documents_to_text.py

import PyPDF2, docx, textract, os

def getOutfileName(infile):
    name_arr = infile.split('.')
    str      = ""
    for i in range(len(name_arr)-1):
        str += name_arr[i] + '.'
        
    return str + 'txt'
    
def process_pdf(infile):  
    with open(infile, 'rb') as pdfFileObj:
        pdfReader = PyPDF2.PdfFileReader(pdfFileObj)
        outFile   = getOutfileName(infile)
        with open(outFile, 'w', encoding = 'utf-8') as outfile:
            for pageNum in range(1, pdfReader.numPages):
                pageObj = pdfReader.getPage(pageNum)
                outfile.write(pageObj.extractText())
                
def process_docx(infile):
    document = docx.Document(infile)
    fullText = []
    for para in document.paragraphs:
        fullText.append('' + para.text)
    doc_string = '\n'.join(fullText)
    
    outFile = getOutfileName(infile)
    with open(outFile, 'w') as outfile:
        for string in fullText:
            outfile.write(doc_string)
            
def process_doc(infile):
    text = textract.process(infile)
    
    outFile = getOutfileName(infile)
    with open(outFile, 'wb') as outfile:
        outfile.write(text)
    

for filename in os.listdir('.'):
    if filename.endswith('.pdf'):
        process_pdf(filename)
    elif filename.endswith('.docx'):
        process_docx(filename)
    elif filename.endswith('.doc'):
        process_doc(filename)