From 49d135c654df39c24459a58c346abdb54aa959c0 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jos=C3=A9=20de=20Jesus=20Filho?= <jjesusfilho@gmail.com>
Date: Wed, 7 Aug 2024 08:57:54 -0300
Subject: [PATCH] Update tjsp_ler_peticoes_diversas.R

---
 R/tjsp_ler_peticoes_diversas.R | 21 +++++++++++++++------
 1 file changed, 15 insertions(+), 6 deletions(-)

diff --git a/R/tjsp_ler_peticoes_diversas.R b/R/tjsp_ler_peticoes_diversas.R
index 2684e8d..ed273c9 100644
--- a/R/tjsp_ler_peticoes_diversas.R
+++ b/R/tjsp_ler_peticoes_diversas.R
@@ -20,20 +20,29 @@ tjsp_ler_peticoes_diversas <- function(arquivos = NULL,diretorio = ".") {
   purrr::map_dfr(arquivos, purrr::possibly(~{
 
 
-    pb$tick()
+   doc <-    xml2::read_html(.x) 
+    
+   processo <- doc |>
+      xml2::xml_find_first("//span[contains(@class,'unj-larger')]") |>
+      xml2::xml_text() |>
+      stringr::str_squish() |>
+      stringr::str_remove_all("[^\\d+\\s]") |>
+      stringr::str_trim()
 
-    processo <- stringr::str_extract(.x, "\\d{20}")
+    cd_processo <- doc |>
+      xml2::xml_find_first("//a[contains(@href,'processo.codigo')]/@href|//form[contains(@action,'processo.codigo')]/@action") |>
+      xml2::xml_text() |>
+      stringr::str_extract("(?<=processo.codigo=)\\w+")
 
-    
 
-       xml2::read_html(.x) |>
+       doc |>
       xml2::xml_find_all(xpath = "//div/h2[contains(text(),'Peti\u00E7\u00F5es diversas')]/../following-sibling::table[1]/tbody/tr") |>
       xml2::xml_text() |>
       stringr::str_trim() |>
       stringr::str_split("\n\\s+") |>
       purrr::map_dfr(stats::setNames, c("data","tipo")) |>
       dplyr::mutate(data = lubridate::dmy(data)) |>
-      tibble::add_column(processo  = processo, .before = 1)
+      tibble::add_column(processo  = processo, cd_processo = cd_processo,  .before = 1)
 
   }, otherwise = NULL))
-}
\ No newline at end of file
+}