Method _identify_floating_images

pdf2docx/layout/Blocks.py:370–385

Identify floating image lines and convert to ImageBlock.

(self, float_image_ignorable_gap:float)

    # internal methods
    # ----------------------------------------------------------------------------------
    def _identify_floating_images(self, float_image_ignorable_gap:float):
        '''Identify floating image lines and convert to ImageBlock.'''
        # group lines by connectivity
        groups = self.group_by_connectivity(dx=-float_image_ignorable_gap, dy=-float_image_ignorable_gap)
        
        # identify floating images
        for group in filter(lambda group: len(group)>1, groups):
            for line in filter(lambda line: line.image_spans, group):
                float_image = ImageBlock().from_image(line.spans[0])
                float_image.set_float_image_block()
                self._floating_image_blocks.append(float_image)

                # remove the original image line from flow layout by setting empty bbox
                line.update_bbox((0,0,0,0))

        return self

    def _remove_overlapped_lines(self, line_overlap_threshold:float):
        '''Delete overlapped lines.