In systats/lolR: Data Sourcing

Packages

pacman::p_load(dplyr, ggplot2, rvest, purrr, xml2, stringr, anytime, magrittr, tidyr, tibble, devtools, jsonlite)
devtools::load_all()

Load Tournament Data

load("data/matches.Rdata")
matches %>%
  count(game_id)

Scrape API

Filter for known API ids

matches_complet <- matches %>% 
  filter(!duplicated(game_id)) %>%
  filter(
    !is.na(game_id),
    !is.na(game_hash),
    !is.na(tournament_id)
  )

download json file per game

download_json <- function(x){
  json_file <- x$api_url %>%
    httr::GET() %>%
    httr::content(., "text", encoding = "UTF-8")

  cat("\n Content avaible? ")
  cat(!is.null(json_file))
  cat(" ")
  cat("for ")
  cat(x$game_id)  

  if(!is.null(json_file)){
    raw <- jsonlite::fromJSON(json_file)
    # json_file <- jsonlite::toJSON(raw)
    write(json_file, file = paste0("data/games/game_", raw$gameId, ".json"))
  }
}

download_json_pos <- purrr::possibly(download_json, otherwise = NULL)
download_json_pro <- lolR::progressively(download_json_pos, nrow(matches_complet))

matches_complet %>% 
  split(1:nrow(.)) %>%
  #.$[1:2] %>% 
  walk(~download_json_pro(.x))

jsonlite::read_json("data/games/game_670035.json")

• Athor Jason APIS

systats/lolR documentation built on May 31, 2019, 6:17 p.m.