Excluded Docling blocks

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

Diff Summary

{
  "parsed_text_blocks": 17,
  "final_body_blocks": 1,
  "excluded_blocks": {
    "count": 16,
    "by_reason": {
      "publisher_landing_page": 9,
      "first_page_author_or_affiliation": 2,
      "first_page_metadata": 2,
      "docling_page_header": 1,
      "first_page_front_matter_heading": 1,
      "outside_body_flow_list_item": 1
    },
    "by_role_guess": {
      "publisher_landing_page": 9,
      "affiliation": 2,
      "page_header": 2,
      "metadata": 1,
      "reference": 1,
      "unknown_text": 1
    },
    "by_risk_level": {
      "low": 15,
      "high": 1
    },
    "high_risk_count": 1,
    "medium_risk_count": 0
  },
  "char_counts": {
    "parsed_text_chars": 1994,
    "final_body_chars": 69,
    "excluded_chars": 1925
  }
}

Truncation

{
  "truncated": false,
  "message": "",
  "truncated_block_count": 0
}

Excluded Page Overlays

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。

Page 1

Page 2

Visual Assets

这是实际图表资产输出，不是审计层重新推断。

Excluded Blocks