Extract Tables from PDF with PDFTools 2.0

pdftools_tables.R

library(pdftools)
library(tidyverse)

parse_tables <- function(url, remove_last = TRUE) {
    
    my_data <- pdf_data(url)
    
    lapply(my_data, function(my_data2) {
        header_row <- my_data2 %>%
            filter(y == min(y))
        
        subdata <- my_data2 %>%
            filter(y != min(y))
        
        first_table_temp <- subdata %>%
            group_by(y) %>%
            summarise(text = paste(text, space, collapse = " ")) %>%
            mutate(text = gsub("TRUE ", "", text),
                   text = gsub("FALSE", ",,", text)) %>%
            mutate(text = sapply(strsplit(text, ",,"), str_trim)) %>%
            rowwise() %>%
            mutate(variable = list(tail(c("", header_row$text), length(text)))) %>%
            ungroup() %>%
            mutate(id = 1:nrow(.))
        
        if (remove_last) {
            first_table_temp <- first_table_temp %>%
                slice(-nrow(.))
        }
        
        first_table_temp %>%
            unnest() %>%
            spread(key = variable, value = text) %>%
            select(-y, -id) %>%
            select(one_of(c("V1", header_row$text))) %>%
            mutate_all(parse_guess)    
    })
}

parse_tables("https://github.com/ropensci/tabulizer/raw/master/inst/examples/data.pdf")