Utility function to get English common name from Wikidata

R/get-common-name.R

@@ -0,0 +1,113 @@
+# Function to retrieve a taxon's common names from Wikidata
+
+# Author: Stefan Vriend
+# Created: 2024-10-24
+# Last updated: 2024-11-11
+
+
+# Load packages -----------------------------------------------------------
+
+library(WikidataQueryServiceR)
+library(dplyr)
+library(stringr)
+#library(WikidataR)
+
+# Get common name through {WikidataQueryServiceR} -----------------
+
+# Note: this code uses SPARQL (an RDF-query language)
+
+# Arguments
+# sci_name: Character specifying one or more scientific names
+# lang: Character specifying one or more languages of the common name
+
+get_common_name <- function(sci_name,
+                            lang = "en") {
+
+  # SPARQL query to select common name from Wikidata
+  query <- paste0('
+      SELECT DISTINCT
+        ?item ?common_name (LANG(?common_name) AS ?lang)
+      WHERE {
+        ?item wdt:P225', '"', sci_name, '"', ';
+              wdt:P1843 ?common_name.
+
+      FILTER(LANGMATCHES(LANG(?common_name), ', '"', lang, '"))
+
+      SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],en". }}
+      ')
+
+  # Send query to Wikidata query service and extract common name
+  common_name <- WikidataQueryServiceR::query_wikidata(sparql_query = query,
+                                                       format = "smart") |> 
+    dplyr::bind_rows() |> 
+    dplyr::mutate(common_name = stringr::str_to_title(common_name)) |> 
+    dplyr::distinct() |> 
+    dplyr::select(-"item")
+
+  # For some taxa (e.g., plants and fungi), non-English common names are stored as 
+  # rdfs:label properties rather than wdt:P1843 properties
+  if(nrow(common_name) != length(lang)) {
+
+    query <- paste0('
+      SELECT DISTINCT
+        ?item ?label ?common_name (LANG(?label) AS ?lang) (LANG(?common_name) AS ?lang_cn)
+      WHERE {
+        ?item wdt:P225', '"', sci_name, '"', ';
+              wdt:P1843 ?common_name;
+              rdfs:label ?label.
+
+      FILTER(LANG(?label) = ', '"', lang, '")
+
+      SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],en". }}
+      ')
+
+    common_name <- WikidataQueryServiceR::query_wikidata(sparql_query = query,
+                                                         format = "smart") |> 
+      dplyr::bind_rows() |> 
+      dplyr::mutate(common_name = dplyr::case_when(lang == "en" & lang_cn == "en" ~ stringr::str_to_title(common_name),
+                                                   lang == "en" & lang_cn != "en" ~ NA_character_,
+                                                   lang != "en" ~ stringr::str_to_title(label))) |> 
+      dplyr::select(-"lang_cn") |> 
+      dplyr::distinct() |> 
+      dplyr::select(-"item", -"label") |> 
+      # For chosen languages that do not have a common name for the taxon,
+      # (e.g., Wikidata does not provide an English common name for the fungus Sparassis crispa)
+      # return NA.
+      dplyr::filter(!(is.na(common_name) & dplyr::n() > 1), .by = "lang")
+
+  }
+
+  return(common_name)
+
+}
+
+
+# (alternative) Get English common name through {WikidataR} ---------------
+
+# Arguments
+# sci_name: Character specifying one or more scientific names
+
+# get_common_name2 <- function(sci_name) {
+#   
+#   purrr::map_chr(.x = sci_name,
+#                  .f = ~{
+#                    
+#                    # Search Wikidata item corresponding to scientific name
+#                    item <- WikidataR::find_item(.x) |> 
+#                      purrr::flatten()
+#                    
+#                    # Retrieve properties from Wikidata item
+#                    wiki <- WikidataR::get_item(item$id)
+#                    
+#                    # Extract English common name
+#                    wiki[[1]]$claims$P1843 |> 
+#                      purrr::pluck("mainsnak", "datavalue", "value") |> 
+#                      dplyr::filter(language == "en") |> 
+#                      dplyr::pull("text") |> 
+#                      purrr::as_vector() |> 
+#                      stringr::str_to_sentence() |>
+#                      unique()
+#                    
+#                  })
+#   
+# }