An R package to analyze the parliamentary records of the 19th legislative period of the Bundestag, the German parliament.
Вы не можете выбрать более 25 тем Темы должны начинаться с буквы или цифры, могут содержать дефисы(-) и должны содержать не более 35 символов.

147 строки
4.5KB

  1. source("config.R")
  2. source("../utils/helpers.R")
  3. library("xml2")
  4. library(tibble)
  5. library(dplyr)
  6. library(magrittr)
  7. # for usage see the example at the end
  8. read_all <- function() {
  9. available_protocols <- list.files(DOWNLOAD_DIR)
  10. res <- lapply(available_protocols, read_one)
  11. sapply(res, `[[`, "redner") %>%
  12. bind_rows() %>%
  13. distinct() ->
  14. redner
  15. sapply(res, `[[`, "reden") %>%
  16. bind_rows() %>%
  17. distinct() ->
  18. reden
  19. sapply(res, `[[`, "talks") %>%
  20. bind_rows() %>%
  21. distinct() ->
  22. talks
  23. list(redner = redner, reden = reden, talks = talks)
  24. }
  25. # this reads all currently parseable data from one xml
  26. read_one <- function(name) {
  27. print(paste("reading", name))
  28. x <- tryCatch(read_xml(paste0(DOWNLOAD_DIR, name)),
  29. error = function(c) NULL)
  30. if (is.null(x)) return(NULL)
  31. cs <- xml_children(x)
  32. verlauf <- xml_find_first(x, "sitzungsverlauf")
  33. rednerl <- xml_find_first(x, "rednerliste")
  34. xml_children(rednerl) %>%
  35. parse_rednerliste() ->
  36. redner
  37. xml_children(verlauf) %>%
  38. xml_find_all("rede") %>%
  39. parse_redenliste() ->
  40. res
  41. list(redner = redner, reden = res$reden, talks = res$talks)
  42. }
  43. xml_get <- function(node, name) {
  44. res <- xml_text %$% xml_find_all(node, name)
  45. if (length(res) == 0) NA_character_
  46. else res
  47. }
  48. # parse one redner
  49. parse_redner <- function(redner_xml) {
  50. redner_id <- xml_attr(redner_xml, "id")
  51. nm <- xml_child(redner_xml)
  52. vorname <- xml_get(nm, "vorname")
  53. nachname <- xml_get(nm, "nachname")
  54. fraktion <- xml_get(nm, "fraktion")
  55. titel <- xml_get(nm, "titel")
  56. rolle <- xml_find_all(nm, "rolle")
  57. if (length(rolle) > 0) {
  58. rolle_lang <- xml_get(rolle, "rolle_lang")
  59. rolle_kurz <- xml_get(rolle, "rolle_kurz")
  60. } else rolle_kurz <- rolle_lang <- NA_character_
  61. c(id = redner_id, vorname = vorname, nachname = nachname, fraktion = fraktion, titel = titel,
  62. rolle_kurz = rolle_kurz, rolle_lang = rolle_lang)
  63. }
  64. # parse one rede
  65. # returns: - a rede (with rede id and redner id)
  66. # - all talks appearing in the rede (with corresponding content)
  67. parse_rede <- function(rede_xml) {
  68. rede_id <- xml_attr(rede_xml, "id")
  69. cs <- xml_children(rede_xml)
  70. cur_redner <- NA_character_
  71. principal_redner <- NA_character_
  72. cur_content <- ""
  73. reden <- list()
  74. for (node in cs) {
  75. if (xml_name(node) == "p") {
  76. klasse <- xml_attr(node, "klasse")
  77. if (!is.na(klasse) && klasse == "redner") {
  78. if (!is.na(cur_redner)) {
  79. rede <- c(rede_id = rede_id,
  80. redner = cur_redner,
  81. content = cur_content)
  82. reden <- c(reden, list(rede))
  83. cur_content <- ""
  84. } else {
  85. principal_redner <- xml_child(node) %>% xml_attr("id")
  86. }
  87. cur_redner <- xml_child(node) %>% xml_attr("id")
  88. } else {
  89. cur_content <- paste0(cur_content, xml_text(node), sep="\n")
  90. }
  91. }
  92. }
  93. rede <- c(rede_id = rede_id,
  94. redner = cur_redner,
  95. content = cur_content)
  96. reden <- c(reden, list(rede))
  97. list(rede = c(id = rede_id, redner = principal_redner),
  98. parts = reden)
  99. }
  100. # creates a tibble of reden and a tibble of talks from a list of xml nodes representing reden
  101. parse_redenliste <- function(redenliste_xml) {
  102. d <- sapply(redenliste_xml, parse_rede)
  103. reden <- simplify2array(d["rede", ])
  104. parts <- simplify2array %$% unlist(d["parts", ], recursive=FALSE)
  105. list(reden = tibble(id = reden["id",], redner = reden["redner",]),
  106. talks = tibble(rede_id = parts["rede_id", ],
  107. redner = parts["redner", ],
  108. content = parts["content", ]))
  109. }
  110. # create a tibble of redner from a list of xml nodes representing redner
  111. parse_rednerliste <- function(rednerliste_xml) {
  112. d <- sapply(rednerliste_xml, parse_redner)
  113. tibble(id = d["id",],
  114. vorname = d["vorname",],
  115. nachname = d["nachname",],
  116. fraktion = d["fraktion",],
  117. titel = d["titel",],
  118. rolle_kurz = d["rolle_kurz",],
  119. rolle_lang = d["rolle_lang",])
  120. }
  121. # -------------------------------
  122. # EXAMPLE USE
  123. # make sure data ist downloaded via fetch.R
  124. res <- read_one("19126-data.xml")
  125. res$redner
  126. res$reden
  127. res$talks
  128. # -------------------------------