import sys

from flask import Flask, jsonify, request, render_template, Config

import pandas as pd

import joblib

import itertools

import spacy

import en_core_web_sm

import datetime

from datetime import datetime

import os

import numpy

from service.preprocessing import Preprocessing

from service.w2v_features import w2vFeatures

from service.extractResults import ShowResultS

from gensim.models import Word2Vec

app = Flask(__name__)

model_path = "modeles/"

MULTILABEL_BINAZER = "multilabel_binarizer.pkl"

TDIDF_VECTORS = "tfidf_vectors.pkl"

MODELE_LOGISTIC_TFIDF = "model_logistic_tfidf.pkl"

MODEL_SGDC_TFIDF = "model_SGDC_tfidf.pkl"

W2V_VECTORS = "w2v_vertors.bin"

MODELE_LOGISTIC_W2V = "model_logistic_w2v.pkl"

MODELE_SGDC_W2V = "model_SGDC_w2v.pkl"

# load multilabels

multilabel_binarizer = joblib.load(model_path + MULTILABEL_BINAZER, 'r')

# modele et vecteurs TFIDF

tfidf_vectorizer = joblib.load(model_path + TDIDF_VECTORS, 'r')

model_logistic_tfidf = joblib.load(model_path + MODELE_LOGISTIC_TFIDF, 'r')

model_sgdc_tfidf = joblib.load(model_path + MODEL_SGDC_TFIDF, 'r')

#modele et vecteurs word2vec

word2vec_vectorizer = Word2Vec.load(model_path + W2V_VECTORS)

model_logistic_word2vec = joblib.load(model_path + MODELE_LOGISTIC_W2V, 'r')

model_sgdc_word2vec = joblib.load(model_path + MODELE_SGDC_W2V, 'r')

#API available in http://ismail2233.pythonanywhere.com/

@app.route('/')

def index():

return render_template('index.html')

@app.route('/tagGenerators', methods=['POST'])

def tagGenerators():

print('passage controller')

result = {}

question = request.form.get("question")

question_service = Preprocessing()

cleaned_question = question_service.text_cleaner(question)

print('result cleaned {}'.format(cleaned_question))

extra_result_service = ShowResultS()

print('getTagTfIdfResult')

tag_tfid_logistic_prob = extra_result_service.getTagTfIdfResultWithProba(cleaned_question, tfidf_vectorizer, model_logistic_tfidf)

result['tfidf_logistic'] = [*tag_tfid_logistic_prob['Pred_Tags_Probab']]

result['tfidf_logistic_proba'] = tag_tfid_logistic_prob['Pred_Tags_Probab']

tag_tfid_sgdc_prob = extra_result_service.getTagTfIdfResultWithProba(cleaned_question, tfidf_vectorizer, model_sgdc_tfidf)

result['tfidf_sgdc'] = [*tag_tfid_sgdc_prob['Pred_Tags_Probab']]

result['tfidf_sgdc_proba'] = tag_tfid_sgdc_prob['Pred_Tags_Probab']

print('word2vec')

tag_w2v_logistic_prob = extra_result_service.getTagWord2vecResultWithProba(cleaned_question, word2vec_vectorizer, model_logistic_word2vec)

result['w2v_logistic'] = [*tag_w2v_logistic_prob['Pred_Tags_Probab']]

result['w2v_logistic_proba'] = tag_w2v_logistic_prob['Pred_Tags_Probab']

tag_w2v_sgdc_prob = extra_result_service.getTagWord2vecResultWithProba(cleaned_question, word2vec_vectorizer, model_sgdc_word2vec)

result['w2v_sgdc'] = [*tag_w2v_sgdc_prob['Pred_Tags_Probab']]

result['w2v_sgdc_proba'] = tag_w2v_sgdc_prob['Pred_Tags_Probab']

return result

if __name__ == "__main__":

app.config['env'] = sys.argv[1]

if app.config.get('env') == 'prod':

app.run(host='178.170.47.69', port=5000)

else:

app.run(debug=True)

# app.run(host='178.170.47.69', port=5000)