Looping over files to create a dataframe

Question

As part of my NLP project at work, I want to loop over all files that are either PDF of docx in the same directory. The end purpose is to create a dataframe with text content of the files in one column and filename in another column. So I wrote a loop to open them up, extract the text, add it to a list as below:

import docx
import PyPDF2


    #wordfile
    def ReadingText(filename):
        doc=docx.Document(filename)
        comp = []
        for par in doc.paragraphs:
            comp.append(par.text)
        return '\n' .join(comp)
    
    #PDF files
    def Readingpdf(pdfname):
        pdfRead=PyPDF2.PdfFileReader(pdfname)
        comp = ""
        for i in range(pdfRead.getNumPages()):
            comp += pdfRead.getPage(i).extractText()
        return comp

After that I will specify directory and perform the loop as below:

directory= "my directory"
all_together = []
name = []

for filename in os.listdir(directory):
    if filename.endswith(".docx"):
        f = ReadingText(filename)
        all_together.append(f)
        name.append(filename)
    elif filename.endswith(".pdf"):
            try:
                pdf = Readingpdf(filename)
                all_together.append(pdf)
                name.append(filename)
            except:
                pass

            
    df = pd.DataFrame({"article":name,"text":all_together})

        
 
df.head(10)

It is doing the job, but I really want to develop myself so I want to know your opinion - be as harsh as you can.