Paddle ocr refined

[1192119703jzx]

This refines mainly aims to make the mmif-visualizer able to work for paddleOCR.

1. Solve the issue of failing to get id by using long_id
2. In add_timeframe method, solve the issue of failing to get start_anno, end_anno
3. In add_text_document method, solve the issue of failing to get the value property in TextDocument annotation
4. In get_ocr_frames method, solve the issue of failing to get frame.frame_num for Alignment annotation between timepoint and TextDocument.
   All the issues above combined to cause the visualizer to fail to print OCR outcome text in Thumbnails-v_1.