script.py

from PIL import Image
import pytesseract
import nltk.data
import re
from newspaper import Article
from googlesearch import search


def google_search(query):  # function to google querys
    search_urls = []
    for i in search(query, tld="com", num=1, start=1, stop=1):
        search_urls.append(i)

    return search_urls


def scrape(url):  # function to web scrape
    article = Article(url)
    article.download()
    article.parse()
    article.nlp()
    text = article.text

    return text


# provide path of your image that you want to give as input.
img = Image.open("test-imgs/test-1.png")

text = pytesseract.image_to_string(img, lang="eng")  # image converted to string


tokenizer = nltk.data.load("tokenizers/punkt/english.pickle")

# print('\n-----\n'.join(tokenizer.tokenize(text)))
# text seperated into sentences.
tokenized_text = "\n \n".join(tokenizer.tokenize(text))
print(tokenized_text)

print("\nDoes the text is recognized correctly ? \n")
print("\nIf the text is not recognized correctly try using a better quality image. \n")
print("\nDo you want to proceed with this text?(y\\n) \n")

if input() == "y" or "Y":
    sentences = tokenized_text.split("\n \n")
    questions = []
    # for loop to seperate questions
    for i in range(len(sentences)):
        if (
            sentences[i].endswith("?")
            or sentences[i].startswith("What")
            or sentences[i].startswith("When")
            or sentences[i].startswith("How")
            or sentences[i].startswith("Why")
            or sentences[i].startswith("Describe")
            or sentences[i].startswith("Explain")
        ):
            questions.append(sentences[i])
            print(sentences[i])

    print("\nDo you want to proceed with these recognized questions?(y\\n) \n")
    if input() == "y" or "Y":
        print("\nPlease wait while your questions are being searched! \n")
        search_querys = []
        # for loop to get URLs of answers
        for i in range(len(questions)):
            query = questions[i]
            output = google_search(query)
            search_querys.append(output)

        solutions = {}
        # for loop to scrape answers from URLs
        for i in range(len(search_querys)):
            keys = questions[i]
            try:
                answers = scrape(search_querys[i][0])
            except Exception as e:
                values = e
            values = answers
            solutions[keys] = values

        # for loopp to display
        for x, y in solutions.items():
            print("\n******************\n")
            print(x)
            print("\n answer = \n")
            print(y)
            print("\n******************\n")

        print("\nAll questions have been answered! \n")

        print("\nDo you want the URLs of the answers?(y\\n)\n")
        if input() == "y" or "Y":
            # for loop to display URLs
            for i in range(len(search_querys)):
                print(search_querys[i][0])

            print("Thank You!")
        else:
            print("Thank You!")
    else:
        print("Thank You!")
else:
    print("Thank You!")