database_creation/eurasianphonology_database_creation.R

In agricolamz/lingtypology: Linguistic Typology and Mapping

library(jsonlite)
json_data <- fromJSON(txt = "https://github.com/macleginn/eurasian-phonologies/raw/master/src/dbase/phono_dbase.json")
json_data <- lapply(json_data, function(x) {
  x[sapply(x, is.null)] <- NA
  unlist(x)
})


lapply(seq_along(json_data), function(i){
  data.frame(names = names(json_data[[i]]),
             value = json_data[[i]],
             id = names(json_data[i]))
}) ->
  res

df <- as.data.frame(do.call("rbind", res))

library(tidyverse)

df %>%
  pivot_wider(names_from = names, values_from = value) %>%
  pivot_longer(names_to = "segment_type", values_to = "segments",
               cols = -c(id, code, type, name, coords1, coords2, gen1,
                         gen2, tones, syllab, cluster, finals, source, comment,
                         contr)) %>%
  filter(!is.na(segments),
         !str_detect(segment_type, "inv")) %>%
  mutate(segment_type = ifelse(str_detect(segment_type, "cons"), "consonant",
                               "vowel")) ->
  result

result$glottocode <- lingtypology::gltc.iso(result$code)

result$glottocode[result$id == 'Darkhat#16'] <- 'dark1243'
result$glottocode[result$id == 'Northwestern Pashto#52'] <- 'nort2646'
result$glottocode[result$id == 'Japhug#72'] <- 'japh1234'
result$glottocode[result$id == 'Zuberoan Basque#86'] <- 'basq1248'
result$glottocode[result$id == 'Beserman (Šamardan)#91'] <- 'bese1243'
result$glottocode[result$id == 'Yodzyak Komi#92'] <- 'komi1267'
result$glottocode[result$id == 'Forest Nenets#100'] <- 'fore1274'
result$glottocode[result$id == 'Solon#132'] <- 'even1259'
result$glottocode[result$id == 'Eastern Khanty (Vakh)#143'] <- 'east2774'
result$glottocode[result$id == 'Eastern Mari#150'] <- 'east2774'
result$glottocode[result$id == 'Dolakha Newar#172'] <- 'newa1246'
result$glottocode[result$id == 'Digor Ossetic#222'] <- 'digo1242'
result$glottocode[result$id == 'South Mustang Tibetan (Jharkhot)#352'] <-
  'lowa1242'
result$glottocode[result$id == 'Tamang (Dhankuta)#360'] <- 'east2347'
result$glottocode[result$id == 'Lizu#362'] <- 'lizu1234'
result$glottocode[result$id == 'Puxi#363'] <- 'sout2728'
result$glottocode[result$id == 'Kurtoep#364'] <- 'kurt1248'
result$glottocode[result$id == 'Sangdam Tibetan#369'] <- 'kham1299'
result$glottocode[result$id == 'Myeik Burmese#370'] <- 'merg1238'
result$glottocode[result$id == 'Elfdalian#377'] <- 'elfd1234'
result$glottocode[result$id == 'Lemi Chin#403'] <- 'east2779'
result$glottocode[result$id == 'Amədya Neo-Aramaic#420'] <- 'east2779'
result$glottocode[result$id == 'Arbel Neo-Aramaic#421'] <- 'lish1245'
result$glottocode[result$id == 'Barwar Neo-Aramaic#423'] <- 'assy1241'
result$glottocode[result$id == 'Rushani#428'] <- 'rush1239'
result$glottocode[result$id == 'Khufi#429'] <- 'khuf1238'

result$code <- unname(lingtypology::iso.gltc(result$glottocode))

result %>%
  separate(id, into = c("lang", "id"), "#") %>%
  mutate(-lang) ->
  result

result %>%
  mutate(contr = str_replace_all(contr,
                                 "André Nikulin (ojovemlouco@gmail.com)",
                                 "André Nikulin, ojovemlouco@gmail.com"),
         contr = str_replace_all(contr,
                                 "Dmitry Nikolayev, dsnikolaev@gmail.com",
                                 "Dmitry Nikolaev, dsnikolaev@gmail.com"),
         type = ifelse(type == "Язык", "language", "dialect")) %>%
  rename(contributor = contr,
         latitude = coords1,
         longitude = coords2,
         iso = code) ->
  result

result %>%
  count(type)

result$language <- lingtypology::lang.gltc(result$glottocode)
eurasianphonology <- result[c("id",
                              "name",
                              "language",
                              "iso",
                              "glottocode",
                              "type",
                              "latitude",
                              "longitude",
                              "gen1",
                              "gen2",
                              "tones",
                              "syllab",
                              "cluster",
                              "finals",
                              "segments",
                              "segment_type",
                              "source",
                              "comment",
                              "contributor")]

save(eurasianphonology, file="data/eurasianphonology.RData", compress= 'xz')