# Imports first

import numpy as np
import pandas as pd
from os import path
from PIL import Image
from wordcloud import WordCloud, STOPWORDS, ImageColorGenerator
import matplotlib.pyplot as plt

# Helper function

def save_wordcloud_from_text(text,file):
  print( "INFO: about to generate the WordCloud for {}".format(file) )
  wordcloud = WordCloud(font_path="/Ubuntu-Th.ttf", width=1920, height=1080).generate(text)
  print( "INFO: setting the figure size for {}".format(file) )
  plt.figure(figsize=(200,100), facecolor='k')
  print( "INFO: setting tight layout for {}".format(file) )
  plt.tight_layout(pad=0)
  print( "INFO: performing the 'imshow' for {}".format(file) )
  plt.imshow(wordcloud, interpolation="bilinear")
  print( "INFO: turning off the axes for {}".format(file) )
  plt.axis("off")
  print( "INFO: saving file {}".format(file) )
  plt.savefig(file, facecolor='k', bbox_inches='tight')

# Main script

file_input    = "snake_public.csv"
file_users    = "users.png"
file_messages = "messages.png"

print( "INFO: about to read in CSV file." )
df = pd.read_csv(file_input, low_memory=False)

print( "INFO: about to extract users column." )
users    = df.ut.dropna(how='any')
text_users = "\n".join(review for review in users.astype(str))
save_wordcloud_from_text( text_users, file_users )

print( "INFO: about to extract messages column." )
messages = df.m.dropna(how='any')
text_messages = "\n".join(review for review in messages.astype(str))
save_wordcloud_from_text( text_messages, file_messages )