Unable to process pdfs - Windows

[fraserpage]

I'm seeing the following on Windows 10. You assistance would be greatly appreciated.

Syntax Warning: Bad annotation destination
Syntax Warning: Bad annotation destination

I see about 30 lines of the above when using trying to process a pdf with pypdfocr filename.pdf.
I see the below with any usage.

WARNING: Could not execute identify to calculate DPI (try installing imagemagick?), so defaulting to 300dpi
Traceback (most recent call last):
File "", line 495, in
File "", line 492, in main
File "", line 474, in go
File "", line 480, in _convert_and_file_email
File "", line 359, in run_conversion
File "C:\Users\Virantha Ekanayake\dev\pypdfocr\build\pypdfocr_windows\out00-PYZ.pyz\pypdfocr_tesseract", line 130, in make_hocr_from_pnms
File "C:\Users\Virantha Ekanayake\dev\pypdfocr\build\pypdfocr_windows\out00-PYZ.pyz\pypdfocr_tesseract", line 96, in _is_version_uptodate
ValueError: invalid literal for int() with base 10: '00dev'

All dependencies are installed.

[flothesof]

Hi @fraserpage

I've had the same issue. In my case, the code for parsing the version string used by tesseract does not work as intended by the author. In particular, my version string was 3.05.00dev which caused the same error as you when the script tried to parse it and determine whether it was correct.

As a workaround, you can add the following bold lines to the file pypdfocr_tesseract.py found in python27\Lib\site-packages\pypdfocr:

for line in ret_output.splitlines():
            if 'tesseract' in line:
                ver_str = line.split(' ')[1]
                **if ver_str.endswith('dev'):
                    ver_str = ver_str[:-3]**

Hope this helps,

Florian

[fraserpage]

Thanks very much @flothesof! That got it working for me.

I'm still seeing the warning about imagemagick. Any clues on that one?
WARNING: Could not execute identify to calculate DPI (try installing imagemagick?), so defaulting to 300dpi

[flothesof]

Hey there!

The warning is normal, the program is just telling us it would like to do
some additional checks before adding ocr. I've found no problems with the
default resolution while using it.

Best regards
Florian
Le 14 juin 2016 21:35, "fraserpage" notifications@github.com a écrit :

Thanks very much @flothesof https://github.com/flothesof! That got it
working for me.

I'm still seeing the warning about imagemagick. Any clues on that one?
WARNING: Could not execute identify to calculate DPI (try installing
imagemagick?), so defaulting to 300dpi

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#43 (comment),
or mute the thread
https://github.com/notifications/unsubscribe/ACQFXaASjNyIuOBM-QYuthAuoKnoaWFLks5qLwJ4gaJpZM4IMFrJ
.

[fraserpage]

Got it. Thanks for your help!

[virantha]

Going to reopen and fix this in source for next release. Thanks for pointing this out, folks!

[dwmcqueen]

Hi - can the exe be fixed with this same patch?

[rasa]

@flothesof: The warning message is actually not normal, but is reporting an error on Windows. This has been fixed in #54

[qi55wyqu]

I'm still running into this problem with the word alpha in Version 0.9.1
My added fix for this (based on flothesof's answer):

checkFileEndings = ['dev', 'alpha']
for line in ret_output.splitlines():
    if 'tesseract' in line:
        ver_str = line.split(' ')[1]
        for fileEnding in checkFileEndings:
            if ver_str.endswith(fileEnding):
                ver_str = ver_str[:-len(fileEnding)]