Initial commit

final_check_v2.py

@@ -0,0 +1,66 @@
+import os, re
+
+dir_path = '/app/working/workspaces/default/novels/r2-ai-xiu-jiao'
+files = sorted(
+    [f for f in os.listdir(dir_path) if f.endswith('.md') and f != 'README.md'],
+    key=lambda x: int(re.search(r'第(\d+)章', x).group(1)) if re.search(r'第(\d+)章', x) else 99
+)
+
+all_text = ''.join(open(os.path.join(dir_path, f)).read() for f in files)
+sizes = [len(open(os.path.join(dir_path, f)).read().encode('utf-8')) for f in files]
+hz = len(re.findall(r'[\u4e00-\u9fff]', all_text))
+
+print('=' * 60)
+print('《AI能代替一切，除了修脚》- 完整度检查')
+print('=' * 60)
+
+print(f'15章完整: {"OK" if len(files) == 15 else "ISSUE: " + str(len(files))}')
+print(f'汉字:~{hz} | 均:{sum(sizes)//15}b | 最短:{min(sizes)}b | 最长:{max(sizes)}b')
+
+print()
+print('--- 审核敏感词扫描 ---')
+for w in ['公司', '集团', '企业']:
+    count = len(re.findall(w, all_text))
+    print(f'  "{w}": {count}处 {"✅" if count == 0 else "⚠️"}')
+for w in ['被裁', '裁员', '辞退', '被优化', '被替代', '被代替']:
+    count = len(re.findall(w, all_text))
+    print(f'  "{w}": {count}处 {"✅" if count == 0 else "⚠️"}')
+
+print()
+print('--- 章节开头模式检查 ---')
+for f in files[:3]:
+    lines = open(os.path.join(dir_path, f)).read().strip().split('\n')
+    for l in lines:
+        s = l.strip()
+        if s and not s.startswith('# ') and len(s) > 5:
+            print(f'  {f[:15]}: {s[:60]}')
+            break
+
+print()
+print('--- AI味残留扫描 ---')
+patterns = {
+    '燃点标签': r'燃点|【燃|（燃',
+    '章末标注': r'第[一二三四五六七八九十百]+章.*完',
+    '深吸一口气': r'深吸一口气|深呼一口气',
+    '握紧/攥紧': r'握紧|攥紧|捏紧',
+    '——独立行': r'^\s*——+\s*$',
+    'meta自嘲': r'不对.*系统|这不是.*小说|作者注',
+    '全书完': r'（全书完）|（全文完）',
+}
+for name, pat in patterns.items():
+    count = len(re.findall(pat, all_text, re.MULTILINE))
+    flag = '⚠️' if count > 0 else '✅'
+    print(f'  {name}: {flag} ({count}处)')
+
+print()
+print('--- 被动受害者叙事扫描 ---')
+bei_pat = r'被[^。！？]{0,20}替代|被[^。！？]{0,20}代替|被[^。！？]{0,20}辞退|被[^。！？]{0,20}裁'
+bei_count = len(re.findall(bei_pat, all_text))
+print(f'  被动被XX叙事: {"✅ 0处" if bei_count == 0 else "⚠️ " + str(bei_count) + "处"}')
+
+score = 100
+if min(sizes) < 1200: score -= 10
+if bei_count > 0: score -= bei_count * 10
+print()
+print(f'评分: {max(score, 0)}/100')
+print(f'评级: {"S - 可直接投番茄" if score >= 90 else "A - 微调后可投"}')