优化AI服务的内容提取逻辑，增强对缺失字段的推断能力，改进后处理机制以提升数据提取的准确性和完整性。

enable_all_fields.py

@@ -0,0 +1,231 @@
+"""
+启用f_polic_field表中所有字段（将state更新为1）
+"""
+import pymysql
+import os
+from datetime import datetime
+
+# 数据库连接配置
+DB_CONFIG = {
+    'host': os.getenv('DB_HOST', '152.136.177.240'),
+    'port': int(os.getenv('DB_PORT', 5012)),
+    'user': os.getenv('DB_USER', 'finyx'),
+    'password': os.getenv('DB_PASSWORD', '6QsGK6MpePZDE57Z'),
+    'database': os.getenv('DB_NAME', 'finyx'),
+    'charset': 'utf8mb4'
+}
+
+TENANT_ID = 615873064429507639
+UPDATED_BY = 655162080928945152
+CURRENT_TIME = datetime.now()
+
+
+def check_field_states(conn):
+    """检查字段状态统计"""
+    cursor = conn.cursor(pymysql.cursors.DictCursor)
+    
+    # 统计各状态的字段数量（使用CAST来正确处理二进制类型）
+    sql = """
+        SELECT 
+            CAST(state AS UNSIGNED) as state_int,
+            field_type,
+            COUNT(*) as count
+        FROM f_polic_field
+        WHERE tenant_id = %s
+        GROUP BY CAST(state AS UNSIGNED), field_type
+        ORDER BY field_type, CAST(state AS UNSIGNED)
+    """
+    cursor.execute(sql, (TENANT_ID,))
+    stats = cursor.fetchall()
+    
+    cursor.close()
+    return stats
+
+
+def get_fields_by_state(conn, state):
+    """获取指定状态的字段列表"""
+    cursor = conn.cursor(pymysql.cursors.DictCursor)
+    
+    sql = """
+        SELECT id, name, filed_code, field_type, CAST(state AS UNSIGNED) as state_int
+        FROM f_polic_field
+        WHERE tenant_id = %s
+        AND CAST(state AS UNSIGNED) = %s
+        ORDER BY field_type, name
+    """
+    cursor.execute(sql, (TENANT_ID, state))
+    fields = cursor.fetchall()
+    cursor.close()
+    
+    return fields
+
+
+def enable_all_fields(conn, dry_run=True):
+    """启用所有字段"""
+    cursor = conn.cursor(pymysql.cursors.DictCursor)
+    
+    # 查询需要更新的字段（使用CAST来正确处理二进制类型）
+    sql = """
+        SELECT id, name, filed_code, field_type, CAST(state AS UNSIGNED) as state_int
+        FROM f_polic_field
+        WHERE tenant_id = %s
+        AND CAST(state AS UNSIGNED) != 1
+        ORDER BY field_type, name
+    """
+    cursor.execute(sql, (TENANT_ID,))
+    fields_to_update = cursor.fetchall()
+    
+    if not fields_to_update:
+        print("✓ 所有字段已经是启用状态，无需更新")
+        cursor.close()
+        return 0
+    
+    print(f"\n找到 {len(fields_to_update)} 个需要启用的字段:")
+    for field in fields_to_update:
+        field_type_str = "输出字段" if field['field_type'] == 2 else "输入字段"
+        print(f"  - {field['name']} ({field['filed_code']}) [{field_type_str}] (当前state={field['state_int']})")
+    
+    if dry_run:
+        print("\n⚠ 这是预览模式（dry_run=True），不会实际更新数据库")
+        cursor.close()
+        return len(fields_to_update)
+    
+    # 执行更新（使用CAST来正确比较）
+    update_sql = """
+        UPDATE f_polic_field
+        SET state = 1, updated_time = %s, updated_by = %s
+        WHERE tenant_id = %s
+        AND CAST(state AS UNSIGNED) != 1
+    """
+    
+    cursor.execute(update_sql, (CURRENT_TIME, UPDATED_BY, TENANT_ID))
+    updated_count = cursor.rowcount
+    conn.commit()
+    cursor.close()
+    
+    return updated_count
+
+
+def main():
+    """主函数"""
+    print("="*80)
+    print("启用f_polic_field表中所有字段")
+    print("="*80)
+    print()
+    
+    try:
+        conn = pymysql.connect(**DB_CONFIG)
+        print("✓ 数据库连接成功")
+    except Exception as e:
+        print(f"✗ 数据库连接失败: {str(e)}")
+        return
+    
+    try:
+        # 1. 检查当前状态统计
+        print("\n正在检查字段状态统计...")
+        stats = check_field_states(conn)
+        
+        print("\n字段状态统计:")
+        total_fields = 0
+        enabled_fields = 0
+        disabled_fields = 0
+        
+        for stat in stats:
+            state_int = stat['state_int']
+            field_type = stat['field_type']
+            count = stat['count']
+            total_fields += count
+            
+            state_str = "启用" if state_int == 1 else "未启用"
+            type_str = "输出字段" if field_type == 2 else "输入字段"
+            
+            print(f"  {type_str} - {state_str} (state={state_int}): {count} 个")
+            
+            if state_int == 1:
+                enabled_fields += count
+            else:
+                disabled_fields += count
+        
+        print(f"\n总计: {total_fields} 个字段")
+        print(f"  启用: {enabled_fields} 个")
+        print(f"  未启用: {disabled_fields} 个")
+        
+        # 2. 显示未启用的字段详情
+        if disabled_fields > 0:
+            print(f"\n正在查询未启用的字段详情...")
+            disabled_fields_list = get_fields_by_state(conn, 0)
+            print(f"\n未启用的字段列表 ({len(disabled_fields_list)} 个):")
+            for field in disabled_fields_list:
+                field_type_str = "输出字段" if field['field_type'] == 2 else "输入字段"
+                print(f"  - {field['name']} ({field['filed_code']}) [{field_type_str}]")
+        
+        # 3. 预览更新（dry_run）
+        print("\n" + "="*80)
+        print("预览更新（不会实际修改数据库）")
+        print("="*80)
+        count_to_update = enable_all_fields(conn, dry_run=True)
+        
+        if count_to_update == 0:
+            print("\n所有字段已经是启用状态，无需更新")
+            return
+        
+        # 4. 确认是否执行更新
+        print("\n" + "="*80)
+        print("准备执行更新")
+        print("="*80)
+        print(f"将更新 {count_to_update} 个字段的状态为启用（state=1）")
+        
+        # 实际执行更新
+        print("\n正在执行更新...")
+        updated_count = enable_all_fields(conn, dry_run=False)
+        
+        print(f"\n✓ 更新成功！共更新 {updated_count} 个字段")
+        
+        # 5. 验证更新结果
+        print("\n正在验证更新结果...")
+        final_stats = check_field_states(conn)
+        
+        print("\n更新后的字段状态统计:")
+        final_enabled = 0
+        final_disabled = 0
+        
+        for stat in final_stats:
+            state_int = stat['state_int']
+            field_type = stat['field_type']
+            count = stat['count']
+            
+            state_str = "启用" if state_int == 1 else "未启用"
+            type_str = "输出字段" if field_type == 2 else "输入字段"
+            
+            print(f"  {type_str} - {state_str} (state={state_int}): {count} 个")
+            
+            if state_int == 1:
+                final_enabled += count
+            else:
+                final_disabled += count
+        
+        print(f"\n总计: {final_enabled + final_disabled} 个字段")
+        print(f"  启用: {final_enabled} 个")
+        print(f"  未启用: {final_disabled} 个")
+        
+        if final_disabled == 0:
+            print("\n✓ 所有字段已成功启用！")
+        else:
+            print(f"\n⚠ 仍有 {final_disabled} 个字段未启用")
+        
+        print("\n" + "="*80)
+        print("操作完成!")
+        print("="*80)
+        
+    except Exception as e:
+        print(f"\n✗ 处理失败: {str(e)}")
+        import traceback
+        traceback.print_exc()
+        conn.rollback()
+    finally:
+        conn.close()
+
+
+if __name__ == '__main__':
+    main()
+

fix_report_card_interview_input_data.py

@@ -0,0 +1,272 @@
+"""
+修复"1.请示报告卡（初核谈话）"模板的input_data字段
+分析模板占位符，根据数据库字段对应关系生成input_data并更新数据库
+"""
+import pymysql
+import json
+import os
+import re
+from datetime import datetime
+from pathlib import Path
+from docx import Document
+
+# 数据库连接配置
+DB_CONFIG = {
+    'host': os.getenv('DB_HOST', '152.136.177.240'),
+    'port': int(os.getenv('DB_PORT', 5012)),
+    'user': os.getenv('DB_USER', 'finyx'),
+    'password': os.getenv('DB_PASSWORD', '6QsGK6MpePZDE57Z'),
+    'database': os.getenv('DB_NAME', 'finyx'),
+    'charset': 'utf8mb4'
+}
+
+TENANT_ID = 615873064429507639
+UPDATED_BY = 655162080928945152
+CURRENT_TIME = datetime.now()
+
+# 模板信息
+TEMPLATE_NAME = "1.请示报告卡（初核谈话）"
+TEMPLATE_CODE = "REPORT_CARD_INTERVIEW"
+BUSINESS_TYPE = "INVESTIGATION"
+TEMPLATE_FILE_PATH = "template_finish/2-初核模版/2.谈话审批/走读式谈话审批/1.请示报告卡（初核谈话）.docx"
+
+
+def extract_placeholders_from_docx(file_path):
+    """从docx文件中提取所有占位符"""
+    placeholders = set()
+    pattern = r'\{\{([^}]+)\}\}'
+    
+    try:
+        doc = Document(file_path)
+        
+        # 从段落中提取占位符
+        for paragraph in doc.paragraphs:
+            text = paragraph.text
+            matches = re.findall(pattern, text)
+            for match in matches:
+                placeholders.add(match.strip())
+        
+        # 从表格中提取占位符
+        for table in doc.tables:
+            for row in table.rows:
+                for cell in row.cells:
+                    for paragraph in cell.paragraphs:
+                        text = paragraph.text
+                        matches = re.findall(pattern, text)
+                        for match in matches:
+                            placeholders.add(match.strip())
+        
+    except Exception as e:
+        print(f"  错误: 读取文件失败 - {str(e)}")
+        return []
+    
+    return sorted(list(placeholders))
+
+
+def get_template_config(conn):
+    """查询模板配置"""
+    cursor = conn.cursor(pymysql.cursors.DictCursor)
+    
+    sql = """
+        SELECT id, name, template_code, input_data, file_path, state
+        FROM f_polic_file_config
+        WHERE tenant_id = %s AND name = %s
+    """
+    cursor.execute(sql, (TENANT_ID, TEMPLATE_NAME))
+    config = cursor.fetchone()
+    cursor.close()
+    
+    return config
+
+
+def get_template_fields(conn, file_config_id):
+    """查询模板关联的字段"""
+    cursor = conn.cursor(pymysql.cursors.DictCursor)
+    
+    sql = """
+        SELECT f.id, f.name, f.filed_code as field_code, f.field_type
+        FROM f_polic_field f
+        INNER JOIN f_polic_file_field ff ON f.id = ff.filed_id
+        WHERE ff.file_id = %s
+        AND f.tenant_id = %s
+        ORDER BY f.field_type, f.name
+    """
+    cursor.execute(sql, (file_config_id, TENANT_ID))
+    fields = cursor.fetchall()
+    cursor.close()
+    
+    return fields
+
+
+def verify_placeholders_in_database(conn, placeholders):
+    """验证占位符是否在数据库中存在对应的字段"""
+    if not placeholders:
+        return {}
+    
+    cursor = conn.cursor(pymysql.cursors.DictCursor)
+    
+    placeholders_list = list(placeholders)
+    placeholders_str = ','.join(['%s'] * len(placeholders_list))
+    
+    # 查询所有字段（包括未启用的）
+    sql = f"""
+        SELECT id, name, filed_code as field_code, field_type, state
+        FROM f_polic_field
+        WHERE tenant_id = %s
+        AND filed_code IN ({placeholders_str})
+    """
+    cursor.execute(sql, [TENANT_ID] + placeholders_list)
+    fields = cursor.fetchall()
+    cursor.close()
+    
+    # 构建字段映射
+    field_map = {f['field_code']: f for f in fields}
+    
+    # 检查缺失的字段
+    missing_fields = set(placeholders) - set(field_map.keys())
+    
+    return {
+        'found_fields': field_map,
+        'missing_fields': missing_fields
+    }
+
+
+def update_input_data(conn, file_config_id, input_data):
+    """更新input_data字段"""
+    cursor = conn.cursor()
+    
+    input_data_str = json.dumps(input_data, ensure_ascii=False)
+    
+    update_sql = """
+        UPDATE f_polic_file_config
+        SET input_data = %s, updated_time = %s, updated_by = %s
+        WHERE id = %s
+    """
+    cursor.execute(update_sql, (input_data_str, CURRENT_TIME, UPDATED_BY, file_config_id))
+    conn.commit()
+    cursor.close()
+
+
+def main():
+    """主函数"""
+    print("="*80)
+    print("修复'1.请示报告卡（初核谈话）'模板的input_data字段")
+    print("="*80)
+    print()
+    
+    # 1. 检查模板文件是否存在
+    template_path = Path(TEMPLATE_FILE_PATH)
+    if not template_path.exists():
+        print(f"✗ 错误: 模板文件不存在 - {TEMPLATE_FILE_PATH}")
+        return
+    
+    print(f"✓ 找到模板文件: {TEMPLATE_FILE_PATH}")
+    
+    # 2. 提取占位符
+    print("\n正在提取占位符...")
+    placeholders = extract_placeholders_from_docx(str(template_path))
+    print(f"✓ 找到 {len(placeholders)} 个占位符:")
+    for i, placeholder in enumerate(placeholders, 1):
+        print(f"  {i}. {{{{ {placeholder} }}}}")
+    
+    # 3. 连接数据库
+    print("\n正在连接数据库...")
+    try:
+        conn = pymysql.connect(**DB_CONFIG)
+        print("✓ 数据库连接成功")
+    except Exception as e:
+        print(f"✗ 数据库连接失败: {str(e)}")
+        return
+    
+    try:
+        # 4. 查询模板配置
+        print(f"\n正在查询模板配置: {TEMPLATE_NAME}")
+        config = get_template_config(conn)
+        
+        if not config:
+            print(f"✗ 未找到模板配置: {TEMPLATE_NAME}")
+            return
+        
+        print(f"✓ 找到模板配置:")
+        print(f"  ID: {config['id']}")
+        print(f"  名称: {config['name']}")
+        print(f"  当前template_code: {config.get('template_code', 'NULL')}")
+        print(f"  当前input_data: {config.get('input_data', 'NULL')}")
+        print(f"  文件路径: {config.get('file_path', 'NULL')}")
+        print(f"  状态: {config.get('state', 0)}")
+        
+        file_config_id = config['id']
+        
+        # 5. 查询模板关联的字段
+        print(f"\n正在查询模板关联的字段...")
+        template_fields = get_template_fields(conn, file_config_id)
+        print(f"✓ 找到 {len(template_fields)} 个关联字段:")
+        for field in template_fields:
+            field_type_str = "输出字段" if field['field_type'] == 2 else "输入字段"
+            print(f"  - {field['name']} ({field['field_code']}) [{field_type_str}]")
+        
+        # 6. 验证占位符是否在数据库中存在
+        print(f"\n正在验证占位符...")
+        verification = verify_placeholders_in_database(conn, placeholders)
+        
+        found_fields = verification['found_fields']
+        missing_fields = verification['missing_fields']
+        
+        print(f"✓ 在数据库中找到 {len(found_fields)} 个字段:")
+        for field_code, field in found_fields.items():
+            field_type_str = "输出字段" if field['field_type'] == 2 else "输入字段"
+            state_str = "启用" if field.get('state', 0) == 1 else "未启用"
+            print(f"  - {field['name']} ({field_code}) [{field_type_str}] [状态: {state_str}]")
+        
+        if missing_fields:
+            print(f"\n⚠ 警告: 以下占位符在数据库中未找到对应字段:")
+            for field_code in missing_fields:
+                print(f"  - {field_code}")
+            print("\n这些占位符仍会被包含在input_data中，但可能无法正确填充。")
+        
+        # 7. 生成input_data
+        print(f"\n正在生成input_data...")
+        input_data = {
+            'template_code': TEMPLATE_CODE,
+            'business_type': BUSINESS_TYPE,
+            'placeholders': placeholders
+        }
+        
+        print(f"✓ input_data内容:")
+        print(json.dumps(input_data, ensure_ascii=False, indent=2))
+        
+        # 8. 更新数据库
+        print(f"\n正在更新数据库...")
+        update_input_data(conn, file_config_id, input_data)
+        print(f"✓ 更新成功!")
+        
+        # 9. 验证更新结果
+        print(f"\n正在验证更新结果...")
+        updated_config = get_template_config(conn)
+        if updated_config:
+            try:
+                updated_input_data = json.loads(updated_config['input_data'])
+                if updated_input_data.get('template_code') == TEMPLATE_CODE:
+                    print(f"✓ 验证成功: template_code = {TEMPLATE_CODE}")
+                if updated_input_data.get('business_type') == BUSINESS_TYPE:
+                    print(f"✓ 验证成功: business_type = {BUSINESS_TYPE}")
+                if set(updated_input_data.get('placeholders', [])) == set(placeholders):
+                    print(f"✓ 验证成功: placeholders 匹配")
+            except Exception as e:
+                print(f"⚠ 验证时出错: {str(e)}")
+        
+        print("\n" + "="*80)
+        print("修复完成!")
+        print("="*80)
+        
+    except Exception as e:
+        print(f"\n✗ 处理失败: {str(e)}")
+        import traceback
+        traceback.print_exc()
+    finally:
+        conn.close()
+
+
+if __name__ == '__main__':
+    main()
+