pyPDF2中的extractText（）函数抛出错误

Question

I am trying to extract text from PDFs so that I can analyze it but when I try to extract the text from a page I receive the following error.

Traceback (most recent call last):
File "C:\Program Files (x86)\eclipse\plugins\org.python.pydev_2.7.4.2013051601\pysrc\pydevd_comm.py", line 765, in doIt
    result = pydevd_vars.evaluateExpression(self.thread_id, self.frame_id, self.expression, self.doExec)

File "C:\Program Files (x86)\eclipse\plugins\org.python.pydev_2.7.4.2013051601\pysrc\pydevd_vars.py", line 376, in evaluateExpression
    result = eval(compiled, updated_globals, frame.f_locals)

File "<string>", line 1, in <module>

File "C:\Python33\lib\site-packages\pypdf2-1.9.0-py3.3.egg\PyPDF2\pdf.py", line 1701, in extractText
    content = ContentStream(content, self.pdf)

File "C:\Python33\lib\site-packages\pypdf2-1.9.0-py3.3.egg\PyPDF2\pdf.py", line 1783, in __init__
    stream = StringIO(stream.getData())

File "C:\Python33\lib\site-packages\pypdf2-1.9.0-py3.3.egg\PyPDF2\generic.py", line 801, in getData
    decoded._data = filters.decodeStreamData(self)

File "C:\Python33\lib\site-packages\pypdf2-1.9.0-py3.3.egg\PyPDF2\filters.py", line 228, in decodeStreamData
    data = ASCII85Decode.decode(data)

File "C:\Python33\lib\site-packages\pypdf2-1.9.0-py3.3.egg\PyPDF2\filters.py", line 170, in decode
    data = [y for y in data if not (y in ' \n\r\t')]

File "C:\Python33\lib\site-packages\pypdf2-1.9.0-py3.3.egg\PyPDF2\filters.py", line 170, in <listcomp>
    data = [y for y in data if not (y in ' \n\r\t')]

TypeError: 'in <string>' requires string as left operand, not int

The relevant code sections follow:

from PyPDF2 import PdfFileReader

for PDF_Entry in self.PDF_List:
    Pdf_File = PdfFileReader(open(PDF_Entry, "rb"))
    for pg_idx in range(0, Pdf_File.getNumPages()):
        page_Content = Pdf_File.getPage(pg_idx).extractText()
        for line in page_Content.split("\n"):
            self.Analyse_Line(line)

The error is thrown at the extractText() line.

Answer 1

It may be worth trying the latest version of PyPDF2, latest as I write this is 1.24.

With that said, I have found the extractText() feature to be very fragile. It works on some documents, fails on others. See some open issues:

https://github.com/mstamy2/PyPDF2/issues/180 and https://github.com/mstamy2/PyPDF2/issues/168

I worked around the problem by using the Poppler command-line utility pdftotext instead, both to classify a doc as image vs text and to get all the content. Has been extremely stable for me - I've run it on thousands of PDF documents. In my experience it also extracts text without further ado from protected/encrypted PDFs.

For example (written for Python 2):

def consult_pdftotext(filename):
    '''
    Runs pdftotext to extract text of pages 1..3.
    Returns the count of characters received.

    `filename`: Name of PDF file to be analyzed.
    '''
    print("Running pdftotext on file %s" % filename, file=sys.stderr)
    # don't forget that final hyphen to say, write to stdout!!
    cmd_args = [ "pdftotext", "-f", "1", "-l", "3", filename, "-" ]
    pdf_pipe = subprocess.Popen(cmd_args, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
    std_out, std_err = pdf_pipe.communicate()
    count = len(std_out)
    return count

HTH

Answer 2

You are doing two things in one line. Try to break things done to get closer to the problem. Change:

page_Content = Pdf_File.getPage(pg_idx).extractText()

into

page = Pdf_File.getPage(pg_idx)
page_Content = page.extractText()

To see where the error happens. Also run the program from the command line not from Eclipse just to make sure it is the same error. You say it happens at extractText() but this line does not show up in the traceback.