AttributeError: 'unicode' object has no attribute 'seek'

[tringger]

Hi Virantha,

I really enjoy what you've built. Thank you very much. I'm reporting an issue I've run into when attempting to copy more than one PDF to the watch folder while using a config.yaml file. Here is the complete output from my command line, including the command. Also, I have imagemagick installed. Even though the message says it can't find "identify", I can run "identify" from the command line without any issues. Also, if I don't use a watch folder, I can run pypdfocr on a single file without any issue.

C:\Users\tringger006\Documents\ARCA\2014\OCR>pypdfocr -w C:\Users\tringger006\Documents\ARCA\2014\OCR -f -c config.yaml -d

Filing of PDFs is enabled

• 2 target filing folders
• 2 keywords

Starting to watch for new pdfs in C:\Users\tringger006\Documents\ARCA\2014\OCR
Starting conversion of C:\Users\tringger006\Documents\ARCA\2014\OCR\Document1.pdf

WARNING: Could not execute identify to calculate DPI (try installing imagemagick?), so defaulting to >300dpi
Completed conversion successfully to >C:\Users\tringger006\Documents\ARCA\2014\OCR\Document1_ocr.pdf
Traceback (most recent call last):
File "C:\Python27\Scripts\pypdfocr-script.py", line 9, in
load_entry_point('pypdfocr==0.7.4', 'console_scripts', 'pypdfocr')()
File "c:\Python27\lib\site-packages\pypdfocr\pypdfocr.py", line 411, in main
script.go(sys.argv[1:])
File "c:\Python27\lib\site-packages\pypdfocr\pypdfocr.py", line 396, in go
filing = self.file_converted_file(ocr_pdffilename, pdf_filename)
File "c:\Python27\lib\site-packages\pypdfocr\pypdfocr.py", line 331, in file_converted_file
filed_path = self.pdf_filer.move_to_matching_folder(ocr_pdffilename)
File "c:\Python27\lib\site-packages\pypdfocr\pypdfocr_pdffiler.py", line 65, in move_to_matching_folder
for page_text in self.iter_pdf_page_text(filename):
File "c:\Python27\lib\site-packages\pypdfocr\pypdfocr_pdffiler.py", line 42, in iter_pdf_page_text
reader = PdfFileReader(filename)
File "c:\Python27\lib\site-packages\PyPDF2\pdf.py", line 800, in init
self.read(stream)
File "c:\Python27\lib\site-packages\PyPDF2\pdf.py", line 1242, in read
stream.seek(-1, 2)
AttributeError: 'unicode' object has no attribute 'seek'

Here are the contents of my config.yaml file, stored in the watch directory:

target_folder: "docs/filed"
default_folder: "docs/filed/manual_sort"
original_move_folder: "docs/originals"

folders:
default_fees:
- appraisal
other:
- text

Thanks so much, again.

[virantha]

Thanks for reporting this! Let me take a look at it...

[nextechinc]

I'm having the exact same problem.

[virantha]

I am unable to reproduce this under mac/linux. Could you send me the pdf you're working with?

[virantha]

Also, please make sure you're running version 0.7.5

[virantha]

Closing issue as can't reproduce.