Extract data from a PDF

wsaldo.csv

wsaldo.R

library('pdftools')
library('stringi')

options(stringsAsFactors = FALSE)

getWSaldo <- function(x, id) {
                k <- sapply(x, "[", id)
                k <- trimws(k)
                k <- gsub("\\.", "", k)
                k <- as.numeric(k)
                return(k)}

txt <- pdf_text("rb_00.....pdf")

extract_wsaldo <- function(text, first_page, last_page) {

  fin <- list()

  for (i in first_page:last_page) {

    x <- text[i]

    x <- gsub("\r\n", "::", x)
    x <- gsub("\\s+", " ", x)
    x <- gsub("- ", "-", x)

    x <- strsplit(x, "::")[[1]]

    x <- data.frame(x, stringsAsFactors = FALSE)

    # W-Saldo für zwei Zeiträume
    x <- data.frame(id =  trimws(x[2, ]), data = x[5:12, ])
    x$region <- trimws(stri_extract_all_regex(x$data, "[[:alpha:]][ \\/[:alpha:]]*"))
    x <- x[x$region %in% c("Ausland/unbek", "Land Berlin",
                           "Alte Bundesländer", "Neue Bundesländer",
                           "Land Brandenburg"), ]
    x$data <- stri_replace_all_regex(x$data, "\\.", "")

    wsaldo <- stri_extract_all_regex(x$data, "-?[[:digit:]][[:digit:]]*")
    wsaldo <- data.frame(ws01_14 = getWSaldo(wsaldo, 4), ws11_14 = getWSaldo(wsaldo, 5))
    x <- cbind(x, wsaldo)

    fin[[i]] <- x

  }

  fin <- do.call("rbind", fin)
  rownames(fin) <- NULL
  fin

}

fin <- extract_wsaldo(txt, 27, 110)

write.csv(tail(fin), file="wsaldo.csv", fileEncoding = "UTF-8", row.names = FALSE)