scraper_pdf_bovespa.R

library(tabulizer)
library(tidyverse)

arquivo <- "C:/Users/sillas.gonzaga/Downloads/gsa.pdf"

tab <- tabulizer::extract_tables(arquivo, encoding = "UTF-8", method = "lattice")

map(tab,dim)

# extrair tabelas com 8 colunas
df_operacoes <- tab %>%
  keep(~ dim(.x)[[2]] == 8) %>%
  imap(~ as.tibble(.) %>% 
         mutate(pagina = .y)) %>% 
  bind_rows()


glimpse(df_operacoes)

# colunas 5 a 8 sao numericas
limpar_col_numerica <- function(x){
  x <- stringr::str_remove_all(x, "[R$]")
  x <- stringr::str_remove_all(x, "[\r]")
  x <- stringr::str_remove_all(x, "[.]")
  x <- stringr::str_replace_all(x, ",", ".")
  as.numeric(x)
}

limpar_col_char <- function(x){
  x <- stringr::str_replace_all(x, "[\r]", " ")
  x <- stringr::str_replace_all(x, "Preferemcal|Preferenciall|Preferenciais|Preferencial", "PN")
  x <- stringr::str_to_upper(x)
  x
}

df_operacoes <- df_operacoes %>% 
  mutate_at(vars(V5:V8), limpar_col_numerica) %>% 
  filter(!is.na(V7)) %>% 
  mutate(V5 = as.integer(V5)) %>% 
  mutate_at(vars(V1:V4), limpar_col_char) %>% 
  group_by(pagina) %>% 
  mutate(linha_pagina = row_number()) %>% 
  ungroup()
  

table(df_operacoes$V4)

df_operacoes %>% 
  group_by(V4) %>% 
  summarise(v = sum(V8))

df_operacoes %>% 
  filter(V4 == "VENDA") %>% 
  group_by(pagina) %>% 
  summarise(s = sum(V8))