---
name: delivery-assurance
description: "Meta-model-agent 完成编译、版式、匿名、页数、引用与提交前合规验收。适用于提交质量验收。"
---
# 提交质量验收
## 数据准备一致性验收
有数据时，核对论文独立“数据预处理”小节、`全部结果.json.data_preparation`、`数据/processed/` 文件哈希和模型代码读取路径一致，并确认处理前后质量统计与泄漏说明真实可复核。无数据时确认存在明确豁免且没有伪造的处理后数据或清洗结论。
## 模型身份与摘要一致性验收
逐问核验审计身份与论文表达卡。正式名称必须显式暴露线性/整数/鲁棒/随机规划、回归、时间序列、状态空间、车辆路径、网络流、多指标决策等标准结构；论文展示名称应简洁并保留主要结构。比较、验证或应用类问题必须说明继承关系，不得强造独立模型。摘要和正文先说明模型或继承关系，再说明核心方法；关键词至少包含一个标准模型族。
`建模报告.md` 的审计身份与 `全部结果.json.model_identity` 必须逐字一致；摘要和正文与论文表达卡保持语义一致，不得复制身份卡字段标签。最终摘要还必须拦截内部术语、完整算法链、未解释方案缩写、数字流水账和缺失的模型边界。
## 发布声明、公式与页面构成验收
- 核对 `全部结果.json.publication_claims` 的来源键、派生关系与正文显示值；同一指标在摘要、正文、图表、表格和结论中不得冲突。
- 核对真实问题数、逐问章节文件和主文件 `\input/\include` 一一闭合，禁止章节存在但最终 PDF 漏问。
- 读取 `latex-math.md`，检查原生公式、符号与量纲、`\eqref/\ref/\cref`、重复标签和硬编码图表公式编号。
- 读取 `page-composition.md` 并在实际 PDF 上执行页面审计。编译成功、页数合规和页面质量是三个独立门禁。
- Overfull box、裁切、标题孤悬、图题分离和不可读小字在冠军模式下不得带入终稿；低占用页面必须人工确认是否来自章节自然结束，而不是用空白或大图填页。
## 必读的版面子协议
- 始终读取 `参考资料/paper-layout/core.md` 与 `body-density.md`。
- PDF/LaTeX 验收读取 `latex-figures.md`。
- PDF/LaTeX 验收同时读取 `latex-math.md` 与 `page-composition.md`。
- DOCX 验收读取 `docx-figures.md`。
- 子协议中的尺寸和页数语义优先于本文件中的旧示例。
DOCX 路线使用 `python 工具/docx_export.py --workspace .` 生成 `论文/数模论文.docx` 和 `论文/docx_report.json`。不得对 DOCX 路线执行 XeLaTeX/BibTeX；验收以 Markdown 正文、DOCX 包、尺寸报告以及可选的 LibreOffice PDF 预览为准。
## 竞赛专用硬门禁
编译前必须读取 `状态/工作流状态.json` 中的竞赛 Profile 和 `参考资料/竞赛规则.md`：
- CUMCM：核验当届官方通知、匿名电子正文、模板、引用、图表及编号/附件要求。
- 51MCM：核验 `51mcmthesis`、`withoutpreface`、报名号/题号/组别字段，以及正文不存在学校、成员、邮箱和电话信息。
- MCM/ICM：核验全英文、12pt、Summary Sheet、已填写 Control Number、A--F 题号、无个人/学校身份信息，完整 PDF 不超过 25 页。
编译成功不能覆盖竞赛 Profile 门禁失败。
## 稳定执行契约
- **执行目标**：完成论文编译、版式、匿名、页数、引用与提交文件的终局验收。
- **调用参数**：[paper-directory]。
- **权威输入**：论文/论文正文.tex、论文/章节/、图表资产、模板规则与竞赛要求。
- **允许交付**：论文/数模论文.pdf、论文/编译日志.log及必要的验收记录；仅可做有边界的修复。
- **禁止写入**：不得越权修改已冻结的上游事实、用户原始文件或本协议未授权的目录。
- **可用工具边界**：Bash(*), Read, Write, Edit, Grep, Glob。
- **最小交付**：PDF可打开、编译无关键错误、引用闭合、模板完整、匿名与页数合规。
- **恢复入口**：优先读取当前工作、状态记录和已有产物，从最近一次通过门禁的位置继续。
- **失败回退**：最多按既定次数修复编译；模型或结果问题必须回退上游，不得在验收阶段擅改事实。
- **收口顺序**：先核对输入，再完成产物，再运行本环节门禁，最后登记状态；门禁未通过不得宣告完成。
Compile and validate: **$ARGUMENTS**
## 运行参数
- **ENGINE = `xelatex`**
- **MAX_COMPILE_ATTEMPTS = 3**
- **PAPER_DIR = `论文/`**
- **MAX_PAGES** / **COMPETITION** — From Additional Parameters.
## 执行流程
### 工作节点 1：Validate environment
```bash
if ! which xelatex 2>/dev/null; then
echo "xelatex not found, attempting install..."
if which miktex 2>/dev/null; then
miktex packages install xetex ctex xecjk gbt7714 fontspec
miktex fndb refresh
elif which initexmf 2>/dev/null; then
initexmf --set-config-value=[MPM]AutoInstall=1
fi
fi
which xelatex && which bibtex && echo "ready" || echo "xelatex/bibtex not found"
fc-list :lang=zh | head -5
kpsewhich gbt7714.sty 2>/dev/null || echo "gbt7714.sty not found (will auto-install on first compile)"
```
### 工作节点 2：Pre-compile cleanup
```bash
if [ -f "工具/compile_utils.sh" ]; then
bash 工具/compile_utils.sh 论文/
elif [ -f "skills/shared-scripts/compile_utils.sh" ]; then
bash skills/shared-scripts/compile_utils.sh 论文/
else
echo "compile_utils.sh not found, manual cleanup needed"
fi
```
This utility auto-handles: special chars cleanup, table format fixes, includegraphics path correction (`图表/` → `../图表/`), hidelinks, 图表/图表/ nesting, math_commands conflicts, wide table resizebox wrapping, narrow table resizebox removal, light-color text fixes, TikZ library injection.
If script not found, perform these steps manually.
**⛔ 表格行结束符修复（Misplaced \noalign 的根因）：**
```bash
# 检测并修复 tabular/longtable 中的单 \ 行结束符（应该是 \\）
# 这是 heredoc 不加引号导致 \\ 被转义为 \ 的常见问题
for f in 论文/章节/*.tex; do
[ -f "$f" ] || continue
# 检测：tabular 环境内，行末只有单个 \ 后跟换行（应该是 \\）
if grep -P '(? /dev/null 2>&1; then
echo "⚠ $(basename $f): 可能有表格行结束符问题（单 \\ 应为 \\\\）"
# 在 tabular/longtable 环境内，把行末的单 \ 替换为 \\
python3 -c "
import re
with open('$f', 'r', encoding='utf-8') as fh:
content = fh.read()
# 只在 tabular/longtable 环境内修复
def fix_table_endings(match):
table = match.group(0)
# 把数据行末尾的单 \ (后跟换行) 替换为 \\\\
fixed = re.sub(r'(?<=&[^&\n]*)\\\s*\n', r'\\\\\\\\\n', table)
return fixed
for env in ['tabular', 'longtable']:
pattern = r'(\\\\begin\{' + env + r'[*]?\}.*?\\\\end\{' + env + r'[*]?\})'
content = re.sub(pattern, fix_table_endings, content, flags=re.DOTALL)
with open('$f', 'w', encoding='utf-8') as fh:
fh.write(content)
" 2>/dev/null
fi
done
```
Additionally validate ref/label matching and embed missing visuals:
```bash
mkdir -p 临时文件
grep -oh '\\ref{[^}]*}' 论文/章节/*.tex 论文/论文正文.tex 2>/dev/null | sort -u > 临时文件/_refs.txt
grep -oh '\\label{[^}]*}' 论文/章节/*.tex 论文/论文正文.tex 2>/dev/null | sort -u > 临时文件/_labels.txt
comm -23 <(sed 's/\\ref/\\label/g' 临时文件/_refs.txt) 临时文件/_labels.txt > 临时文件/_missing_labels.txt
cat 临时文件/_missing_labels.txt
```
**⛔ 中文引号统一为全角引号（避免 PDF 出现 ''乱码或两个堆叠反引号）：**
中文论文用 xeCJK，全角引号 `"..."` 会自动化渲染为漂亮的对称弯引号。
LaTeX 风格 `` ``...'' `` 在中文字体下会呈现成两个堆叠的反引号，很丑。
ASCII 直引号 `"..."` 在 LaTeX 中渲染为右右引号 `''...''`。
统一替换为全角引号。
```bash
# 把所有错误引号统一为全角引号
for f in 论文/章节/*.tex; do
[ -f "$f" ] || continue
python3 -c "
import re
content = open('$f', 'r', encoding='utf-8').read()
# 保护数学环境
placeholders = []
def stash(m):
placeholders.append(m.group(0)); return f'\\x00M{len(placeholders)-1}\\x00'
content = re.sub(r'\\\$[^\\\$]*\\\$', stash, content)
content = re.sub(r'\\\\\\[.*?\\\\\\]', stash, content, flags=re.DOTALL)
# 保护 \begin{verbatim}/lstlisting 等代码环境
content = re.sub(r'\\\\begin\{(verbatim|lstlisting|minted)\}.*?\\\\end\{\\1\}', stash, content, flags=re.DOTALL)
# 1. LaTeX 风格 ``...'' → 全角双引号
content = re.sub(r\"\`\`([^\`'\\\\n]+?)''\", '\u201c\\\\g<1>\u201d', content)
# 2. ASCII 直引号 \"...\" → 全角双引号（成对处理）
parts = content.split('\"')
if len(parts) > 2:
result = parts[0]
for i, p in enumerate(parts[1:], 1):
result += ('\u201c' if i % 2 == 1 else '\u201d') + p
content = result
# 还原
for i, ph in enumerate(placeholders):
content = content.replace(f'\\x00M{i}\\x00', ph)
open('$f', 'w', encoding='utf-8').write(content)
" 2>/dev/null
done
```
If labels are missing, find corresponding figure/table code in `图表/*.tex` and embed into the correct section file.
Also verify compile_utils.sh output for "UNEMBEDDED" warnings against `figure_manifest.json`. Only an item selected for publication is a blocking omission.
**MANDATORY FIX LOOP — do NOT proceed to compilation until every `publish=true` figure and every table selected by the manuscript embedding plan is embedded:**
```bash
UNEMBED=0
# Validate PDF visuals
for pdf in 图表/*.pdf; do
[ -f "$pdf" ] || continue
bn=$(basename "$pdf")
python3 -c "import json,sys; d=json.load(open('图表/figure_manifest.json',encoding='utf-8')); sys.exit(0 if any(x.get('publish') and x.get('path','').replace('\\\\','/').endswith('/'+sys.argv[1]) for x in d.get('figures',[])) else 1)" "$bn" || continue
grep -rq "$bn" 论文/章节/*.tex 论文/论文正文.tex 2>/dev/null || { echo "UNEMBEDDED PDF: $bn"; UNEMBED=$((UNEMBED+1)); }
done
# Validate only TABLE_*.tex files selected by the manuscript embedding plan
for tbl in 图表/TABLE_*.tex; do
[ -f "$tbl" ] || continue
bn=$(basename "$tbl")
python3 -c "import json,sys; d=json.load(open('图表/figure_manifest.json',encoding='utf-8')); sys.exit(0 if any(x.get('publish') and x.get('path','').replace('\\\\','/').endswith('/'+sys.argv[1]) for x in d.get('tables',[])) else 1)" "$bn" || continue
# Validate if any label from this table file appears in sections
for lbl in $(grep -oh '\\label{[^}]*}' "$tbl" 2>/dev/null); do
grep -rq "$lbl" 论文/章节/*.tex 论文/论文正文.tex 2>/dev/null || { echo "UNEMBEDDED TABLE: $lbl (from $bn)"; UNEMBED=$((UNEMBED+1)); }
done
done
# Raw 图表引用.tex labels are diagnostic; the published asset loops above remain authoritative
echo "Total unembedded: $UNEMBED"
```
If UNEMBED > 0, you MUST fix all selected publication items before compiling. For each unembedded item:
- **PDF figure**: copy the `\begin{figure}...\end{figure}` block from `图表/图表引用.tex` into the target section
- **TABLE_*.tex**: copy the `\begin{table}...\end{table}` block from `图表/TABLE_*.tex` into the target section (use `\input{../图表/TABLE_xxx.tex}` or paste the tabular code directly)
- Add 1-2 sentences of lead-in text before and 3-5 sentences of analysis after each embedded item
- Re-run the count verify above — repeat until UNEMBED = 0
**Do NOT compile with unembedded publication figures or selected正文 tables. Unpublished diagnostic assets are allowed to remain outside the paper.**
```bash
# Diagnose generated PDFs; only publish=true omissions are blocking
for pdf in 图表/*.pdf; do
[ -f "$pdf" ] || continue
bn=$(basename "$pdf")
python3 -c "import json,sys; d=json.load(open('图表/figure_manifest.json',encoding='utf-8')); sys.exit(0 if any(x.get('publish') and x.get('path','').replace('\\\\','/').endswith('/'+sys.argv[1]) for x in d.get('figures',[])) else 1)" "$bn" || continue
grep -rq "$bn" 论文/章节/*.tex 论文/论文正文.tex 2>/dev/null || echo "⚠ $bn not referenced"
done
```
### 执行节点 3：Compile (manual steps, no latexmk)
```bash
cd 论文/
xelatex -interaction=nonstopmode 论文正文.tex
bibtex 论文正文
xelatex -interaction=nonstopmode 论文正文.tex
xelatex -interaction=nonstopmode 论文正文.tex
```
### 执行节点 4：Error fix loop (MANDATORY — do NOT skip)
After each compilation, verify `编译日志.log` for CRITICAL errors. **You MUST fix ALL errors before declaring compilation complete.**
```bash
# Count critical errors
MATH_ERR=$(grep -c 'Bad math environment delimiter\|Missing \$ inserted\|begin{document} ended by' 论文/编译日志.log 2>/dev/null || echo 0)
LR_ERR=$(grep -c 'Not allowed in LR mode' 论文/编译日志.log 2>/dev/null || echo 0)
UNDEF_CS=$(grep -c 'Undefined control sequence' 论文/编译日志.log 2>/dev/null || echo 0)
TOTAL_ERR=$((MATH_ERR + LR_ERR))
echo "Math errors: $MATH_ERR, LR mode errors: $LR_ERR, Undefined CS: $UNDEF_CS"
if [ "$TOTAL_ERR" -gt 0 ]; then
echo "CRITICAL: $TOTAL_ERR errors — MUST FIX prior to proceeding"
# Show error locations
grep -B2 'Bad math\|Missing \$ inserted\|begin{document} ended\|Not allowed in LR mode' 论文/编译日志.log | grep -E '^\./|^l\.' | head -20
fi
```
**Error fix rules (iterate up to 5 times, not 3):**
1. **Math environment errors** (`Bad math environment delimiter`, `Missing $ inserted`, `\begin{document} ended by \end{equation}`):
- Read the error location from 编译日志.log (e.g., `./章节/3_model_theory.tex:42`)
- Open the file and find the broken math: usually `\X(t)$` should be `$X(t)$`, or `\mu$` should be `$\mu$`
- Common cause: a sed/cleanup script stripped the opening `$` but left the closing `$`
- Fix: ensure every math expression has matching `$...$` or `\[...\]` delimiters
- **Do NOT use broad sed patterns to fix math** — read each error location and fix individually
2. **LR mode errors** (`Not allowed in LR mode`):
- Usually caused by `\begin{figure}` or `\begin{table}` inside a paragraph without proper separation
- Fix: add `\par` or blank line before the float environment
3. **Undefined control sequence**:
- Missing package → add `\usepackage{xxx}` to 论文正文.tex preamble
- Typo in command → fix the command name
4. **BibTeX failures**:
- If BibTeX fails because of earlier LaTeX errors, fix the LaTeX errors first, then recompile
- Verify that `\bibliography{references}` and `\bibliographystyle{plainnat}` exist in 论文正文.tex
- Verify that references.bib has no syntax errors (unmatched braces, missing commas)
**After each fix, recompile and reverify:**
```bash
cd 论文/
xelatex -interaction=nonstopmode 论文正文.tex
bibtex 论文正文 2>&1 | tail -5
xelatex -interaction=nonstopmode 论文正文.tex
xelatex -interaction=nonstopmode 论文正文.tex
cd ..
# Reverify
MATH_ERR=$(grep -c 'Bad math environment delimiter\|Missing \$ inserted' 论文/编译日志.log 2>/dev/null || echo 0)
echo "Remaining math errors: $MATH_ERR"
```
**⛔ Do NOT proceed to Step 5 until MATH_ERR = 0 and LR_ERR = 0.** BibTeX will also fail if there are LaTeX errors upstream — fix LaTeX first.
When fixing errors in 论文正文.tex, only fix the specific error (e.g., add a missing package, fix a typo). Avoid rewrite or restructure 论文正文.tex — the template's preamble, cover page, page margins, section numbering format, and header/footer settings must remain unchanged.
### 执行节点 5：Post-compile checks
```bash
bash 工具/compile_check.sh 论文/ 2>/dev/null || bash skills/shared-scripts/compile_check.sh 论文/
```
The script checks: PDF existence/size, undefined references, overfull hbox, TOC, abstracts, bibliography entries, citation count, citation format (上标/顺序/合并), unused figures, figure stacking, TikZ diagram presence.
**⛔ MANDATORY: 引用格式问题必须修复（不可忽略）：**
compile_check.sh 会检查 3 类引用格式问题，任何一类 FAIL/WARN 都必须修复：
1. **上标格式 FAIL** → 如果 bibliographystyle 是 `plain/plainnat/unsrt` 但 `\cite{}` 没有上标
- 修复方法A：改 bibliographystyle 为 `gbt7714-numerical` 或 `plainnat` + `\usepackage[numbers,square,super]{natbib}`
- 修复方法B：把正文里所有 `\cite{x}` 改为 `\upcite{x}` 或 `\textsuperscript{\cite{x}}`
2. **多引用顺序 WARN** → `\cite{c,a,b}` 但全文 a 先出现
- 修复方法：改成 `\cite{a,b,c}`（按全文首次出现顺序）
- 用以下 bash 找出所有需修复的位置：
```bash
grep -rnP '\\(up)?cite\{[^}]+\}' 论文/章节/*.tex 论文/论文正文.tex
```
3. **连续引用未合并 WARN** → `\cite{a}\cite{b}` 或 `\cite{a} \cite{b}`
- 修复方法：合并为 `\cite{a,b}`
- 批量修复：
```bash
# 找出全部需合并的位置
grep -rnoP '\\(up)?cite\{[^}]+\}\s*\\(up)?cite\{[^}]+\}' 论文/章节/*.tex 论文/论文正文.tex
```
**⛔ 修复后必须重新编译验证：** `xelatex → bibtex → xelatex → xelatex`，然后重新运行 compile_check.sh 确认无 FAIL。
Additionally verify:
```bash
# List of visuals / list of tables (stats competition requirement)
[ -s 论文/论文正文.lof ] && echo "✅ 插图清单已产出" || echo "⚠ 插图清单为空"
[ -s 论文/论文正文.lot ] && echo "✅ 表格清单已产出" || echo "⚠ 表格清单为空"
```
**Template format verification (stats competition)**:
```bash
echo "=== 模板格式校验 ==="
if grep -q 'stats\|统计建模' 论文/论文正文.tex 2>/dev/null || [ "$COMPETITION" = "stats" ]; then
# Validate page margins (is expected to be 2.54cm top/bottom, 3.17cm left/right)
grep -q '2.54cm' 论文/论文正文.tex && echo "✅ 页边距无误" || echo "⚠ 页边距可能不对（应为上下2.54cm，左右3.17cm）"
# Validate Chinese section numbering
grep -q 'chinese{section}' 论文/论文正文.tex && echo "✅ 中文章节编号" || echo "⚠ 缺少中文章节编号（一、二、三...）"
# Validate cover page
grep -q '参赛学校\|参赛作品\|作品编号' 论文/论文正文.tex && echo "✅ 封面出现" || echo "⚠ 缺少封面"
# Verify abstract format (should be \section*{摘要}, not \begin{abstract})
grep -q 'section\*.*摘要' 论文/论文正文.tex && echo "✅ 摘要格式无误" || echo "⚠ 摘要格式可能不对"
# Validate natbib (stats uses natbib, not gbt7714)
grep -q 'natbib' 论文/论文正文.tex && echo "✅ natbib 引用格式" || echo "⚠ 缺少 natbib（统计建模应用 natbib）"
# Validate no header line
grep -q 'headrulewidth.*0pt' 论文/论文正文.tex && echo "✅ 无页眉线" || echo "⚠ 可能有页眉线"
# Validate listoffigures / listoftables
grep -q 'listoffigures' 论文/论文正文.tex && echo "✅ 插图清单命令" || echo "⚠ 缺少 \\listoffigures"
grep -q 'listoftables' 论文/论文正文.tex && echo "✅ 表格清单命令" || echo "⚠ 缺少 \\listoftables"
fi
```
If any format checks fail, Meta-model-agent should fix 论文正文.tex to match the template format before recompiling.
### 执行节点 6：Competition compliance
Verify items:
1. **Page count**: determine which parts count under the active competition Profile and require the count to be `≤ MAX_PAGES`. Separately reject substantively incomplete chapters; never require the paper to fill the page ceiling.
2. **Anonymous**: no team info (队号, 队员, 指导老师)
3. **Abstract exists** (数模竞赛: at least Chinese; 统计建模: both Chinese and English)
4. **TOC exists** (required by MathorCup etc.)
5. **Code appendix exists**
6. **No undefined references/citations**
“Code appendix exists”不得只检查标题或“代码/程序”关键词。必须读取 `程序/code_manifest.json`，逐项核验源码存在、SHA-256 未陈旧、`required_in_appendix=true` 的文件均有 `CODE_FILE` 标记和真实 `\lstinputlisting`/代码块。三种赛制均只核验主程序和逐问核心实现；辅助源码只能列入支撑材料清单，附录显示名必须为英文代码名。
CUMCM 页数必须读取编译后的 `AbstractStart/AbstractEnd` 与 `BodyStart/BodyEnd` 标签：摘要不超过 1 页，正文不超过 30 页，附录页数单独报告且不计入正文。模板含目录、标签缺失或仅用字符数估算时直接失败。
#### 正文充分性诊断（与页数上限分开）
不要再把“正文达到 MAX_PAGES 的 80%”作为硬目标。`MAX_PAGES` 只表示上限。即使页数较少，也只有在论证链缺失时才扩充正文。
```bash
echo "=== 页数不足诊断 ==="
echo "页数上限: MAX_PAGES；另行检查正文论证充分性"
echo ""
echo "=== 各章节字符数（找出最薄的章节）==="
for f in 论文/章节/*.tex; do
chars=$(wc -c < "$f")
echo " $(basename $f): $chars 字符 (~$(echo "scale=1; $chars/900" | bc) 页)"
done
echo ""
echo "=== 推荐扩充的章节（字符数最少的 3 个）==="
for f in $(ls -S 论文/章节/*.tex | tail -3); do
chars=$(wc -c < "$f")
echo " ⚠ $(basename $f): 仅 $chars 字符"
done
```
只有出现以下问题时才标记为 CRITICAL 并扩充：
- 某个子问题缺少模型机制或必要公式推导
- 结果只有数字或图表，没有验证和解释
- 摘要/结论中的关键主张无法在正文定位
扩充时只添加有证据来源的内容：
- Read 建模报告.md and 计算结果.md for detailed content
- Add unexpanded derivations, result analysis, validation, parameter discussions, and limitations
- Recompile after expansion
### 执行节点 7：⛔ FINAL QUALITY GATE (must ALL pass before finishing)
Run all checks and verify every item passes. **Do NOT output the report until all CRITICAL items are resolved.**
```bash
echo "=========================================="
echo " FINAL QUALITY GATE"
echo "=========================================="
GATE_FAIL=0
# 1. PDF exists and non-trivial
if [ -f 论文/数模论文.pdf ] && [ $(wc -c < 论文/数模论文.pdf) -gt 100000 ]; then
echo "✅ PDF exists ($(wc -c < 论文/数模论文.pdf) bytes)"
else
echo "❌ PDF missing or too small"; GATE_FAIL=$((GATE_FAIL+1))
fi
# 2. No LaTeX errors
MATH_ERR=$(grep -c 'Bad math environment delimiter\|Missing \$ inserted' 论文/编译日志.log 2>/dev/null || echo 0)
LR_ERR=$(grep -c 'Not allowed in LR mode' 论文/编译日志.log 2>/dev/null || echo 0)
[ "$((MATH_ERR+LR_ERR))" -eq 0 ] && echo "✅ No LaTeX errors" || { echo "❌ $MATH_ERR math + $LR_ERR LR errors"; GATE_FAIL=$((GATE_FAIL+1)); }
# 3. Bibliography not empty
BBL_ENTRIES=$(grep -c '\\bibitem' 论文/论文正文.bbl 2>/dev/null || echo 0)
[ "$BBL_ENTRIES" -gt 0 ] && echo "✅ Bibliography: $BBL_ENTRIES entries" || { echo "❌ Bibliography empty (BibTeX failed)"; GATE_FAIL=$((GATE_FAIL+1)); }
# 4. No unembedded publish=true visuals
UNEMBED=0
for pdf in 图表/*.pdf; do
[ -f "$pdf" ] || continue
bn=$(basename "$pdf")
python3 -c "import json,sys; d=json.load(open('图表/figure_manifest.json',encoding='utf-8')); sys.exit(0 if any(x.get('publish') and x.get('path','').replace('\\\\','/').endswith('/'+sys.argv[1]) for x in d.get('figures',[])) else 1)" "$bn" || continue
grep -rq "$bn" 论文/章节/*.tex 论文/论文正文.tex 2>/dev/null || UNEMBED=$((UNEMBED+1))
done
[ "$UNEMBED" -eq 0 ] && echo "✅ All publication visuals embedded" || { echo "❌ $UNEMBED publish=true visuals not embedded in manuscript"; GATE_FAIL=$((GATE_FAIL+1)); }
# 5. Page count
PAGE_EST=0
for f in 论文/章节/*.tex; do [ -f "$f" ] || continue; c=$(wc -c < "$f"); PAGE_EST=$((PAGE_EST + c)); done
PAGE_EST=$((PAGE_EST / 900))
echo " Body estimate: ~$PAGE_EST pages; MAX_PAGES is the submission ceiling, not a fill target"
# 6. Overfull vbox (table/visual overflow)
VBOX_ERR=$(grep -c 'Overfull.*vbox' 论文/编译日志.log 2>/dev/null || echo 0)
[ "$VBOX_ERR" -eq 0 ] && echo "✅ No table/visual overflow" || { echo "❌ $VBOX_ERR overfull vbox — tables may be cut off"; GATE_FAIL=$((GATE_FAIL+1)); }
# 7. AI writing patterns (itemize in body)
AI_LISTS=0
for f in 论文/章节/*.tex; do
[ -f "$f" ] || continue
echo "$(basename $f)" | grep -qi 'appendix\|附录' && continue
c=$(grep -c '\\begin{itemize}' "$f" 2>/dev/null || echo 0)
AI_LISTS=$((AI_LISTS + c))
done
[ "$AI_LISTS" -eq 0 ] && echo "✅ No bullet lists in body" || { echo "❌ $AI_LISTS itemize — convert to prose"; GATE_FAIL=$((GATE_FAIL+1)); }
# 8. Template integrity — compare preamble against established template
echo "--- Template integrity ---"
TMPL=""
for t in 模板/stats_论文正文.tex 模板/cumcm_论文正文.tex 模板/mcm_论文正文.tex 模板/bachelor_论文正文.tex 模板/master_论文正文.tex 模板/journal_论文正文.tex; do
[ -f "$t" ] && TMPL="$t" && break
done
if [ -n "$TMPL" ] && [ -f 论文/论文正文.tex ]; then
# Extract preamble (before \begin{document}) from both files
TMPL_PRE=$(sed -n '1,/\\begin{document}/p' "$TMPL" | grep '\\usepackage\|\\documentclass\|\\ctexset\|\\pagestyle\|\\renewcommand.*headrulewidth\|\\listoftables\|\\listoffigures\|\\cline\|\\bibliography' | sort)
MAIN_PRE=$(sed -n '1,/\\begin{document}/p' 论文/论文正文.tex | grep '\\usepackage\|\\documentclass\|\\ctexset\|\\pagestyle\|\\renewcommand.*headrulewidth\|\\listoftables\|\\listoffigures\|\\cline\|\\bibliography' | sort)
MISSING=$(comm -23 <(echo "$TMPL_PRE") <(echo "$MAIN_PRE") 2>/dev/null | head -5)
if [ -z "$MISSING" ]; then
echo "✅ Template preamble intact"
else
echo "❌ Template preamble was modified — these lines from template are missing in 论文正文.tex:"
echo "$MISSING" | sed 's/^/ /'
GATE_FAIL=$((GATE_FAIL+1))
fi
# Validate cover page structure (if template has one)
if grep -q '参赛学校\|参赛作品' "$TMPL" 2>/dev/null; then
grep -q '参赛学校\|参赛作品' 论文/论文正文.tex 2>/dev/null && echo "✅ Cover page present" || { echo "❌ Cover page missing/rewritten"; GATE_FAIL=$((GATE_FAIL+1)); }
grep -q 'cline{2-2}' 论文/论文正文.tex 2>/dev/null && echo "✅ Cover underlines (cline)" || { echo "❌ Cover cline missing"; GATE_FAIL=$((GATE_FAIL+1)); }
fi
# Validate bracket placeholders not remaining
if grep -q '\[论文标题\]\|\[学校名称\]\|\[队员1\]\|\[指导老师\]\|\[竞赛年份\]\|\[届数\]\|\[中文摘要内容\]' 论文/论文正文.tex 2>/dev/null; then
echo "❌ Unreplaced bracket placeholders in 论文正文.tex"; GATE_FAIL=$((GATE_FAIL+1))
else
echo "✅ All placeholders replaced"
fi
# Validate hand-written visual/table list (anti-pattern)
if grep -P '^(表|图)\d+\.' 论文/论文正文.tex 2>/dev/null | head -1 | grep -q '.'; then
echo "❌ Hand-written figure/table list (use \\listoftables/\\listoffigures)"; GATE_FAIL=$((GATE_FAIL+1))
fi
else
echo " (no template found for comparison, skipping)"
fi
# 9. Visual plan reconciliation (validate planning docs)
echo "--- Visual plan validate ---"
PLAN_FIGS=0; ACTUAL_FIGS=$(ls 图表/*.pdf 2>/dev/null | wc -l)
for plan in 选题规划.md 论文规划.md 问题分析.md; do
[ -f "$plan" ] || continue
pf=$(grep -ci 'fig_\|图.*：\|visual.*:' "$plan" 2>/dev/null || echo 0)
[ "$pf" -gt "$PLAN_FIGS" ] && PLAN_FIGS=$pf
done
if [ "$PLAN_FIGS" -gt 0 ]; then
echo " Planned: ~$PLAN_FIGS visuals, Actual: $ACTUAL_FIGS PDFs"
[ "$ACTUAL_FIGS" -ge "$PLAN_FIGS" ] && echo "✅ Visual count meets original plan" || echo " ⚠ Visual count is below the early plan; confirm omitted items are registered publish=false with reasons"
else
echo " No visual plan found, actual: $ACTUAL_FIGS PDFs"
fi
# 10. TikZ 几何/算法/架构图
# TikZ 由 systems-diagramming 编译成 图表/结构示意图.pdf, 通过 \includegraphics 嵌入
# (不是 sections 里的裸 tikzpicture 代码)。两种形态都算已嵌入。
TIKZ_IN_PAPER=$(grep -rl 'tikzpicture\|结构示意图\|tikz_' 论文/章节/*.tex 论文/论文正文.tex 2>/dev/null | wc -l)
# ⛔ 只在"真实产出了 tikz_*.pdf 产物"时才硬核对(避免规划提了但用 DrawIO 替代时误报失败)
TIKZ_PDF_TOTAL=0; TIKZ_PDF_MISSING=0
for tpdf in 图表/结构示意图.pdf 图表/结构示意图_*.pdf 图表/tikz_*.pdf; do
[ -f "$tpdf" ] || continue
python3 -c "import json,sys; d=json.load(open('图表/figure_manifest.json',encoding='utf-8')); sys.exit(0 if any(x.get('publish') and x.get('path','').replace('\\\\','/').endswith('/'+sys.argv[1]) for x in d.get('figures',[])) else 1)" "$(basename "$tpdf")" || continue
TIKZ_PDF_TOTAL=$((TIKZ_PDF_TOTAL+1))
grep -rq "$(basename "$tpdf")" 论文/章节/*.tex 论文/论文正文.tex 2>/dev/null || TIKZ_PDF_MISSING=$((TIKZ_PDF_MISSING+1))
done
if [ "$TIKZ_PDF_TOTAL" -gt 0 ]; then
# 有 publish=true 的 TikZ 产物 → 务必全部嵌入
if [ "$TIKZ_PDF_MISSING" -gt 0 ]; then
echo "❌ $TIKZ_PDF_MISSING 张 TikZ PDF 未嵌入任何章节"; GATE_FAIL=$((GATE_FAIL+1))
else
echo "✅ TikZ diagrams embedded ($TIKZ_PDF_TOTAL)"
fi
elif grep -qi 'tikz\|架构\|路线图\|几何示意\|算法流程' 论文规划.md 问题分析.md 2>/dev/null; then
# 规划提到 TikZ 但没产出 tikz_*.pdf(可能已用 DrawIO 替代) → 仅提醒, 不判失败
[ "$TIKZ_IN_PAPER" -gt 0 ] && echo "✅ TikZ embedded" || echo " ⚠ 规划提到 TikZ 但未发现 tikz_*.pdf(可能已用 DrawIO 替代, 不阻塞)"
else
[ "$TIKZ_IN_PAPER" -gt 0 ] && echo "✅ TikZ diagrams embedded ($TIKZ_IN_PAPER)" || echo " (no TikZ planned)"
fi
# 11. Citations in body
CITE_COUNT=$(grep -roh '\\cite{' 论文/章节/*.tex 论文/论文正文.tex 2>/dev/null | wc -l)
[ "$CITE_COUNT" -gt 0 ] && echo "✅ Citations: $CITE_COUNT" || { echo "❌ No citations in body text"; GATE_FAIL=$((GATE_FAIL+1)); }
# 12. No placeholders remaining
PLACEHOLDERS=$(grep -rl 'PLACEHOLDER\|待补充\|TODO\|\[论文标题\]\|\[中文摘要内容\]' 论文/章节/*.tex 论文/论文正文.tex 2>/dev/null | wc -l)
[ "$PLACEHOLDERS" -eq 0 ] && echo "✅ No placeholders" || { echo "❌ $PLACEHOLDERS files have placeholders"; GATE_FAIL=$((GATE_FAIL+1)); }
# 13. Undefined references
UNDEF_REFS=$(grep -c 'LaTeX Warning.*Reference.*undefined' 论文/编译日志.log 2>/dev/null || echo 0)
[ "$UNDEF_REFS" -eq 0 ] && echo "✅ No undefined refs" || { echo "❌ $UNDEF_REFS undefined references"; GATE_FAIL=$((GATE_FAIL+1)); }
# 14. Overfull hbox（零容忍）
HBOX_ERR=$(grep -c 'Overfull.*hbox' 论文/编译日志.log 2>/dev/null || echo 0)
[ "$HBOX_ERR" -eq 0 ] && echo "✅ No overfull hbox" || { echo "❌ $HBOX_ERR overfull hbox — fix wide tables/formulas"; GATE_FAIL=$((GATE_FAIL+1)); }
# 15. Visual stacking
STACKING=0
for f in 论文/章节/*.tex; do [ -f "$f" ] || continue; s=$(awk '/\\end\{(figure|table)\}/{a=1;t=0;next} a&&/\\begin\{(figure|table)\}/{if(t<3)c++;a=0;next} a&&/[a-zA-Z\x80-\xff]{3,}/{t++} a&&t>=3{a=0} END{print c+0}' "$f" 2>/dev/null); STACKING=$((STACKING+s)); done
[ "$STACKING" -eq 0 ] && echo "✅ No visual stacking" || { echo "❌ $STACKING visual stacking — add analysis text between visuals"; GATE_FAIL=$((GATE_FAIL+1)); }
# 16. TOC
if grep -q 'tableofcontents' 论文/论文正文.tex 2>/dev/null; then
[ -s 论文/论文正文.toc ] && echo "✅ TOC generated" || {
echo "❌ TOC empty — running extra compile pass..."
cd 论文/
xelatex -interaction=nonstopmode 论文正文.tex > /dev/null 2>&1
xelatex -interaction=nonstopmode 论文正文.tex > /dev/null 2>&1
cd ..
[ -s 论文/论文正文.toc ] && echo "✅ TOC generated following extra compile" || { echo "❌ TOC still empty"; GATE_FAIL=$((GATE_FAIL+1)); }
}
fi
# 17. Abstracts (Chinese papers)
if grep -q 'ctex' 论文/论文正文.tex 2>/dev/null; then
grep -rq '摘.*要' 论文/章节/*.tex 论文/论文正文.tex 2>/dev/null && echo "✅ Chinese abstract" || { echo "❌ No Chinese abstract"; GATE_FAIL=$((GATE_FAIL+1)); }
grep -rq 'Abstract' 论文/章节/*.tex 论文/论文正文.tex 2>/dev/null && echo "✅ English abstract" || { echo "❌ No English abstract"; GATE_FAIL=$((GATE_FAIL+1)); }
fi
# 18. Run compile_check.sh + writing_check.sh for full details
echo ""
echo "--- Full validate scripts ---"
bash 工具/compile_check.sh 论文/ 2>/dev/null || bash skills/shared-scripts/compile_check.sh 论文/ 2>/dev/null
bash 工具/writing_check.sh 论文/ 2>/dev/null || bash skills/shared-scripts/writing_check.sh 论文/ 2>/dev/null
WC_EXIT=$?
[ "$WC_EXIT" -eq 0 ] && echo "✅ Writing checks passed" || { echo "❌ Writing checks failed (exit=$WC_EXIT)"; GATE_FAIL=$((GATE_FAIL+1)); }
# 19. 符号阐明 longtable 核验
echo "--- Symbol table format ---"
for f in 论文/章节/*.tex; do
[ -f "$f" ] || continue
if grep -q '\\section{符号说明}\|\\section.*符号' "$f" 2>/dev/null; then
if grep -q '\\begin{longtable}' "$f" 2>/dev/null; then
echo "✅ 符号阐明采用 longtable"
elif grep -q '\\begin{table}' "$f" 2>/dev/null; then
echo "❌ 符号阐明仍用 table（应转 longtable 防分页）"; GATE_FAIL=$((GATE_FAIL+1))
fi
fi
done
# 20. 正文长表格核验（>12 行必须摘要化；需跨页时使用 longtable）
echo "--- Long table validate ---"
for f in 论文/章节/*.tex; do
[ -f "$f" ] || continue
echo "$(basename $f)" | grep -qi 'symbol\|appendix\|A_code' && continue
if grep -q '\\begin{tabular}' "$f" 2>/dev/null; then
ROW_COUNT=$(awk '/\\begin\{tabular\}/,/\\end\{tabular\}/' "$f" 2>/dev/null | grep -c '&' || echo 0)
[ "$ROW_COUNT" -gt 12 ] && { echo "❌ $(basename $f): $ROW_COUNT 行结果表应在正文摘要化，完整表移至附录并按需使用 longtable"; GATE_FAIL=$((GATE_FAIL+1)); }
fi
done
# 21. babel[english] 冲突
echo "--- babel validate ---"
if grep -q 'ctex\|cumcmthesis\|gmcmthesis' 论文/论文正文.tex 2>/dev/null; then
grep -q 'babel.*english' 论文/论文正文.tex 2>/dev/null && { echo "❌ 中文论文有 babel[english]"; GATE_FAIL=$((GATE_FAIL+1)); } || echo "✅ 无 babel 冲突"
fi
# 22. 数值一致性（JSON vs 论文）
echo "--- Numerical consistency ---"
if [ -f 图表/全部结果.json ]; then
python3 -c "
import json, re, os
with open('图表/全部结果.json','r',encoding='utf-8') as f: outcomes=json.load(f)
def extract(obj,p=''):
n={}
if isinstance(obj,dict):
for k,v in obj.items(): n.update(extract(v,f'{p}.{k}'))
elif isinstance(obj,(int,float)) and not isinstance(obj,bool):
if 0.0013 else 0)
" 2>/dev/null
[ $? -ne 0 ] && GATE_FAIL=$((GATE_FAIL+1))
fi
# 22.5 "太完美"结果检测（AI 编造或过拟合特征）
echo "--- Unrealistic values validate ---"
if [ -f 图表/全部结果.json ]; then
python3 -c "
import json
with open('图表/全部结果.json','r',encoding='utf-8') as f: data=json.load(f)
suspicious = []
def validate(name, val):
if not isinstance(val,(int,float)) or isinstance(val,bool): return
key = name.lower()
if any(w in key for w in ['r2','r_squared','accuracy','acc','precision','recall','f1','auc']) and val > 0.999:
suspicious.append(f'{name}={val:.4f} 过于完美（>0.999）')
if any(w in key for w in ['rmse','mae','mse','loss']) and val == 0:
suspicious.append(f'{name}=0 完美误差')
if ('p_value' in key or 'pvalue' in key) and val == 0:
suspicious.append(f'{name}=0 完美显著')
if any(w in key for w in ['improvement','speedup','gain','提升']) and val > 10:
suspicious.append(f'{name}={val} 提升过大（{val*100:.0f}%）')
def walk(obj, path=''):
if isinstance(obj,dict):
for k,v in obj.items(): walk(v, f'{path}.{k}')
elif isinstance(obj,list):
for i,v in enumerate(obj): walk(v, f'{path}[{i}]')
else: validate(path, obj)
walk(data)
if suspicious:
print(f'🚩 {len(suspicious)} 处可疑的完美结果（可能过拟合或数值编造）:')
for s in suspicious[:5]: print(f' {s}')
print(' 需在论文中阐明合理性，或回到 computational-realization 核验数据泄漏')
else:
print('✅ 数值合理性通过')
" 2>/dev/null
fi
# 22.6 合理性复核章节是否出现（务必）
echo "--- 合理性复核章节核验 ---"
if [ -f 计算结果.md ]; then
if grep -q '合理性审查\|数值合理\|背景对照\|sanity.*verify' 计算结果.md 2>/dev/null; then
echo "✅ 计算结果.md 涵盖合理性复核章节"
else
echo "❌ 计算结果.md 缺少合理性复核章节 — 回到 computational-realization Phase 1.7 补充"
GATE_FAIL=$((GATE_FAIL+1))
fi
fi
# 22.7 数据源时间戳一致性（代码 vs 图形与表格 vs 论文）
echo "--- 数据源时间戳一致性 ---"
if [ -f 图表/全部结果.json ]; then
JSON_TIME=$(stat -c %Y 图表/全部结果.json 2>/dev/null || stat -f %m 图表/全部结果.json 2>/dev/null || echo 0)
# 核验是否有 PDF 图形与表格比 JSON 旧（阐明代码重跑了但图没更新）
STALE_FIGS=0
for pdf in 图表/*.pdf; do
[ -f "$pdf" ] || continue
PDF_TIME=$(stat -c %Y "$pdf" 2>/dev/null || stat -f %m "$pdf" 2>/dev/null || echo 0)
if [ "$JSON_TIME" -gt "$PDF_TIME" ] && [ "$((JSON_TIME - PDF_TIME))" -gt 60 ]; then
echo " ⚠ $(basename $pdf) 比 全部结果.json 旧 — 图形与表格数据可能过期"
STALE_FIGS=$((STALE_FIGS+1))
fi
done
if [ "$STALE_FIGS" -gt 0 ]; then
echo " ❌ $STALE_FIGS 张图形与表格可能采用了旧数据（JSON 更新后图形与表格未重新产出）"
echo " → 推荐重跑 evidence-visualization 环节更新图形与表格"
GATE_FAIL=$((GATE_FAIL+1))
else
echo "✅ 数据源时间戳保持一致"
fi
fi
# 22.8 正文长表格检测（>12 行的表格不应在正文中完整展开）
echo "--- 正文长表格检测 ---"
LONG_TABLES=0
for f in 论文/章节/*.tex; do
[ -f "$f" ] || continue
# 跳过附录文件
echo "$(basename $f)" | grep -qi 'appendix\|附录\|A_code' && continue
# 统计各个 tabular/longtable 环境内的数据行数（\\ 的数量）
python3 -c "
import re
with open('$f', 'r', encoding='utf-8', errors='ignore') as fh:
content = fh.read()
# 找全部 tabular/longtable 环境
for env in ['tabular', 'longtable', 'tabular*']:
pattern = r'\\\\begin\{' + env + r'[*]?\}.*?\\\\end\{' + env + r'[*]?\}'
for match in re.finditer(pattern, content, re.DOTALL):
table_text = match.group()
row_count = table_text.count('\\\\\\\\') - table_text.count('\\\\hline') // 2
if row_count > 12:
print(f'$(basename $f): {env} 有 {row_count} 行（>12）')
" 2>/dev/null | while read line; do
echo " ❌ $line — 正文表格超过 12 行，应截断（前3+后3+省略），完整版放附录"
LONG_TABLES=$((LONG_TABLES+1))
done
done
[ "$LONG_TABLES" -eq 0 ] && echo "✅ 正文无超长表格" || GATE_FAIL=$((GATE_FAIL+1))
# 23. AI 写作痕迹（列表环境）
echo "--- AI writing patterns ---"
AI_LISTS=0
for f in 论文/章节/*.tex; do
[ -f "$f" ] || continue
echo "$(basename $f)" | grep -qi 'appendix\|A_code' && continue
c=$(grep -c '\\begin{itemize}\|\\begin{enumerate}' "$f" 2>/dev/null || echo 0)
AI_LISTS=$((AI_LISTS+c))
done
[ "$AI_LISTS" -le 3 ] && echo "✅ AI patterns: $AI_LISTS lists" || { echo "❌ $AI_LISTS lists in body — convert to prose"; GATE_FAIL=$((GATE_FAIL+1)); }
# 24. 元叙述泄露
echo "--- Meta content leak ---"
META=0
for f in 论文/章节/*.tex; do
[ -f "$f" ] || continue
l=$(grep -ci 'RESULTS\.md\|META_MODEL_AGENT\.md\|MODELING_REPORT\|PROBLEM_ANALYSIS\|latex_includes' "$f" 2>/dev/null || echo 0)
META=$((META+l))
done
[ "$META" -eq 0 ] && echo "✅ No meta leaks" || { echo "❌ $META meta content leaks"; GATE_FAIL=$((GATE_FAIL+1)); }
# 25. 过度声称
echo "--- Overclaiming ---"
OC=0
for f in 论文/章节/*.tex; do
[ -f "$f" ] || continue
for w in "首次提出" "首次发现" "完美" "最优的" "无可比拟" "前所未有" "开创性" "革命性"; do
c=$(grep -c "$w" "$f" 2>/dev/null || echo 0); OC=$((OC+c))
done
done
[ "$OC" -eq 0 ] && echo "✅ No overclaiming" || echo "⚠ $OC overclaiming instances"
echo ""
echo "=========================================="
if [ "$GATE_FAIL" -eq 0 ]; then
echo " ✅ ALL CRITICAL CHECKS PASSED — ready to submit"
else
echo " ❌ $GATE_FAIL CRITICAL FAILURES — MUST FIX prior to finishing"
echo " Go back and fix each ❌ item, recompile, then re-run this gate."
fi
echo "=========================================="
```
**⛔ If GATE_FAIL > 0, you MUST go back and fix every ❌ item, recompile, and re-run this gate. Do NOT output the final report with any ❌ remaining. Repeat until GATE_FAIL = 0.**
### 执行节点 8：Output report
Competition name, status, PDF path, total pages, body pages, compliance pass/fail.
## 核心规则
- No latexmk — manual step-by-step compilation
- Avoid delete .bbl file after compilation (bibliography data)
- Figure paths auto-corrected by compile_utils.sh: `图表/` → `../图表/`
- Total/countable pages ≤ MAX_PAGES. Page count is a ceiling;正文充分性按逐问论证链独立检查。
- Every embedded LaTeX figure must stay within `\linewidth` and `0.70\textheight` with `keepaspectratio`; every DOCX image must stay within the computed page body width. Oversized or unreadably shrunken figures are gate failures.
- Anonymous: no team info in body
- Primary output: `论文/数模论文.pdf`, temp files: `临时文件/`