baixar obras do governo

obras.R

library(tabulizer)
library(tidyverse)
library(stringr)
library(abjutils)
u <- "http://legis.senado.leg.br/sdleg-getter/documento/download/a5e8c92c-86c1-45bb-99bd-3ad098905e81"
tab <- tabulizer::extract_tables(u)

arrumar <- function(x) {
  x %>% 
    tolower() %>% 
    str_trim() %>% 
    str_replace_all('[[:space:]]+', '_') %>% 
    str_replace_all('%', 'p') %>% 
    str_replace_all('r\\$', 'reais') %>% 
    rm_accent()
}

tab_tidy <- tab %>% 
  map(as_tibble) %>% 
  bind_rows() %>% 
  set_names(arrumar(.[1,])) %>% 
  slice(-1) %>% 
  mutate_all(funs(str_replace_all(., '[[:space:]]+', ' '))) %>% 
  mutate_all(str_trim)

write_csv(tab_tidy, 'obras.csv')