# Script: filter different markets( just for cake oil market and sugar, cereals...)

#' function reads balance detailed, split it into "biofuels" and "non-biofuels"
#'
#' @param balance_detailed A data frame.
#' @param p_biofuels boolean.

#' @param products products from dimdefs.xml

#' @return links
#' @export
#'

prelinks <- function(balance_detailed, p_biofuels = TRUE,products){


  # balance_detailed =balance_detailed_Baseline
  # products information

  added <- products  %>% dplyr::select(key,itemName,color) %>% as_tibble()

  # add variable key, itemName and color from products

  balance_detailed <- left_join(balance_detailed, added, by = c(".i4" = "key"))

  # balance_detailed <- left_join(balance_detailed, dimdef_product[,c(2,3,9)], by = c(".i4" = "key"))

  # rename new variable
  colnames(balance_detailed)[13] <- "Commodities"
  colnames(balance_detailed)[1] <- "key"

  balance_detailed$Commodities <- sub("\\[.*?\\]", "", balance_detailed$Commodities)


  library(magrittr)
  # prod_list_fullname <- balance_detailed %>% select(Commodities) %>% extract2(1)

  # remove character begin with [ and end with ]
  # prod_list_fullname <- gsub("\\[.*?\\]", "", prod_list_fullname)
  # p_biofuels = T

  if (p_biofuels){
    selected_list = c("Rape seed",
                      "Soya seed",
                      "Sunflower seed",
                      "Rape seed oil",
                      "Soya oil",

                      "Sunflower seed oil",

                      "Soya cake",
                      "Sunflowe seed cake",
                      "Rape seed cake",
                      "Destilled dried grains from bio-ethanol processing",
                      "Pulses",
                      "Bio ethanol")
  }else{
    selected_list = c("Wheat","Grain maize","Barley", "Other cereals","Poultry meat",
                      "Sugar", "Fish and other acquatic products", "Rice milled")
  }

  #
  balance_detailed_ <- balance_detailed %>% as.data.frame()

  balance_detailed_selected <-  filter(balance_detailed_, balance_detailed_[,13] %in% selected_list) %>% as_tibble()

  #rename "Destilled dried grains from bio-ethanol processing" -> "DDGS"
  # balance_detailed_selected <- balance_detailed_selected %>%
  #   mutate(Commodities = ifelse(Commodities!="Destilled dried grains from bio-ethanol processing",Commodities,"DDGS"))


  # filter
  balance_detailed_all <- balance_detailed_selected %>%
    select(-.i1)  %>% # remove region
    select(-.i5)  %>% # remove year: 2030
    select(-interv_ch)  %>%
    group_by(Commodities) %>%
    summarise_at(vars(supply:exports),sum,na.rm=TRUE)

  library(magrittr)
  prod_list_fullname <- balance_detailed_selected %>% select(Commodities) %>% extract2(1)

  #print(prod_list_fullname)
  vars_to_run <- prod_list_fullname


  left_side <- c("supply","imports")

  links_list <- list()
  #nodes_list <- list()

  vars_to_run <- vars_to_run[!vars_to_run %in%
                               c("Soya oil",
                                 "Soya cake",
                                 "Rape seed oil",
                                 "Rape seed cake",

                                 "Sunflower seed oil",

                                 "Sunflowe seed cake",
                                 "Bio ethanol",
                                 "Destilled dried grains from bio-ethanol processing"
                               )]
  # "Rape seed"      "Sunflower seed" "Soya seed"      "Pulses"
  for (i in vars_to_run){

    data_df <- reshape2::melt(balance_detailed_all)

    data_df_1 <- data_df %>% filter(Commodities == i)
    data_df_1_l <- data_df_1  %>% filter(variable %in% left_side )
    data_df_1_r <- data_df_1  %>% filter(!variable %in% left_side)

    data_df_1_l_agg <- data_df_1_l %>% group_by(Commodities) %>% dplyr::summarise(value=sum(value))

    data_df_1_l_agg$variable <- paste0(i)

    first <- merge(data_df_1_l, data_df_1_l_agg, by="Commodities",all=TRUE)

    first <- first[,c("variable.x","variable.y","value.x")]
    colnames(first)[3] <- "value"


    second <- merge(data_df_1_l_agg,data_df_1_r,by="Commodities",all=TRUE)
    second <- second[,c("variable.x","variable.y","value.y")]
    colnames(second)[3] <- "value"

    if (p_biofuels){
      if (i != "Pulses"){
        if (i == "Soya seed") {produced  = c("Soya oil","Soya cake")}
        else if (i == "Rape seed") {produced  = c("Rape seed oil","Rape seed cake") }

        else if (i == "Sunflower seed") {produced  = c("Sunflower seed oil","Sunflowe seed cake") }

        else {print(i);break}

        data_df_proc_help <- balance_detailed_all %>% filter(Commodities %in% produced)
        data_df_proc_help <- reshape2::melt(data_df_proc_help)
        data_df_proc_help <- data_df_proc_help %>% filter(variable== "supply" )
        data_df_proc_help$variable <- paste(data_df_proc_help$Commodities)

        data_df_proc_help$Commodities <- i
        colnames(data_df_proc_help) <- c("variable.x","variable.y","value")

        second <- rbind(second,data_df_proc_help)
        second <- second %>% filter(variable.y!="processing")

        data_df_proc_help <- balance_detailed_all %>% filter(Commodities %in% produced)
        data_df_proc_help <- reshape2::melt(data_df_proc_help)
        data_df_proc_help <- data_df_proc_help %>% filter(variable!="supply")
        colnames(data_df_proc_help) <- c("variable.x","variable.y","value")
        second <- rbind(second,data_df_proc_help)
        second_imports <- second %>% filter(variable.y=="imports")
        second <- second %>% filter(!variable.y=="imports")
        colnames(second_imports) <- c("variable.y","variable.x","value")
        second <- rbind(second,second_imports)
      }

    }

    combined <- rbind(first,second)
    links_list[[i]] <- combined


    links <- rbindlist(links_list)
    links <- unique(links)

  }

  if(p_biofuels){

    # links[grepl("Bio",variable.x) | grepl("Bio",variable.y)]
    setnames(links,old=c("variable.x","variable.y"),new=c("source","target"))
    # links[grepl("Bio",source) | grepl("Bio",target)]

    i <- c("Bio ethanol","Destilled dried grains from bio-ethanol processing")

    data_df_proc_help <- balance_detailed_all %>% filter(Commodities %in% i)
    data_df_proc_help <- reshape2::melt(data_df_proc_help)
    data_df_proc_help <- data_df_proc_help %>% filter(variable=="supply")
    data_df_proc_help$variable <- paste(data_df_proc_help$Commodities)

    data_df_proc_help$Commodities <- "biofuels"
    colnames(data_df_proc_help) <- c("variable.x","variable.y","value")

    data_df <- reshape2::melt(balance_detailed_all)

    data_df_1 <- data_df[data_df$Commodities %in% i,]
    data_df_1_l <- data_df_1[data_df_1$variable %in% left_side,]
    data_df_1_l$variable <- as.character(data_df_1_l$variable)
    data_df_1_l <- data_df_1_l %>% filter(variable!="supply")
    data_df_1_r <- data_df_1[!data_df_1$variable %in% left_side,]
    data_df_1_r$variable <- as.character(data_df_1_r$variable)

    data_df_1_l_agg <- data_df_1_l%>%group_by(Commodities) %>% dplyr::summarise(value=sum(value))