from io import StringIO
from pdfminer.pdfinterp import PDFResourceManager, PDFPageInterpreter
from pdfminer.converter import TextConverter
from pdfminer.layout import LAParams
from pdfminer.pdfpage import PDFPage
import os
#import sys, getopt
from pathlib import Path
import re
import time

def convert(fname, pages=None):
    #Read a pdf, returns its text content as a string
    if not pages:
        pagenums = set()
    else:
        pagenums = set(pages)

    output = StringIO()
    manager = PDFResourceManager()
    converter = TextConverter(manager, output, laparams=LAParams())
    interpreter = PDFPageInterpreter(manager, converter)

    infile = open(fname, 'rb')

    for page in PDFPage.get_pages(infile, pagenums,check_extractable=False):
        interpreter.process_page(page)
    infile.close()
    converter.close()
    text = output.getvalue()
    output.close
    return text

def convertMultiple(pdfDir, txtDir):
    if pdfDir == "": pdfDir = os.getcwd() + "/"  # if no pdfDir passed in
    for pdf in Path(pdfDir).glob(r'**/*.pdf'):   # iterate through pdfs in pdf directory
        if '-' in str(pdf):                      #rename the file if it has "-" in the firename
            pdf_legal=str(pdf).replace('-','_')
            os.rename(str(pdf),str(pdf_legal))
        else:
            pdf_legal=pdf
        fileExtension = str(pdf).split(".")[-1] #i.e. extension after the dot
        if fileExtension == "pdf":
            pdfFilename = str(pdf_legal)
            text = convert(pdfFilename)         #get string of text content of pdf
            textFilename = ((txtDir +'/'+ str(pdf).split('\\')[-1]).rstrip(".pdf"))+ ".txt"
            textFile = open(textFilename, "w", encoding='utf-8')  # make text file
            textFile.write(text)  # write text to text file
            textFile.close()
# walk through all the sub-folders with specified level(depth) to look through
def folders_in_path(path):
    if not Path.is_dir(path):
        raise ValueError("argument is not a directory")
    yield from filter(Path.is_dir, path.iterdir())

def folders_in_depth(path, depth):
    if 0 > depth:
        raise ValueError("depth smaller 0")
    if 0 == depth:
        yield from folders_in_path(path)
    else:
        for folder in folders_in_path(path):
            yield from folders_in_depth(folder, depth-1)

os.chdir(r"C:/Users/xxxx") #set the working directory
#print(os.getcwd())

lst_folders=[]
evaluators=[]

#get a list of all the folder names(evaluators)
for folder in folders_in_depth(Path.cwd(),0):
        lst_folders.append(folder)
        evaluator=str(folder).split("\\")[-1]
        evaluators.append(evaluator)

start_time = time.time() #start the timer at this point of time
if __name__ == '__main__':
    for foldername in lst_folders:                                    #iterate over each folder
        evaluator=str(foldername).split("\\")[-1]
        evaluator_folder_path = txtDir+'/'+evaluator

        try:                                                          #create a folder for each evaluator
            os.mkdir(evaluator_folder_path)
            print('folder %s has been created'%evaluator_folder_path) #mark the current folder under conversion
        except OSError:
            print ("Creation of the directory %s failed" % foldername)

        convertMultiple(str(foldername), evaluator_folder_path)
        print("--- %s seconds ---" % (time.time() - start_time))