explore ICES advice with tabulizer

ICES_tab.R

## Exploring ICES advice ##

library(tabulizer)
library(glue)
library(dplyr)
library(stringr)
library(purrr)
library(httr)
library(jsonlite)

rawsd <- jsonlite::fromJSON("http://sd.ices.dk/services/odata3/StockListDWs3")$value %>% 
  filter(ActiveYear == 2017) %>% 
  mutate(stock_code = dplyr::case_when(YearOfLastAssessment <= 2016 ~ PreviousStockKeyLabel,
                                YearOfLastAssessment >= 2017 ~ StockKeyLabel,
                                TRUE ~ NA_character_),
         advice_url = glue::glue("http://www.ices.dk/sites/pub/Publication%20Reports/Advice/{YearOfLastAssessment}/{YearOfLastAssessment}/{stock_code}.pdf"),
         valid_url = purrr::map(advice_url, httr::http_error) == FALSE) %>% 
  select(StockKeyDescription,
         StockKeyLabel,
         YearOfLastAssessment,
         advice_url,
         valid_url)

## This downloads all of the first pages of advice into a data.frame. The same concept could be done with tables
advice_table <- rawsd %>% 
  filter(valid_url == TRUE) #%>% # Note, there are ~15 that don't have a valid url... this should be explored
  # mutate(advice_sentence = purrr::map(advice_url, tabulizer::extract_tables, pages = 1, method = "data.frame"))

td <- tabulizer::extract_tables(advice_table$advice_url[1], guess = FALSE, method = "data.frame")
td <- data.frame(VAL = as.character(td[[1]]), stringsAsFactors = FALSE)

## Advice value
## TAC over time
## Total wanted and unwanted catch
## Advice at FMSY




start_row <- grep("ICES stock advice", td$VAL)
end_row <- grep("Stock development over time", td$VAL)

td$VAL[start_row]


library(icesVocab)