It's not possible to use a character whitelist with a spacebar on windows

[art0007i]

I hope I am not overlooking something obvious but from my experiments it seems not possible to use a character whitelist which has a spacebar in it on windows

I tried:
pytesseract.image_to_string(img, config=conf)
where conf is
"--psm 6 -c tessedit_char_whitelist=' 0123456789'"
'--psm 6 -c tessedit_char_whitelist=" 0123456789"'
"--psm 6 -c tessedit_char_whitelist=\' 0123456789\'"
'--psm 6 -c tessedit_char_whitelist=\" 0123456789\"'

none of those worked, it seems that the underlying code is running shlex.split with posix=False on windows. this makes it ALWAYS split on a spacebar ignoring quotes, I edited pytesseract's source code to set posix=True even on windows and it fixed my issue.

[stefan6419846]

This seems to be related to the (possibly wrongly) closed issue #569.

[tgonzalez89]

Yup, this is indeed an issue. I also had the same problem as you recently. You'll have to either modify the pytesseract code like you mentioned or do your own lightweight wrapper to run Tesseract (it's not hard, just a few lines of code). Devs please fix this! This is a really annoying issue.