install.packages("jsonlite")
install.packages("httr")
install.packages("base64enc")

install.packages("namsor")

install.packages("devtools")
library(devtools)
install_github("namsor/namsor-r-sdk2")

library(namsor)

var.first.name <- 'John' # character | 
var.last.name <- 'Smith' # character | 
var.country.iso2 <- 'US' # character | 

#Infer the likely gender of a name, given a local context (ISO2 country code).
api.instance <- PersonalApi$new()
# Configure API key authorization: api_key 
api.instance$apiClient$apiKeys['X-API-KEY'] <- '83ad28e3b3f92de1494181624864bec6';
result <- api.instance$GenderGeo(var.first.name, var.last.name, var.country.iso2)
dput(result)
print(result)

#prepare workspace
if(!'devtools' %in% installed.packages()){
  install.packages(devtools)
}

if(!'readr' %in% installed.packages()){
  install.packages("readr")
}

if(!'namsor' %in% installed.packages()){
  install_github("namsor/namsor-r-sdk2")
}

#call packages
library(devtools)
library(namsor)
library(readr)

#call api
httr::set_config(httr::config(ssl_verifypeer = 0L))
api.instance <- PersonalApi$new()
api.instance$apiClient$apiKeys['X-API-KEY'] <- '83ad28e3b3f92de1494181624864bec6';

#call function
extractor <- function(names){
  
### Function calls api to assign gender by attributes ###
  
  #Input:
  
  #names - data.frame with 4 columns:
  # 1. id(char)
  # 2.first name(char)
  # 3. second name(char)
  # 4. country(char)
  
  
  #Output:
  #names - data.frame with 5 columns:
  # 1. id(char)
  # 2.first name(char)
  # 3. second name(char)
  # 4. country(char)
  # 5. gender(char)
  


## SPLIT TO LIST
names_list <- split(names, 
                    factor(rep(seq(ceiling(nrow(names)/100)), 
                               each = 100)[1:nrow(names)]))

## ADD ID
names_list <- lapply(names_list, function(x){
  data.frame('new_id' = seq(nrow(x)), 
             'id' = x[, 1],
             'first_name' = x[, 2], 
             'last_name' = x[, 3], 
             'country' = x[, 4])
})

#create r_gender
r_gender <- lapply(names_list, function(x){
  apply(x, 1, function(y){PersonalNameGeoIn$new(as.character(y[1]),y[3], y[4])})
})
#combine
r_gender <- lapply(r_gender, function(x){c(x)})
#create var gender
var_gender <- lapply(r_gender, function(x){
  BatchPersonalNameGeoIn$new(personalNames = x)})
#call api
api.instance <- PersonalApi$new()
api.instance$apiClient$apiKeys['X-API-KEY'] <- '83ad28e3b3f92de1494181624864bec6';
#get result
result_gender <- lapply(var_gender,
                        function(x){
                          api.instance$GenderFullGeoBatch(
                            batch.personal.name.geo.in=x)})
#extract gender
gender <- lapply(result_gender, function(x){lapply(x$personalNames, 
                                                   function(y){y$likelyGender})})
  
#add gender
names$gender <- unlist(gender)

#return
return(names)
}

#load small data
df1 <- read.csv('example2_csv.csv', stringsAsFactors = FALSE)

#convert to char
df1[, 1] <- sapply(df1[, 1], as.character)

#call function
df1_with_names <- extractor(df1)

#load big data
df2 <- data.frame(read_delim("example1_pipe_input.txt", 
                             "|", escape_double = FALSE, trim_ws = TRUE, 
                             col_names = FALSE))
#all char
str(df2)

#call function
df2_with_names <- extractor(df2)