in src/chug/doc/doc_read_processor.py [0:0]
def _expand_anno(self, anno, count: int):
expanded_annos = []
for i in range(count):
sample = {}
for k, v in anno.items():
if k == '_parse':
gt_parse = {}
gt_parse['num_pages'] = v['num_pages']
gt_parse['page_indices'] = [v['page_indices'][i]]
gt_parse['page_text'] = [v['page_text'][i]]
sample[k] = gt_parse
else:
sample[k] = v[i] if isinstance(v, (list, tuple)) else v
expanded_annos.append(sample)
return expanded_annos