[Feat] Add CLI support for OCR

[simont2k]

This PR:

• Add CLI to perform OCR
• options:

--input_path INPUT_PATH
                      path to input image or PDF file (default: None)
--det_arch DET_ARCH   name of the detection architecture or the model itself to use (default: db_resnet50)
--reco_arch RECO_ARCH
                      name of the recognition architecture or the model itself to use (default: crnn_vgg16_bn)
--assume_straight_pages, --no-assume_straight_pages
                      assume only straight pages without rotated textual elements (default: True)
--straighten_pages    attempt to straighten skewed pages before analysis (default: False)
--preserve_aspect_ratio, --no-preserve_aspect_ratio
                      preserve aspect ratio when resizing pages (default: True)
--symmetric_pad       apply symmetric padding (default: False)
--det_bs DET_BS       batch size for detection (default: 2)
--reco_bs RECO_BS     batch size for recognition (default: 128)
--detect_orientation  automatically detect page orientation (default: False)
--detect_language     detect language of the text (default: False)
--output OUTPUT       path to output results in JSON format (default: results.json)

• usage example: doctr-cli --input_path sample.png