feat: 添加修复upload_path脚本

fix: PDF对比功能无法显示原文 - uploads和outputs使用同一UUID，添加upload_path字段
fix: 实现对比查看功能
2026-04-17 09:43:18 +08:00 · 2026-04-17 09:32:21 +08:00 · 2026-04-16 23:45:57 +08:00 · 2026-04-16 23:33:40 +08:00
6 changed files with 373 additions and 20 deletions
--- a/app.py
+++ b/app.py
@@ -374,16 +374,17 @@ def upload_pdf():
        output_path = cache_path
    else:
        # 需要翻译
-        # 保存上传文件
-        upload_dir = os.path.join(UPLOAD_DIR, str(uuid.uuid4()))
+        # 保存上传文件 - 使用同一个UUID确保uploads和outputs目录关联
+        session_uuid = str(uuid.uuid4())
+        upload_dir = os.path.join(UPLOAD_DIR, session_uuid)
        os.makedirs(upload_dir, exist_ok=True)
        upload_path = os.path.join(upload_dir, filename)
        
        with open(upload_path, 'wb') as f:
            f.write(file_content)
        
-        # 创建输出路径
-        output_dir = os.path.join(OUTPUT_DIR, str(uuid.uuid4()))
+        # 创建输出路径 - 使用相同的UUID
+        output_dir = os.path.join(OUTPUT_DIR, session_uuid)
        os.makedirs(output_dir, exist_ok=True)
        output_path = os.path.join(output_dir, f"{filename}_translated.md")
        
@@ -400,6 +401,7 @@ def upload_pdf():
        translate_params=json.dumps({'instruction': instruction}) if instruction else None,
        status='processing' if not from_cache else 'completed',
        progress=0 if not from_cache else 100,
+        upload_path=upload_path if not from_cache else None,  # 保存上传路径
        output_path=output_path,
        from_cache=from_cache
    )
@@ -609,16 +611,71 @@ def compare_view(translation_id):
        return jsonify({'error': '请登录后使用此功能'}), 401
    
    translation = Translation.query.get(translation_id)
-    if not translation or translation.user_id != user.id:
+    if not translation or (translation.user_id != user.id and user.user_type != 'admin'):
        return jsonify({'error': '无权访问'}), 403
    
-    # 生成对比文件
-    # TODO: 实现对比功能
+    # 读取翻译结果文件
+    translated_content = ''
+    if translation.output_path and os.path.exists(translation.output_path):
+        try:
+            with open(translation.output_path, 'r', encoding='utf-8') as f:
+                translated_content = f.read()
+        except Exception as e:
+            translated_content = f'读取失败: {str(e)}'
+    
+    # 从翻译结果中提取各页内容
+    # 翻译结果格式是Markdown，包含"## 第 X 页"分隔
+    original_pages = []
+    translated_pages = []
+    
+    if translated_content:
+        # 解析翻译结果的页面结构
+        import re
+        page_pattern = r'## 第 (\d+) 页\n\n(.*?)\n\n---'
+        matches = re.findall(page_pattern, translated_content, re.DOTALL)
+        
+        for page_num, content in matches:
+            translated_pages.append({
+                'page': int(page_num),
+                'content': content.strip()
+            })
+    
+    # 如果有原文内容存储，获取原文
+    original_content = ''
+    
+    # 优先从数据库存储的upload_path获取原PDF
+    possible_paths = []
+    
+    if translation.upload_path:
+        possible_paths.append(translation.upload_path)
+    
+    # 备用方案：尝试从路径推断（兼容旧数据）
+    upload_dir = os.path.dirname(translation.output_path.replace('outputs', 'uploads').replace('_translated.md', '')) if translation.output_path else ''
+    if upload_dir:
+        possible_paths.append(
+            translation.output_path.replace('outputs', 'uploads').replace('_translated.md', '') if translation.output_path else ''
+        )
+        possible_paths.append(os.path.join(upload_dir, translation.original_filename))
+    
+    for pdf_path in possible_paths:
+        if pdf_path and os.path.exists(pdf_path) and pdf_path.endswith('.pdf'):
+            try:
+                from pypdf import PdfReader
+                reader = PdfReader(pdf_path)
+                for page in reader.pages:
+                    text = page.extract_text()
+                    if text:
+                        original_content += text + '\n\n'
+            except:
+                pass
+            break
    
    return jsonify({
        'id': translation.id,
-        'original': '原文内容',
-        'translated': '译文内容'
+        'filename': translation.original_filename,
+        'original': original_content or '原文内容未找到（可能PDF已被删除或为扫描版）',
+        'translated': translated_content,
+        'pages': translated_pages
    })


--- a/logs/app.log
+++ b/logs/app.log
@@ -0,0 +1,187 @@
+ * Serving Flask app 'app'
+ * Debug mode: on
+WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead.
+ * Running on all addresses (0.0.0.0)
+ * Running on http://127.0.0.1:19000
+ * Running on http://192.168.2.17:19000
+Press CTRL+C to quit
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/services.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/app.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/app.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/pdf_translate.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/pdf_translate.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/pdf_translate.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/pdf_translate.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/pdf_translate.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/pdf_translate.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/pdf_translate.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/pdf_translate.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/pdf_translate.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/pdf_translate.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/pdf_translate.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+127.0.0.1 - - [16/Apr/2026 21:50:38] "GET /api/config HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:08] "GET /admin/translations HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:08] "GET /static/img/favicon.svg HTTP/1.1" 304 -
+/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/app.py:53: LegacyAPIWarning: The Query.get() method is considered legacy as of the 1.x series of SQLAlchemy and becomes a legacy construct in 2.0. The method is now available as Session.get() (deprecated since: 2.0) (Background on SQLAlchemy 2.0 at: https://sqlalche.me/e/b8d9)
+  return User.query.get(user_id)
+192.168.2.10 - - [16/Apr/2026 21:58:13] "GET / HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:13] "GET /static/css/style.css HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 21:58:13] "GET /static/js/main.js HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 21:58:13] "GET /api/user/info HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:13] "GET /static/img/favicon.svg HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 21:58:16] "GET /history HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:16] "GET /static/css/style.css HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 21:58:18] "GET / HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:18] "GET /static/css/style.css HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 21:58:18] "GET /static/js/main.js HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 21:58:18] "GET /api/user/info HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:26] "POST /api/upload HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:26] "GET /api/task/17baff33-894b-4e56-8975-cc2e6359cd66 HTTP/1.1" 200 -
+/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/app.py:481: LegacyAPIWarning: The Query.get() method is considered legacy as of the 1.x series of SQLAlchemy and becomes a legacy construct in 2.0. The method is now available as Session.get() (deprecated since: 2.0) (Background on SQLAlchemy 2.0 at: https://sqlalche.me/e/b8d9)
+  translation = Translation.query.get(translation_id)
+192.168.2.10 - - [16/Apr/2026 21:58:26] "GET /api/status/4 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:28] "GET /api/task/17baff33-894b-4e56-8975-cc2e6359cd66 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:28] "GET /api/status/4 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:30] "GET /api/task/17baff33-894b-4e56-8975-cc2e6359cd66 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:30] "GET /api/status/4 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:32] "GET /api/task/17baff33-894b-4e56-8975-cc2e6359cd66 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:32] "GET /api/status/4 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:34] "GET /api/task/17baff33-894b-4e56-8975-cc2e6359cd66 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:34] "GET /api/status/4 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:37] "GET /api/task/17baff33-894b-4e56-8975-cc2e6359cd66 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 21:58:37] "GET /api/status/4 HTTP/1.1" 200 -
+/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/app.py:519: LegacyAPIWarning: The Query.get() method is considered legacy as of the 1.x series of SQLAlchemy and becomes a legacy construct in 2.0. The method is now available as Session.get() (deprecated since: 2.0) (Background on SQLAlchemy 2.0 at: https://sqlalche.me/e/b8d9)
+  translation = Translation.query.get(translation_id)
+192.168.2.10 - - [16/Apr/2026 21:58:37] "GET /api/result/4 HTTP/1.1" 200 -
+127.0.0.1 - - [16/Apr/2026 22:00:01] "GET /api/health HTTP/1.1" 404 -
+127.0.0.1 - - [16/Apr/2026 22:00:33] "POST /api/upload HTTP/1.1" 200 -
+127.0.0.1 - - [16/Apr/2026 22:00:38] "GET /api/status/5 HTTP/1.1" 200 -
+127.0.0.1 - - [16/Apr/2026 22:01:15] "GET /api/status/5 HTTP/1.1" 200 -
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/services.py', reloading
+[翻译任务] 开始翻译，使用配置: https://open.bigmodel.cn/api/paas/v4
+[翻译任务] 开始翻译，使用配置: https://open.bigmodel.cn/api/paas/v4
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+127.0.0.1 - - [16/Apr/2026 22:20:01] "GET /api/health HTTP/1.1" 404 -
+/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/app.py:53: LegacyAPIWarning: The Query.get() method is considered legacy as of the 1.x series of SQLAlchemy and becomes a legacy construct in 2.0. The method is now available as Session.get() (deprecated since: 2.0) (Background on SQLAlchemy 2.0 at: https://sqlalche.me/e/b8d9)
+  return User.query.get(user_id)
+192.168.2.10 - - [16/Apr/2026 22:24:27] "GET / HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:24:27] "GET /static/css/style.css HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 22:24:27] "GET /static/js/main.js HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 22:24:27] "GET /api/user/info HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:24:27] "GET /static/img/favicon.svg HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 22:24:31] "GET /admin/translations HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:24:31] "GET /static/img/favicon.svg HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 22:24:33] "POST /admin/translation/5/toggle-share HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:24:33] "GET /admin/translations HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:24:33] "GET /static/img/favicon.svg HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 22:24:34] "POST /admin/translation/4/toggle-share HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:24:34] "GET /admin/translations HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:24:34] "GET /static/img/favicon.svg HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 22:24:42] "GET /admin/translation/5 HTTP/1.1" 200 -
+/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/app.py:519: LegacyAPIWarning: The Query.get() method is considered legacy as of the 1.x series of SQLAlchemy and becomes a legacy construct in 2.0. The method is now available as Session.get() (deprecated since: 2.0) (Background on SQLAlchemy 2.0 at: https://sqlalche.me/e/b8d9)
+  translation = Translation.query.get(translation_id)
+192.168.2.10 - - [16/Apr/2026 22:24:42] "GET /api/result/5 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:24:45] "GET /admin/translations HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:24:46] "GET /admin/translation/4 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:24:47] "GET /api/result/4 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:24:48] "GET /admin/translations HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:24:52] "GET / HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:24:52] "GET /static/css/style.css HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 22:24:52] "GET /static/js/main.js HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 22:24:52] "GET /static/img/favicon.svg HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 22:24:52] "GET /api/user/info HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:24:54] "GET / HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:24:54] "GET /static/css/style.css HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 22:24:54] "GET /static/js/main.js HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 22:24:54] "GET /api/user/info HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:24:54] "GET /static/img/favicon.svg HTTP/1.1" 304 -
+192.168.2.10 - - [16/Apr/2026 22:25:00] "POST /api/upload HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:25:00] "GET /api/task/ec82bf8c-70b3-474a-b94f-64598c81f7fa HTTP/1.1" 200 -
+/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/app.py:481: LegacyAPIWarning: The Query.get() method is considered legacy as of the 1.x series of SQLAlchemy and becomes a legacy construct in 2.0. The method is now available as Session.get() (deprecated since: 2.0) (Background on SQLAlchemy 2.0 at: https://sqlalche.me/e/b8d9)
+  translation = Translation.query.get(translation_id)
+192.168.2.10 - - [16/Apr/2026 22:25:00] "GET /api/status/6 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:25:02] "GET /api/task/ec82bf8c-70b3-474a-b94f-64598c81f7fa HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:25:02] "GET /api/status/6 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:25:05] "GET /api/task/ec82bf8c-70b3-474a-b94f-64598c81f7fa HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:25:05] "GET /api/status/6 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:25:07] "GET /api/task/ec82bf8c-70b3-474a-b94f-64598c81f7fa HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:25:07] "GET /api/status/6 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:25:09] "GET /api/task/ec82bf8c-70b3-474a-b94f-64598c81f7fa HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:25:09] "GET /api/status/6 HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:25:11] "GET /api/task/ec82bf8c-70b3-474a-b94f-64598c81f7fa HTTP/1.1" 200 -
+192.168.2.10 - - [16/Apr/2026 22:25:11] "GET /api/status/6 HTTP/1.1" 200 -
+127.0.0.1 - - [16/Apr/2026 22:40:01] "GET /api/health HTTP/1.1" 404 -
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/services.py', reloading
+[翻译任务] 开始翻译，使用配置: https://open.bigmodel.cn/api/paas/v4
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/services.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/services.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/services.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/services.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+ * Detected change in '/home/xian/.openclaw/workspace-coder/works/pdf-translate-web-v2/services.py', reloading
+ * Restarting with stat
+ * Debugger is active!
+ * Debugger PIN: 154-698-244
+127.0.0.1 - - [16/Apr/2026 23:00:01] "GET /api/health HTTP/1.1" 404 -
+127.0.0.1 - - [16/Apr/2026 23:03:32] "GET / HTTP/1.1" 200 -
--- a/models.py
+++ b/models.py
@@ -169,6 +169,7 @@ class Translation(db.Model):
    error_message = db.Column(db.Text, nullable=True)
    
    # 输出
+    upload_path = db.Column(db.String(255), nullable=True)  # 原始PDF文件路径
    output_path = db.Column(db.String(255), nullable=True)  # 翻译结果文件路径
    
    # 时间戳
--- a/scripts/fix_upload_path.py
+++ b/scripts/fix_upload_path.py
@@ -0,0 +1,67 @@
+#!/usr/bin/env python3
+"""
+修复旧数据的 upload_path 字段
+通过 file_hash 匹配 uploads 目录中的 PDF 文件
+"""
+
+import sqlite3
+import os
+import hashlib
+
+DB_PATH = 'instance/pdf_translate.db'
+UPLOADS_DIR = 'uploads'
+
+def compute_file_hash(filepath):
+    """计算文件MD5"""
+    hasher = hashlib.md5()
+    with open(filepath, 'rb') as f:
+        hasher.update(f.read())
+    return hasher.hexdigest()
+
+def main():
+    conn = sqlite3.connect(DB_PATH)
+    cursor = conn.cursor()
+    
+    # 获取所有 upload_path 为空的翻译记录
+    cursor.execute('SELECT id, file_hash, original_filename, output_path FROM translations WHERE upload_path IS NULL')
+    records = cursor.fetchall()
+    
+    print(f"找到 {len(records)} 条需要修复的记录")
+    
+    if not records:
+        print("无需修复")
+        conn.close()
+        return
+    
+    # 找所有上传的PDF文件
+    pdf_files = []
+    for root, dirs, files in os.walk(UPLOADS_DIR):
+        for f in files:
+            if f.endswith('.pdf'):
+                pdf_files.append(os.path.join(root, f))
+    
+    print(f"找到 {len(pdf_files)} 个PDF文件")
+    
+    # 按hash匹配
+    fixed_count = 0
+    for record in records:
+        id, file_hash, filename, output_path = record
+        
+        # 找匹配hash的文件
+        for pdf_path in pdf_files:
+            try:
+                pdf_hash = compute_file_hash(pdf_path)
+                if pdf_hash == file_hash:
+                    print(f"ID {id}: 找到匹配 {pdf_path}")
+                    cursor.execute('UPDATE translations SET upload_path = ? WHERE id = ?', (pdf_path, id))
+                    conn.commit()
+                    fixed_count += 1
+                    break
+            except Exception as e:
+                print(f"处理 {pdf_path} 失败: {e}")
+    
+    conn.close()
+    print(f"修复完成，共修复 {fixed_count} 条记录")
+
+if __name__ == '__main__':
+    main()
--- a/static/js/main.js
+++ b/static/js/main.js
@@ -70,6 +70,9 @@ document.getElementById('uploadForm').addEventListener('submit', async function(
        currentTranslationId = result.translation_id;
        currentTaskId = result.task_id;
        
+        // 更新按钮状态为翻译中
+        btnText.textContent = '翻译中...';
+        
        // 如果使用缓存，直接显示结果
        if (result.from_cache) {
            document.getElementById('cacheNotice').style.display = 'block';
@@ -201,18 +204,55 @@ document.getElementById('viewCompare')?.addEventListener('click', async function
 function showCompareView(data) {
    const resultContent = document.getElementById('resultContent');
    
-    resultContent.innerHTML = `
-        <div class="compare-container">
-            <div class="compare-panel original">
-                <h5>原文</h5>
-                <div class="content">${escapeHtml(data.original)}</div>
+    // 如果有分页数据，按页显示
+    if (data.pages && data.pages.length > 0) {
+        let html = '<div class="compare-container">';
+        
+        for (const page of data.pages) {
+            html += `
+                <div class="compare-section mb-4">
+                    <h5 class="text-center mb-3">第 ${page.page} 页</h5>
+                    <div class="row">
+                        <div class="col-md-6">
+                            <div class="card">
+                                <div class="card-header">原文（OCR识别）</div>
+                                <div class="card-body"><pre style="white-space: pre-wrap;">${escapeHtml(data.original || '原文内容')}</pre></div>
+                            </div>
+                        </div>
+                        <div class="col-md-6">
+                            <div class="card">
+                                <div class="card-header">译文</div>
+                                <div class="card-body">${renderMarkdown(page.content)}</div>
+                            </div>
+                        </div>
+                    </div>
+                </div>
+            `;
+        }
+        
+        html += '</div>';
+        resultContent.innerHTML = html;
+    } else {
+        // 单一对比视图
+        resultContent.innerHTML = `
+            <div class="compare-container">
+                <div class="row">
+                    <div class="col-md-6">
+                        <div class="card">
+                            <div class="card-header">原文</div>
+                            <div class="card-body"><pre style="white-space: pre-wrap;">${escapeHtml(data.original || '无原文内容')}</pre></div>
+                        </div>
+                    </div>
+                    <div class="col-md-6">
+                        <div class="card">
+                            <div class="card-header">译文</div>
+                            <div class="card-body">${renderMarkdown(data.translated || '无译文内容')}</div>
+                        </div>
+                    </div>
+                </div>
            </div>
-            <div class="compare-panel translated">
-                <h5>译文</h5>
-                <div class="content">${renderMarkdown(data.translated)}</div>
-            </div>
-        </div>
-    `;
+        `;
+    }
 }

 // 重新翻译
--- a/test_simple.txt
+++ b/test_simple.txt
@@ -0,0 +1 @@
+This is a test PDF file. Hello World. This is a simple test document.
Author	SHA1	Message	Date
coder	2e428b2500	feat: 添加修复upload_path脚本	2026-04-17 09:43:18 +08:00
coder	9569898f33	fix: PDF对比功能无法显示原文 - uploads和outputs使用同一UUID，添加upload_path字段	2026-04-17 09:32:21 +08:00
coder	3479cbd04c	fix: 实现对比查看功能 - 读取翻译结果文件内容 - 尝试从原PDF提取原文（OCR识别的文字） - 按页面分块显示对比 - 前端支持分页对比展示	2026-04-16 23:45:57 +08:00
coder	e5c9ea322a	fix: 上传成功后按钮状态改为翻译中上传PDF成功后立即更新按钮文字为翻译中...而不是保持上传中...	2026-04-16 23:33:40 +08:00
				`@@ -0,0 +1 @@`
				`This is a test PDF file. Hello World. This is a simple test document.`