feat: PDF对比用iframe显示原PDF文件（支持扫描版PDF）

app.py

@@ -673,12 +673,33 @@ def compare_view(translation_id):
     return jsonify({
         'id': translation.id,
         '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）',
+        'original': original_content or '',
+        'original_pdf_url': f'/api/original-pdf/{translation.id}' if translation.upload_path else None,
         'translated': translated_content,
         'pages': translated_pages
     })
 
 
+@app.route('/api/original-pdf/<int:translation_id>')
+def get_original_pdf(translation_id):
+    """获取原始PDF文件"""
+    user = get_current_user()
+    if not user:
+        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 and user.user_type != 'admin'):
+        return jsonify({'error': '无权访问'}), 403
+    
+    if not translation.upload_path or not os.path.exists(translation.upload_path):
+        return jsonify({'error': '原PDF文件不存在'}), 404
+    
+    return send_file(translation.upload_path, 
+                     mimetype='application/pdf',
+                     as_attachment=False,
+                     download_name=translation.original_filename)
+
+
 # ==================== 路由: 用户系统 ====================
 @app.route('/login', methods=['GET', 'POST'])
 def login():

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()

templates/translation.html

@@ -84,11 +84,21 @@
                     const response = await fetch(`/api/compare/${translationId}`);
                     const result = await response.json();
                     
+                    // 原文面板：如果有PDF URL用iframe显示，否则显示提取的文本
+                    let originalHtml = '';
+                    if (result.original_pdf_url) {
+                        originalHtml = `<iframe src="${result.original_pdf_url}" style="width:100%;height:500px;border:none;"></iframe>`;
+                    } else if (result.original && result.original.length > 0) {
+                        originalHtml = `<div style="white-space:pre-wrap;font-family:monospace;">${escapeHtml(result.original)}</div>`;
+                    } else {
+                        originalHtml = '<div class="text-muted">原文内容未找到（可能PDF已被删除）</div>';
+                    }
+                    
                     document.getElementById('resultContent').innerHTML = `
                         <div class="compare-container">
                             <div class="compare-panel original">
-                                <h5>原文</h5>
+                                <h5>原文 PDF</h5>
-                                <div>${escapeHtml(result.original)}</div>
+                                ${originalHtml}
                             </div>
                             <div class="compare-panel translated">
                                 <h5>译文</h5>
@@ -98,7 +108,7 @@
                     `;
                     
                 } catch (error) {
-                    alert('加载对比失败');
+                    alert('加载对比失败: ' + error.message);
                 }
             } else {
                 loadResult();