backend.R

library(dplyr)
library(ggplot2)
library(lubridate)
#library(plotly)
#library(leaflet)
.root <- rprojroot::is_rstudio_project$make_fix_file()
xfun::download_file("https://www.dropbox.com/s/2rryka3cprtgfok/heritageDataRevised.Rdata?dl=1", mode="wb")
load("heritageDataRevised.Rdata")


# define user, heritage structure

## define complete structure
user <- list()
user <- 
  list(
    sessionData = 
      list(
        county_chosen = character(0),
        filtered_data = vector("list",0),
        game = 
          list(
            puzzle_guess = vector("list",0)
          )
      ),
    checkIns = vector("list",0),
    visits = vector("list",0)
  )
# heritage structure
heritage <-
  list(
    data = heritageData,
    show_county = function(){},
    filter = list(
      countyChosen = function(county){}
    ),
    game = list(
      puzzle_guess = function(){}
    ),
    validate_checkIn = function(returnedGPS){},
    track = function(list_tracks){}
  )


# show by county --------------------------------------------
# App function: user wants to zoom into map that shows only the point of interests belong to certain county
# when select choose by county on APP, it calls
heritage$show_county() -> user$sessionData$county_chosen
# which returns a character vector of counties on screen
#  1. prompt names of counties that he can choose on screen
#  2. once chosen, return user's choice and save in the according object element

# Then program continues
heritage$filter$countyChosen(user$sessionData$county_chosen) -> 
  user$sessionData$filtered_data

# countyChosen returns caseIds that fit the chosen criterion.
# The backend return user$sessionData$filtered_data to the frontend which will redraw the map

# play puzzle -------------------------------------------------------------
# when user selects play puzzle guess, it calls
source("/Users/frankchao/Desktop/109-2-app101/weekly_progress/frank/puzzle.R")
heritage$game$puzzle_guess <- puzzle_guess
heritage$game$puzzle_guess() -> game_result

#  the return is a list of elements: 
#  timestamp=a character of ISO8601 time when the user starts to play
#  answer=a caseId of correct answer
#  options=a character vector of caseIds of heritage options
#  and user_choice=a caseId of user's choice among options
# game_result$timestamp
# game_result$answer
# game_result$options
# game_result$user_choice

# record result
user$sessionData$game$puzzle_guess <- 
  append(
    user$session$game$puzzle_guess,
    list(game_result)
  )






# check in ----------------------------------------------------------------
# user check in in the webapp UI, 
#browseURL("https://tpemartin.github.io/109-2-app101/checkin")
# the frontend returns a numeric vector of GPS
# c(latitude=..., longitude=...)
returnedGPS = c(lon=120.4407, lat=23.47734)
heritage$validate_checkIn <- function(returnedGPS){
         time_start <- format_ISO8601(now(),usetz = T)
         location <-{
            lon  <-  heritageData$longitude
            lat  <-  heritageData$latitude
            lon[is.na(lon)] <- 0
            lat[is.na(lat)] <- 0
            matrix(lon) -> col1
            matrix(lat) -> col2
            cbind(col1,col2) 
         }
         if(T %in% (sp::spDistsN1(location,returnedGPS,longlat=T) <= 0.1)){
                return(
                      list(
                          timestamp = time_start ,
                          caseId = {
                                     which.min(sp::spDistsN1(location,returnedGPS,longlat=T)) ->
                                              index
                                     heritageData$caseId[index]
                                   }
                      )
                )
         }else{
                return(NULL)
         }
}
heritage$validate_checkIn(returnedGPS) -> checkInResult
# which check if the returnGPS is within 100 meters of any heritage site, 
# If not, checkInResult = NULL;
# If yes, 
# checkInResult = list(
#   timestamp= a string of ISO8601 time when the user checks in,
#   caseId= the closest site's caseId
# )
# sp::spDistsN1 can calculate km distance for you
if(!is.null(checkInResult)){
  user$checkIns <- append(
    user$checkIns, list(checkInResult)
  )
}


# track -------------------------------------------------------------------
# when user turn on tracking in the frontend, user's phone will track his GPX 
source(
  file.path(.root(),"R/field_trips_all.R")
)
# return user's list_tracks


heritage$track(list_tracks) -> 
  newTraces
# where newTraces is a list of length(list_tracks) for element 1, it transform list_trackes[[1]]
#list_tracks[[1]][c("lon","lat")] -> trace1
#newTraces <- list()
# newTraces[[1]] <- 
#   list(
#     timestamp=date of visit
#     trace=trace1,
#     caseId=
#   )

user$visits <- append(
  user$visits, newTraces
)