R/fct_download_read_dfp_zip_file.R

In msperlin/GetCVMData: Reading Annual and Quarterly Financial Reports from CVM

#' Downloads and reads DFP datasets
#'
#' @param url_in Url of file to download and read
#' @inheritParams get_dfp_data
#'
#' @return A dataframe with dfp data
#' @export
#'
#' @examples
#'
#' \dontrun{ # keep cran check fast
#' my_url <- 'http://dados.cvm.gov.br/dados/CIA_ABERTA/DOC/DFP/DRE/DADOS/dre_cia_aberta_2010.zip'
#' df_out <- download_read_dfp_zip_file(url_in = my_url,
#'                                      companies_cvm_codes = NULL,
#'                                      type_format = 'con',
#'                                      clean_data = TRUE)
#' }
download_read_dfp_zip_file <- function(url_in,
                                       companies_cvm_codes,
                                       type_format,
                                       cache_folder = 'gcvmd_cache', clean_data) {

  # create folder
  dir_zip <- file.path(cache_folder, 'DFP_zip_files')
  if (!dir.exists(dir_zip)) dir.create(dir_zip, recursive = TRUE)

  message('\tDowloading ', basename(url_in), appendLF = TRUE)
  dest_file <- file.path(dir_zip, basename(url_in))

  flag_dl <- my_download_file(dl_link = url_in,
                              dest_file = dest_file,
                              max_dl_tries = 10, be_quiet = TRUE)

  message('\t\tUnzipping', appendLF = TRUE)
  # unzip file in tempdir
  #message('\t\t\tunzipping file')
  unzip_dir <- file.path(tempdir(), tools::file_path_sans_ext(
    basename(url_in) ) )
  utils::unzip(zipfile = dest_file,exdir = unzip_dir,
        junkpaths = TRUE)

  unzipped_files <- list.files(unzip_dir, full.names = TRUE)

  # filter by type and format
  type_files_format <- as.character(
    purrr::map(stringr::str_match_all(basename(unzipped_files), 'aberta_(.*)_\\d\\d\\d\\d'),
               2)
  )

  idx <- (type_files_format %in% type_format)
  unzipped_files <- unzipped_files[idx]

  #message('\t\t\t\tfound ', length(unzipped_files), ' files')
  #message('\t\t\treading files', appendLF = FALSE)

  df_out <- dplyr::bind_rows(purrr::map(unzipped_files,
                                        read_dfp_csv, clean_data = clean_data))

  # filter by company
  if (!is.null(companies_cvm_codes)) {
    idx <- df_out$CD_CVM %in% companies_cvm_codes
    df_out <- df_out[idx, ]
  }

  message('\t\tGot ', nrow(df_out), ' rows | ', length(unique(df_out$CD_CVM)), ' companies')

  # clean up zip dir
  unlink(unzip_dir, recursive = TRUE)

  return(df_out)

}