这里是差集:Docling 全量解析 text block 减去最终会进入正文的 block。优先人工检查 risk=high / medium。
layout_review.html excluded_blocks.json excluded_blocks.tsv final_body_blocks.tsv visual_assets.tsv original.pdf
{
"parsed_text_blocks": 125,
"final_body_blocks": 23,
"excluded_blocks": {
"count": 102,
"by_reason": {
"after_back_matter_stop": 37,
"docling_page_footer": 23,
"outside_body_flow_caption": 9,
"docling_page_header": 7,
"outside_body_flow_page_footer": 5,
"inside_front_matter": 4,
"body_heading": 3,
"first_page_metadata": 2,
"implicit_abstract": 2,
"page_margin_footer": 2,
"abstract_boundary_dates": 1,
"back_matter_heading": 1,
"before_body_started": 1,
"docling_caption": 1,
"document_ui": 1,
"document_web_address": 1,
"front_matter_heading": 1,
"non_body_heading": 1
},
"by_role_guess": {
"page_footer": 31,
"reference": 29,
"caption": 10,
"page_header": 8,
"body_candidate_excluded": 5,
"back_matter_heading": 4,
"page_margin_footer": 4,
"body_heading": 3,
"abstract_candidate": 2,
"metadata": 2,
"front_matter_heading": 1,
"page_margin_header": 1,
"title_candidate": 1,
"unknown_text": 1
},
"by_risk_level": {
"low": 94,
"high": 4,
"medium": 4
},
"high_risk_count": 4,
"medium_risk_count": 4
},
"char_counts": {
"parsed_text_chars": 38796,
"final_body_chars": 20226,
"excluded_chars": 18570
}
}
{
"truncated": true,
"message": "Body extraction stopped at page 8 block #/texts/96: Conflict of Interest: The authors declare no competing fi nancial interest.. 37 following text blocks were excluded as after_back_matter_stop.",
"stop_trigger": {
"ref": "#/texts/96",
"page": 8,
"layout_order": 87,
"role_guess": "back_matter_heading",
"body_decision_reason": "back_matter_heading",
"text_preview": "Conflict of Interest: The authors declare no competing fi nancial interest."
},
"first_truncated_block": {
"ref": "#/texts/97",
"page": 8,
"layout_order": 88,
"role_guess": "back_matter_heading",
"body_decision_reason": "after_back_matter_stop",
"text_preview": "Acknowledgment. The research described in this paper is part of the Chemical Imaging Initiative at Paci fi c Northwest National Laboratory (PNNL). It was conducted under the Laboratory Directed Research and Development …"
},
"truncated_block_count": 37,
"truncated_pages": [
8
],
"by_role_guess": {
"reference": 29,
"back_matter_heading": 3,
"page_footer": 3,
"page_header": 1,
"page_margin_footer": 1
}
}
这些框显示被排除块,同时叠加真实图表资产框。青色虚线表示该 text block 被图表资产 caption 吸收;红色 STOP 是截断触发点,红色框是截断后被排除的块。