# -*- coding: utf-8 -*-
"""
Created on Thu Oct 12 15:43:55 2017

@author: Trista
"""
import csv
from nltk.tokenize import word_tokenize
from nltk.corpus import stopwords
import re
from nltk.stem.snowball import SnowballStemmer

key_word=[ 'flu','fever', 'chills','sweats','aches','pains', 'fatigue','coughing','breathing','nausea','vomiting','diarrhea','lymph','death']
stemmer = SnowballStemmer('english')
key_word1 = [stemmer.stem(w) for w in key_word]

f_text = open("sypmtom_text_1.csv", "w", encoding = 'utf-8')
f_text.write("ID,created at,location,text,symptom\n")
f_text_total = open("sypmtom_text_total_1.csv", "w", encoding = 'utf-8')
f_text_total.write("ID,created at,location,text,symptom\n")

stoplist = stopwords.words('english')
words=[]
key_text=[]
text=[]
text_total=[]
f=csv.reader(open('Microblogs.csv','r'))
for i in f:
    docs=word_tokenize(i[3])
    docs1 = [w for w in docs if w not in stoplist]
    docs2=[w.lower() for w in docs1]
    docs3=[w for w in docs2 if re.search('^[a-z]+$',w)]  
    docs4 = [stemmer.stem(w) for w in docs3]
    words.append(docs4)    
    for x in docs4:
        if x in key_word1:
            key_text.append(x)
            f_text.write("%s,%s,%s,%s,%s\n" %(i[0],i[1],i[2],str(i[3]),x))
            f_text_total.write("%s,%s,%s,%s,%s\n" %(i[0],i[1],i[2],str(i[3]),x))
        else:
            f_text_total.write("%s,%s,%s,%s,%s\n" %(i[0],i[1],i[2],str(i[3]),'NA'))
    if len(key_text)>0:
        text.append(i)
    key_text=[]
    text_total.append(i)

f_text.close()
f_text_total.close()