The OCR library itself supports only image formats as input and is "innocent" with regards to this issue ;)
The OCR library itself supports only image formats as input and is "innocent" with regards to this issue ;)