I have the following Python script. The purpose of the script is to extract text from PDF's. I use textract for that because soon I realized there is no easy way to check if a page contains an image or not. So I extract the whole text using textract.

The workflow is like this. main() parses each pdf file from a folder, I extract the text, I search for keyword strikes and then I export the result to a csv file inside folder output_results.

import textract
import os
import csv

class PdfMiner():

    path = os.getcwd() + '/folderForPdf/'
    output_path = os.getcwd() + '/output_results/'

    def __init__(self):
        pass

    def main(self):
        for self.filename in os.listdir(self.path):
            self.text = (PdfMiner().extract_text_from_pdf(self.path + self.filename))
            self.keyword_strike_dict = PdfMiner().keyword_strike(self.text)
            if bool(self.keyword_strike_dict):
                PdfMiner().output_to_csv(self.filename, self.keyword_strike_dict)


    def keyword_strike(self, text, keyword_strike_dict={}):
        '''keyword_strike function counts how many times a specific keyword occurs'''
        self.keyword_strike_dict = {}
        self.text = text
        self.keywords_list = PdfMiner().extract_keywords()
        for keyword in self.keywords_list:
            if keyword in text.decode('utf-8'):
                self.keyword_strike_dict[keyword] = text.decode('utf-8').count(keyword)
        return self.keyword_strike_dict

    def extract_keywords(self, keywords_list=None):
        '''function extract_keywords extract the keywords from file keywords.txt, into a list'''
        keywords_list = []
        with open('keywords.txt', 'r', encoding='utf8') as keywords_file:
            for keyword in keywords_file:
                keywords_list.append(keyword.strip('\n'))
        return keywords_list

    def extract_text_from_pdf(self, file_destination, text=None):
        '''extract_text_from_pdf'''
        self.file_destination = file_destination
        text = textract.process(self.file_destination, method='tesseract', language='eng', encoding='utf-8')
        return text

    def output_to_csv(self, *args, **kwargs):
        '''output_csv exports results to csv'''
        self.filename = args[0]
        self.keyword_strike_dict = args[1]
        self.output_file_path = PdfMiner().output_path + self.filename.strip('.pdf')
        with open(self.output_file_path + '.csv', 'w+', newline='') as csvfile:
            row_writer = csv.writer(csvfile, delimiter=',')
            row_writer.writerow(['keyword', 'keyword_count'])
            for keyword, keyword_count in self.keyword_strike_dict.items():
                print(keyword, keyword_count)
                row_writer.writerow([keyword, keyword_count])

if __name__ == "__main__":
    PdfMiner().main()
```

I have the following Python script. The purpose of the script is to extract text from PDFs. I use textract for that because soon I realized there is no easy way to check if a page contains an image or not. So I extract the whole text using textract.

The workflow is like this. main() parses each pdf file from a folder, I extract the text, I search for keyword strikes and then I export the result to a csv file inside folder output_results.

import textract
import os
import csv

class PdfMiner():

    path = os.getcwd() + '/folderForPdf/'
    output_path = os.getcwd() + '/output_results/'

    def __init__(self):
        pass

    def main(self):
        for self.filename in os.listdir(self.path):
            self.text = (PdfMiner().extract_text_from_pdf(self.path + self.filename))
            self.keyword_strike_dict = PdfMiner().keyword_strike(self.text)
            if bool(self.keyword_strike_dict):
                PdfMiner().output_to_csv(self.filename, self.keyword_strike_dict)


    def keyword_strike(self, text, keyword_strike_dict={}):
        '''keyword_strike function counts how many times a specific keyword occurs'''
        self.keyword_strike_dict = {}
        self.text = text
        self.keywords_list = PdfMiner().extract_keywords()
        for keyword in self.keywords_list:
            if keyword in text.decode('utf-8'):
                self.keyword_strike_dict[keyword] = text.decode('utf-8').count(keyword)
        return self.keyword_strike_dict

    def extract_keywords(self, keywords_list=None):
        '''function extract_keywords extract the keywords from file keywords.txt, into a list'''
        keywords_list = []
        with open('keywords.txt', 'r', encoding='utf8') as keywords_file:
            for keyword in keywords_file:
                keywords_list.append(keyword.strip('\n'))
        return keywords_list

    def extract_text_from_pdf(self, file_destination, text=None):
        '''extract_text_from_pdf'''
        self.file_destination = file_destination
        text = textract.process(self.file_destination, method='tesseract', language='eng', encoding='utf-8')
        return text

    def output_to_csv(self, *args, **kwargs):
        '''output_csv exports results to csv'''
        self.filename = args[0]
        self.keyword_strike_dict = args[1]
        self.output_file_path = PdfMiner().output_path + self.filename.strip('.pdf')
        with open(self.output_file_path + '.csv', 'w+', newline='') as csvfile:
            row_writer = csv.writer(csvfile, delimiter=',')
            row_writer.writerow(['keyword', 'keyword_count'])
            for keyword, keyword_count in self.keyword_strike_dict.items():
                print(keyword, keyword_count)
                row_writer.writerow([keyword, keyword_count])

if __name__ == "__main__":
    PdfMiner().main()