R/mfl_transactions.R

In ffscrapr: API Client for Fantasy Football League Platforms

#### MFL TRANSACTIONS ####

#' Get full transactions table
#'
#' @param conn the list object created by `ff_connect()`
#' @param transaction_type parameter to return transactions of the specified type. Types are: `WAIVER`, `BBID_WAIVER`, `FREE_AGENT`, `TRADE`, `IR`, `TAXI`, `AUCTION_INIT`, `AUCTION_BID`, `AUCTION_WON`, or `*` for all. Can also pass a comma-separated string.  Defaults to `*`. Note that only the types listed above are actually cleaned and processed by ffscrapr - you will need to make a custom api request with `mfl_getendpoint()` to receive other things.
#' @param ... other arguments passed the API request (case sensitive parameters such as FRANCHISE, DAYS, COUNT etc) - see <https://api.myfantasyleague.com/2022/api_info?STATE=test&CCAT=export&TYPE=transactions> for more details.
#'
#' @describeIn ff_transactions MFL: returns all transactions, including auction, free agents, IR, TS, waivers, and trades.
#'
#' @examples
#' \donttest{
#' try({ # try only shown here because sometimes CRAN checks are weird
#'   dlf_conn <- mfl_connect(2019, league_id = 37920)
#'   ff_transactions(dlf_conn)
#' }) # end try
#' }
#' @export

ff_transactions.mfl_conn <- function(conn, transaction_type = "*", ...) {

  df_transactions <- mfl_getendpoint(conn, "transactions", TRANS_TYPE = transaction_type) %>%
    purrr::pluck("content", "transactions", "transaction") %>%
    tibble::tibble() %>%
    tidyr::unnest_wider(1) %>%
    dplyr::mutate_at("timestamp", ~ as.numeric(.x) %>% .as_datetime())

  if (nrow(df_transactions) == 0) {
    return(NULL)
  }

  if (!"comments" %in% names(df_transactions)) df_transactions$comments <- NA_character_

  transaction_functions <- list(
    auction = .mfl_transactions_auction,
    free_agent = .mfl_transactions_freeagent,
    injured_reserve = .mfl_transactions_injuredreserve,
    taxi_squad = .mfl_transactions_taxisquad,
    bbid_waiver = .mfl_transactions_bbid_waiver,
    trade = .mfl_transactions_trade
  )

  players_endpoint <- mfl_players(conn) %>%
    dplyr::select("player_id", "player_name", "pos", "team")

  purrr::map_dfr(transaction_functions, rlang::exec, df_transactions) %>%
    dplyr::arrange(dplyr::desc(.data$timestamp)) %>%
    dplyr::left_join(
      players_endpoint,
      by = "player_id"
    ) %>%
    dplyr::left_join(
      dplyr::select(ff_franchises(conn), "franchise_id", "franchise_name"),
      by = c("franchise" = "franchise_id")
    ) %>%
    dplyr::select(
      dplyr::any_of(c("timestamp", "type", "type_desc",
        "franchise_id" = "franchise", "franchise_name",
        "player_id", "player_name", "pos", "team",
        "bbid_spent", "trade_partner", "comments"
      )),
      dplyr::everything()
    )
}

## AUCTION ##
#' @noRd
#' @keywords internal

.mfl_transactions_auction <- function(df_transactions) {
  auction_transactions <- df_transactions %>%
    dplyr::filter(.data$type %in% c("AUCTION_INIT", "AUCTION_BID", "AUCTION_WON"))

  if (nrow(auction_transactions) == 0) {
    return(NULL)
  }

  auction_transactions %>%
    dplyr::select("timestamp", "type", "franchise", "transaction") %>%
    tidyr::separate("transaction", into = c("player_id", "bid_amount", "comments"), sep = "\\|", fill = "right") %>%
    dplyr::mutate(
      "comments" = ifelse(stringr::str_length(.data$comments) == 0,
        NA_character_,
        .data$comments
      ),
      bid_amount = stringr::str_extract(.data$bid_amount, "[0-9,\\.]+") %>% as.numeric()
    ) %>%
    dplyr::filter(!is.na(.data$bid_amount))
}

## TRADE ##
# Create a tibble where each team's transaction is represented in one line (two equal but opposite lines per trade)
#' @noRd
#' @keywords internal

.mfl_transactions_trade <- function(df_transactions) {
  trade_transactions <- df_transactions %>%
    dplyr::filter(.data$type == "TRADE")

  if (nrow(trade_transactions) == 0) {
    return(NULL)
  }

  parsed_trades <- trade_transactions %>%
    dplyr::select(
      "timestamp", "type",
      "franchise", "franchise1_gave_up",
      "franchise2", "franchise2_gave_up",
      "comments"
    ) %>%
    dplyr::mutate_at(c("franchise1_gave_up", "franchise2_gave_up"), ~ stringr::str_replace(.x, ",$", "")) %>%
    dplyr::mutate_at(c("franchise1_gave_up", "franchise2_gave_up"), ~ stringr::str_split(.x, ","))

  df <- parsed_trades %>%
    dplyr::select(
      "timestamp",
      "type",
      "franchise" = "franchise2",
      "franchise2" = "franchise",
      "franchise1_gave_up" = "franchise2_gave_up",
      "franchise2_gave_up" = "franchise1_gave_up",
      "comments"
    ) %>%
    dplyr::bind_rows(parsed_trades) %>%
    dplyr::rename(
      "trade_partner" = "franchise2",
      "traded_for" = "franchise2_gave_up",
      "traded_away" = "franchise1_gave_up"
    ) %>%
    dplyr::arrange(dplyr::desc(.data$timestamp)) %>%
    tidyr::pivot_longer(c("traded_away", "traded_for"),
      names_to = "type_desc",
      values_to = "player_id"
    ) %>%
    tidyr::unnest("player_id")

  return(df)
}

## FREE AGENTS ##
# Create a tibble where each free agency transaction is represented on one line
#' @noRd
#' @keywords internal

.mfl_transactions_freeagent <- function(df_transactions) {
  fa_transactions <- df_transactions %>%
    dplyr::filter(.data$type == "FREE_AGENT")

  if (nrow(fa_transactions) == 0) {
    return(NULL)
  }

  parsed_fa <- fa_transactions %>%
    dplyr::select("timestamp", "type", "transaction", "franchise", "comments") %>%
    tidyr::separate("transaction", c("added", "dropped"), sep = "\\|") %>%
    dplyr::mutate_at(c("added", "dropped"), ~ stringr::str_replace(.x, ",$", "")) %>%
    tidyr::pivot_longer(c("added", "dropped"), names_to = "type_desc", values_to = "player_id") %>%
    tidyr::separate_rows(c("player_id"), sep = ",") %>%
    dplyr::filter(.data$player_id != "")

  parsed_fa %>%
    dplyr::select(
      "timestamp",
      "type",
      "franchise",
      "type_desc",
      "player_id",
      "comments"
    ) %>%
    dplyr::arrange(dplyr::desc(.data$timestamp))
}

## IR ##
# Create a tibble where each injured reserve transaction is represented on one line
#' @noRd
#' @keywords internal

.mfl_transactions_injuredreserve <- function(df_transactions) {
  ir_transactions <- df_transactions %>%
    dplyr::filter(.data$type == "IR")

  if (nrow(ir_transactions) == 0) {
    return(NULL)
  }

  parsed_ir <- ir_transactions %>%
    dplyr::select("timestamp", "type", "activated", "deactivated", "franchise", "comments") %>%
    dplyr::mutate_at(c("activated", "deactivated"), ~ stringr::str_replace(.x, ",$", "")) %>%
    tidyr::pivot_longer(c("activated", "deactivated"), names_to = "type_desc", values_to = "player_id") %>%
    tidyr::separate_rows("player_id", sep = ",") %>%
    dplyr::filter(.data$player_id != "")

  parsed_ir %>%
    dplyr::select(
      "timestamp",
      "type",
      "franchise",
      "type_desc",
      "player_id",
      "comments"
    ) %>%
    dplyr::arrange(dplyr::desc(.data$timestamp))
}

## TAXI SQUAD ##
# Create a tibble where each taxi squad player is represented on one line
#' @noRd
#' @keywords internal

.mfl_transactions_taxisquad <- function(df_transactions) {
  ts_transactions <- df_transactions %>%
    dplyr::filter(.data$type == "TAXI")

  if (nrow(ts_transactions) == 0) {
    return(NULL)
  }

  parsed_ts <- ts_transactions %>%
    dplyr::select("timestamp", "type", "demoted", "promoted", "franchise", "comments") %>%
    dplyr::mutate_at(c("promoted", "demoted"), ~ stringr::str_replace(.x, ",$", "")) %>%
    tidyr::pivot_longer(c("promoted", "demoted"), names_to = "type_desc", values_to = "player_id") %>%
    tidyr::separate_rows("player_id", sep = ",") %>%
    dplyr::filter(.data$player_id != "")

  parsed_ts %>%
    dplyr::select(
      "timestamp",
      "type",
      "franchise",
      "type_desc",
      "player_id",
      "comments"
    ) %>%
    dplyr::arrange(dplyr::desc(.data$timestamp))
}

## BBID Waiver ##
# Create a tibble where each injured reserve transaction is represented on one line
#' @noRd
#' @keywords internal

.mfl_transactions_bbid_waiver <- function(df_transactions) {
  bbid_transactions <- df_transactions %>%
    dplyr::filter(.data$type == "BBID_WAIVER")

  if (nrow(bbid_transactions) == 0) {
    return(NULL)
  }

  parsed_bbid <- bbid_transactions %>%
    dplyr::select("timestamp", "type", "franchise", "transaction", "comments") %>%
    tidyr::separate("transaction", c("player_id", "bbid_spent", "dropped"), sep = "\\|") %>%
    dplyr::mutate_at(c("player_id", "dropped"), ~ stringr::str_replace(.x, ",$", "")) %>%
    dplyr::mutate_at("bbid_spent", as.numeric)

  parsed_bbid_drops <- parsed_bbid %>%
    dplyr::select("timestamp", "type", "dropped", "franchise", "comments") %>%
    tidyr::separate_rows("dropped", sep = ",") %>%
    tidyr::pivot_longer("dropped", names_to = "type_desc", values_to = "player_id") %>%
    dplyr::filter(.data$player_id != "") %>%
    dplyr::bind_rows(parsed_bbid) %>%
    dplyr::mutate_at("type_desc", tidyr::replace_na, "added")

  parsed_bbid_drops %>%
    dplyr::select(
      "timestamp",
      "type",
      "franchise",
      "type_desc",
      "player_id",
      "bbid_spent",
      "comments"
    ) %>%
    dplyr::arrange(dplyr::desc(.data$timestamp))
}

## Will need to write functions to parse each of these, then row bind them back together afterwards.
# WAIVER - FCFS
# BBID_AUTO_PROCESS_WAIVERS
# WAIVER_REQUEST
# BBID_WAIVER_REQUEST
# SURVIVOR_PICK
# POOL_PICK