def iter_semantic_graphic_for_layout_document(
self, layout_document: LayoutDocument,
extract_graphic_assets: bool) -> Iterable[SemanticGraphic]:
page_by_page_number_map = self.get_page_by_page_number_map(
layout_document)
LOGGER.debug('cv model: page_by_page_number_map=%r',
page_by_page_number_map)
has_cv_semantic_graphic: bool = False
for page_image in self.page_image_iterable:
LOGGER.debug('page_image: %r', page_image)
page_number = page_image.page_number
with PIL.Image.open(page_image.page_image_path) as image:
for semantic_graphic in self.iter_semantic_graphic_for_image(
image,
extract_graphic_assets=extract_graphic_assets,
page_number=page_number,
page=page_by_page_number_map.get(page_number)):
has_cv_semantic_graphic = True
yield semantic_graphic
if not has_cv_semantic_graphic:
LOGGER.info(
'no graphics detected using cv model, falling back to regular graphics'
)
yield from get_semantic_graphic_list_for_layout_graphic_list(
layout_document.iter_all_graphics(),
extract_graphic_assets=extract_graphic_assets)
def iter_semantic_graphic_for_layout_document(
self, layout_document: LayoutDocument,
extract_graphic_assets: bool) -> Iterable[SemanticGraphic]:
return get_semantic_graphic_list_for_layout_graphic_list(
layout_document.iter_all_graphics(),
extract_graphic_assets=extract_graphic_assets)