main.py

import time
import cohere
import os
from stopwords import stopwords
import training_data
from cohere.responses.classify import Example
from flask import Flask, request
from selenium import webdriver
from selenium.webdriver.common.by import By
from selenium.webdriver.chrome.options import Options
from selenium.webdriver.support import expected_conditions as EC


options = Options()
options.add_argument("--headless=new")


app = Flask(__name__)



@app.route("/acceptData", methods=["POST"])
def get_data():
    print(request.headers)

    request_data = request.form.to_dict()
    data_type = int(request_data['type'])

    inputs = []

    if data_type == 0:
        driver = webdriver.Chrome(options=options)
        driver.get(request_data["data"])
        time.sleep(3)
        try:
            EC.element_to_be_clickable((By.XPATH, "//button[contains(text(),'More')]"))
            more_buttons = driver.find_elements(By.XPATH, "//button[contains(text(),'More')]")

            for button in more_buttons:
                try:
                    driver.execute_script("arguments[0].scrollIntoView();", button)
                    button.click()
                except Exception as click_error:
                    print("Could not click button:", click_error)
        except Exception as e:
            print("Error:", e)
        inputs = [str(response.text) for response in driver.find_elements(By.CSS_SELECTOR, "div.MyEned")]
        print(inputs)
        driver.quit()
    elif data_type == 1:
        inputs = request_data["data"].split(",")
        print(inputs)
    elif data_type == 2:
        inputs = request.files["file"].read().decode("utf-8").split(",")
        for i in range(len(inputs)):
            inputs[i] = inputs[i].strip()

    co = cohere.Client(f'{os.getenv("COHERE_KEY")}')

    pos_reviews = []
    neg_reviews = []

    examples = map(lambda x: Example(text=x[0], label=x[1]), training_data.training_data)

    response = co.classify(
        model='large',
        inputs=inputs,
        examples=examples,
    )

    def class_reviews():
        res_class = response.classifications

        in_and_pred = zip(inputs, res_class)

        for i in in_and_pred:
            pred = i[1].predictions[0]
            if pred == "neutral" or pred == "positive":
                pos_reviews.append(i[0])
            else:
                neg_reviews.append(i[0])

        return pos_reviews, neg_reviews

    def tokenize(pos_reviews, neg_reviews):
        pos_words = {}
        neg_words = {}
        for i in pos_reviews:
            for j in co.tokenize(i).token_strings:
                if pos_words.get(j) is None:
                    pos_words[j] = 1
                else:
                    pos_words[j] += 1

        for i in neg_reviews:
            for j in co.tokenize(i).token_strings:
                if neg_words.get(j) is None:
                    neg_words[j] = 1
                else:
                    neg_words[j] += 1

        return pos_words, neg_words

    pos, neg = class_reviews()
    pos_dict, neg_dict = tokenize(pos, neg)
    cleaned_pos_dict = dict(sorted({k.strip(): v for (k, v) in pos_dict.items() if not k.strip().lower() in stopwords}
                                   .items(), key=lambda item: item[1], reverse=True))
    cleaned_neg_dict = dict(sorted({k.strip(): v for (k, v) in neg_dict.items() if not k.strip().lower() in stopwords}
                                   .items(), key=lambda item: item[1], reverse=True))

    return {
        'PositiveList': list(cleaned_pos_dict.items())[:5],
        'NegativeList': list(cleaned_neg_dict.items())[:5],
        'PositiveInput': pos,
        'NegativeInput': neg
    }


if __name__ == "__main__":
    app.run(debug=True, port=4000)