From 7957698c75fe49b2b3e46d5c3b6407a95cd815bc Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jos=C3=A9=20de=20Jesus=20Filho?= <jjesusfilho@gmail.com>
Date: Mon, 4 Mar 2024 16:13:32 -0300
Subject: [PATCH] Update tjsp_ler_tabela_docs_cd_processo.R

---
 R/tjsp_ler_tabela_docs_cd_processo.R | 48 +++++++++++++++-------------
 1 file changed, 26 insertions(+), 22 deletions(-)

diff --git a/R/tjsp_ler_tabela_docs_cd_processo.R b/R/tjsp_ler_tabela_docs_cd_processo.R
index 4d81478..44d94d8 100644
--- a/R/tjsp_ler_tabela_docs_cd_processo.R
+++ b/R/tjsp_ler_tabela_docs_cd_processo.R
@@ -7,57 +7,61 @@
 #' @export
 #'
 tjsp_ler_tabela_docs_cd_processo <- function(arquivos = NULL, diretorio = "."){
-
+  
   if (is.null(arquivos)) {
-
+    
     arquivos <- list.files(diretorio,full.names = TRUE)
-
+    
   }
-
+  
   pb <- progress::progress_bar$new(total = length(arquivos))
-
+  
   purrr::map_dfr(arquivos,purrr::possibly(~{
-
+    
     pb$tick()
-
-    cd_processo <- stringr::str_extract(.x,"(?<=processo_)\\w+")
-
+    
+    cd_processo_pg <- stringr::str_extract(.x,"(?<=processo_pg_)\\w+?(?=_)")
+    cd_processo_sg <- stringr::str_extract(.x,"(?<=processo_sg_)\\w+")
+    
     suppressMessages({
-
+      
       doc <-   .x |>
         xml2::read_html() |>
         xml2::xml_text() |>
         stringr::str_extract("(?<=requestScope = )\\X+?(?=;)") |>
         jsonlite::fromJSON()
-
-
+      
+      
       doc_name <- tibble::tibble(doc_name= doc$data$title) |>
         tibble::rownames_to_column("id_doc")
-
+      
       paginas  <- doc$children[[2]]$data$indicePagina
-
+      
       df <- purrr::imap_dfr(doc$children,~{
-
+        
         url_doc <-  .x$data$parametros
-
+        
         pagina_inicial <- .x$data$title |>
           stringr::str_extract("\\d+")
-
+        
         pagina_final <- .x$data$title |>
           stringr::str_extract("\\d+$")
-
+        
         tibble::tibble(id_doc = .y, pagina_inicial, pagina_final, url_doc) |>
           dplyr::mutate(id_doc = as.character(id_doc))
-
+        
       }) |>
         dplyr::left_join(doc_name) |>
         dplyr::select(id_doc, doc_name, pagina_inicial, pagina_final, url_doc) |>
         dplyr::mutate(url_doc = paste0("https://esaj.tjsp.jus.br/pastadigital/getPDF.do?",url_doc)) |>
         dplyr::group_by(id_doc) |>
         dplyr::ungroup() |>
-        tibble::add_column(cd_processo, .before =1)
-
+        tibble::add_column(cd_processo_pg, .before =1) |> 
+        tibble::add_column(cd_processo_sg, .after  = 1) |> 
+        dplyr::mutate(instancia = ifelse(is.na(cd_processo_sg), 1, 2), .after = 2)
+      
     })
-
+    
   }, NULL))
 }
+