = "application/json")) %>%
+      content(type = "text", encoding = "UTF-8") %>% fromJSON() %>%
+      .$response %>%
+      .$docs %>%
+      as.data.frame()
+  }
+  bind_rows(res)
+}

skrammel/R/get_editions.R

@@ -0,0 +1,19 @@
+#' Get editions
+#'
+#' FIXME
+#'
+#' @param live logical value indicating if list of authors should be retrieved live, or from internal data.
+#'
+#' @return None
+#'
+#' @examples
+#' get_editions()
+#'
+#' @export
+
+
+get_editions <- function(){
+  jsonlite::read_json("https://api.kb.dk/data/rest/api/editions",
+                            simplifyVector = T)
+}
+

skrammel/R/get_genres.R

@@ -0,0 +1,27 @@
+#' Get genres
+#'
+#' FIXME
+#'
+#' @param live logical value indicating if list of authors should be retrieved live, or from internal data.
+#'
+#' @return None
+#'
+#' @examples
+#' get_genres()
+#'
+#' @export
+
+
+
+get_genres <- function(){
+  base_url <- "https://api.kb.dk/data/text"
+
+  test <- rvest::read_html(base_url)
+
+  res <- test |>
+    rvest::html_element("#genre_ssi") |>
+    rvest::html_elements("option") |>
+    rvest::html_attrs() |>
+    unlist()
+  unname(res[nchar(res)>0])
+}

skrammel/R/get_subcollections.R

@@ -0,0 +1,24 @@
+#' Get subcollections
+#'
+#' FIXME
+#'
+#' @param live logical value indicating if list of authors should be retrieved live, or from internal data.
+#'
+#' @return None
+#'
+#' @examples
+#' get_subcollections()
+#'
+#' @export
+
+get_subcollections <- function(){
+test <- jsonlite::read_json("https://api.kb.dk/data/rest/api/text?q=&rows=0&facet=on&facetfield=subcollection_ssi",
+                            simplifyVector = F)
+test <- test$facet_counts$facet_fields$subcollection_ssi
+test <- test[seq(1, length(test), 2)]
+  unlist(test)
+
+
+# Med mindre ovenstående fejler - ellers returner:
+#   c("adl", "gv",  "tfs", "sks")
+}

skrammel/R/luftfotos.R

@@ -0,0 +1,36 @@
+# Installer httr2, hvis det ikke allerede er installeret
+# install.packages("httr2")
+
+# Load httr2 pakken
+library(httr2)
+
+# Definer API-endpointet
+api_url <- "http://www.kb.dk/cop/syndication/images/luftfo/2011/maj/luftfoto/subject203/"
+
+# Opret en query string med parametrene
+query_params <- list(
+  bbo = "10.395490670074423,55.22227193719089,10.296785378326376,55.18994697228769",
+  zoom = 14,
+  lat = 55.20611273543719,
+  lng = 10.346138024200382,
+  page = 1,
+  q_fritekst = "",
+  q_stednavn = "",
+  q_bygningsnavn = "",
+  q_person = "",
+  q_adresse = "",
+  notBefore = 1920,
+  notAfter = 1970,
+  category = "subject203",
+  itemType = "all",
+  thumbnailSize = "",
+  format = "xml"
+)
+
+# Send GET-anmodningen med parametre som query string
+response <- request(api_url) %>%
+  req_url_query(!!!query_params) %>%
+  req_perform()
+
+response
+

skrammel/R/utils-pipe.R

@@ -0,0 +1,14 @@
+#' Pipe operator
+#'
+#' See \code{magrittr::\link[magrittr:pipe]{\%>\%}} for details.
+#'
+#' @name %>%
+#' @rdname pipe
+#' @keywords internal
+#' @export
+#' @importFrom magrittr %>%
+#' @usage lhs \%>\% rhs
+#' @param lhs A value or the magrittr placeholder.
+#' @param rhs A function call using the magrittr semantics.
+#' @return The result of calling `rhs(lhs)`.
+NULL

skrammel/README.md

@@ -0,0 +1,4 @@
+# KB-API
+Code to pull data from the Royal Library API
+
+