キャッシュレス消費者還元事業登録リストのデータフレーム化

kameiten_list.R

library(tidyverse)
library(tabulizer)
tweak_kameiten_df <- function(data) {
  data %>%
    dplyr::filter(!is.na(`No.`)) %>%
    dplyr::mutate(還元率 = units::set_units(還元率, "%")) %>%
    dplyr::arrange(`No.`) %>%
    tibble::as_tibble()
}
kameiten_na <- function() {
  tibble::tribble(
    ~No., ~都道府県, ~市区町村, ~事業所名_屋号, ~業種_大, ~業種_小, ~還元率,
    NA_integer_, NA_character_, NA_character_, NA_character_ ,NA_character_, NA_character_, NA_real_
  )
}
extract_kameiten <- function(path, start_page, end_page) {
  df <-
    tabulizer::extract_tables(file = path,
                   pages = seq.int(start_page, end_page),
                   output = "data.frame") %>%
    purrr::map(~ janitor::remove_empty(.x, "cols"))
  df_a <-
    df %>%
    purrr::keep(~ ncol(.x) == 6)
  df_b <-
    df %>%
    purrr::keep(~ ncol(.x) == 5)
  if (length(df_a) > 0) {
    df_a <-
      df_a %>%
      purrr::map(~ tibble::as_tibble(.x) %>%
            purrr::set_names(paste0("tmp", seq_len(6))) %>%
            tidyr::separate(col = tmp1, into = c("No.", "都道府県"), sep = "[[:space:]]+")) %>%
      purrr::reduce(rbind) %>%
      purrr::set_names(c("No.", "都道府県", "市区町村", "事業所名_屋号", "業種_大", "業種_小", "還元率")) %>%
      readr::type_convert(col_types = "icccccn")
  } else {
    df_a <-
      kameiten_na()
  }
  if (length(df_b) > 0) {
    df_b <-
      df_b %>%
      purrr::map(~ tibble::as_tibble(.x) %>%
            tidyr::separate(col = `No..都道府県`, into = c("No.", "都道府県"), sep = "[[:space:]]+") %>%
            dplyr::mutate(`事業所名_屋号` = stringr::str_replace(`事業所名.屋号.`, "(.*)[[:space:]](.+)", "\\1"),
                          `業種_大`       = stringr::str_replace(`事業所名.屋号.`, ".*[[:space:]](.+)", "\\1"))) %>%
      purrr::reduce(rbind) %>%
      dplyr::rename(業種_小  = 業種) %>%
      dplyr::select(`No.`, 都道府県, 市区町村, 事業所名_屋号, 業種_大, 業種_小, 還元率) %>%
      readr::type_convert(col_types = "icccccn")
  } else {
    df_b <-
     kameiten_na()
  }
  rbind(
    df_a,
    df_b) %>%
    tweak_kameiten_df()
}
extract_kameiten_ec <- function(path, start_page, end_page) {
  tabulizer::extract_tables(path,
                            pages = seq.int(start_page, end_page),
                            output = "data.frame") %>%
    purrr::reduce(rbind) %>%
    purrr::set_names(c("No.", "事業所名_屋号", "還元率")) %>%
    readr::type_convert(col_types = "icn") %>%
    tweak_kameiten_df()
}

pdf_path <- "kameiten_touroku_list.pdf"
# 1 固定店舗... 3 ~ 5536 ------------------------------------------------------
d01_kotei_tenpo <-
  extract_kameiten(pdf_path, 3, 5536)
# 2 楽天市場 5537 ~ 5904 ----------------------------------------------------------------
d02_rakuten <-
  extract_kameiten_ec(pdf_path, 5537, 5904)
# 3 Yahoo!ショッピング 5905 ~ 6245 ----------------------------------------------------------
d03_yahoo <-
  extract_kameiten_ec(pdf_path, 5905, 6245)
# 4 その他ECサイト 6246 ~ 6360 --------------------------------------------------------------
d04_other <-
  extract_kameiten_ec(pdf_path, 6246, 6360)