def get_TWord_from_TBlock()

in tpipelinegeofinder/textractgeofinder/tgeofinder.py [0:0]

• 31 lines of code
• 4 McCabe index (conditional complexity)

    def get_TWord_from_TBlock(self, block: t2.TBlock) -> TWord:
        bbox_width = block.geometry.bounding_box.width
        bbox_height = block.geometry.bounding_box.height
        bbox_left = block.geometry.bounding_box.left
        bbox_top = block.geometry.bounding_box.top
        x_min = round(bbox_left * self.doc_width)
        y_min = round(bbox_top * self.doc_height)
        x_max = round(x_min + (bbox_width * self.doc_width))
        y_max = round(y_min + (bbox_height * self.doc_height))
        page_number = block.page if block.page else 1
        if block.text:
            text = make_alphanum_and_lower_for_non_numbers(block.text)
            original_text = block.text
        elif block.selection_status:
            text = block.selection_status
            original_text = block.selection_status
        else:
            text = ""
            original_text = ""

        return TWord(text=text,
                     original_text=original_text,
                     confidence=block.confidence,
                     id=block.id,
                     text_type=block.block_type,
                     ymin=y_min,
                     xmin=x_min,
                     ymax=y_max,
                     xmax=x_max,
                     page_number=page_number,
                     doc_width=self.doc_width,
                     doc_height=self.doc_height)