7python之把宠物档案加入到用户问题中及去除历史记录中的脏数据

写在前面的话

不回。刚上号那一段时间，还是比较在乎每个上号人的。就像新店开业，对每一个顾客都很热情。当时一天上午，一个叫“夏日暖阳阳的”，突然就加了微信，说上号。我说好好好，赶紧准备好窗口，以最快的速度，把码发给她，结果没音了，我发微信也不回了，那一天，我每隔一两小时给她微一次，本来晚上21点离开图书馆，我等到了22点，还是不回，我失落的回去了。过了一周，我刚到图书馆楼下，她说：现在有时间了，上号。我非常兴奋，摩托车没停稳，倒了，来不及扶，卡卡跑到楼上，打开电脑，发码让她扫，又没音了。等了五个小时，我彻底怒了，这是溜我玩的吗，你就是真急事了，也说一声啊。果断拉黑，逼自己忘记这个人这件事。对于不讲礼貌的人，没必要在乎她，再也不见了您。

[200+100]-------->底部有张生活照片(头条号运营：大家想全托管上号的联系我哦，每天让你得个早餐钱，微信号: qhz_toutiao)

【关键词】python、ragflow、宠物档案、脏数据

一、宠物档案加入用户问题（三级）

描述：现在需要把宠物档案加入用户问题，减少基础问询的提问问题数量。

开工：

第一步：看代码（四级）

20250509时间段：12:06-13:00

看下前端把当前宠物信息传过来没有，在新建会话时。

在新建会话时，宠物档案是传进来的，如下：

图7a-1

注：接下来，先把这个信息存到一张动物表里。

第二步：动物表（四级）

20250510时间段：01:46-02:00

先建一张表，按照上面截图中的字段，如下：

CREATE TABLE `animal` (
  `id` varchar(32) NOT NULL,
  `create_time` bigint DEFAULT NULL,
  `create_date` datetime DEFAULT NULL,
  `update_time` bigint DEFAULT NULL,
  `update_date` datetime DEFAULT NULL,
  `animal_id` int(11) DEFAULT 0,
  `animal` varchar(50) DEFAULT NULL,
  `animal_type` varchar(100) DEFAULT NULL,
  `avatar` varchar(100) DEFAULT NULL,
  `birthday` varchar(50) DEFAULT NULL,
  `conversation_id` varchar(32) DEFAULT NULL,
  `dialog_id` varchar(32) DEFAULT NULL,
  `is_choosed` tinyint(1) DEFAULT 0,
  `is_deleted` tinyint(1) DEFAULT 0,
  `neuter` tinyint(1) DEFAULT 0,
  `nick` varchar(50) DEFAULT NULL,
  `openid` varchar(50) DEFAULT NULL,
  `sex` varchar(5) DEFAULT NULL,
  `speical_time` varchar(50) DEFAULT NULL,
  `vaccination` varchar(50) DEFAULT NULL,  
  PRIMARY KEY (`id`),
  KEY `animal_create_time` (`create_time`),
  KEY `animal_create_date` (`create_date`),
  KEY `animal_update_time` (`update_time`),
  KEY `animal_update_date` (`update_date`),
  KEY `animal_dialog_id` (`dialog_id`),
  KEY `animal_conversation_id` (`conversation_id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_0900_ai_ci;

注：运行一下试试。效果如下：

图7a-2

注：接下来，就是保存宠物档案信息到表，有就更新，没有就增加。

第三步：保存宠物档案（四级）

20250510时间段：02:05-04:00

写程序如下：

            ###保存宠物档案信息
            if 'id' in req and req["id"]:
                save_data = {
                    "id": get_uuid(),
                    "animal_id": req["id"],
                    "animal": req["animal"],
                    "animal_type": req["animal_type"], 
                    "avatar": req["avatar"],
                    "birthday": req["birthday"], 
                    "conversation_id": req["conversation_id"],
                    "dialog_id": req["dialog_id"],
                    "is_choosed": req["is_choosed"], 
                    "is_deleted": req["is_deleted"],
                    "neuter": req["neuter"], 
                    "nick": req["nick"],
                    "openid": req["openid"],
                    "sex": req["sex"], 
                    "speical_time": req["speical_time"],
                    "vaccination": req["vaccination"]
                }
                AnimalService.save(**save_data)

注：这个通过接口模拟来做，但这个涉及到登录验证，看下能不能攻破。

第四步：本地模拟（四级）

20250512时间段：16:52-17:00

先搞一下数据出来，如下：

{
  "add_time": "2025-05-09 12:30:13",
  "animal": "狗狗",
  "animal_type": "澳大利亚牧牛犬",
  "avatar": "",
  "birthday": "2025-05-09",
  "conversation_id": "b1210beb53d245199e3f9a8a93641237",
  "dialog_id": "9cf3d9eaf77c11ef8dc500e003c42347",
  "enable_search": false,
  "id": "38",
  "is_choosed": "0",
  "is_deleted": "0",
  "message": [
    {
      "role": "assistant",
      "content": "你好"
    }
  ],
  "modify_time": "2025-05-09 12:30:13",
  "name": "你好",
  "neuter": "0",
  "nick": "土豆",
  "openid": "oi87O6xKvGO_LwalOFaXmJ8Yswi8",
  "sex": "男",
  "speical_time": "非特殊时期",
  "vaccination": "已接种"
}

注：接下来，写个测试用例，如下：

def test_set_conversation(client):
    log.info("test_expert_diagnosis")
    '''
     测试 设置会话时保存宠物档案信息
    '''
    json_data = {
        "add_time": "2025-05-09 12:30:13",
        "animal": "狗狗",
        "animal_type": "澳大利亚牧牛犬",
        "avatar": "",
        "birthday": "2025-05-09",
        "conversation_id": "b1210beb53d245199e3f9a8a93641237",
        "dialog_id": "9cf3d9eaf77c11ef8dc500e003c42347",
        "enable_search": false,
        "id": "38",
        "is_choosed": "0",
        "is_deleted": "0",
        "message": [
            {
            "role": "assistant",
            "content": "你好"
            }
        ],
        "modify_time": "2025-05-09 12:30:13",
        "name": "你好",
        "neuter": "0",
        "nick": "土豆",
        "openid": "oi87O6xKvGO_LwalOFaXmJ8Yswi8",
        "sex": "男",
        "speical_time": "非特殊时期",
        "vaccination": "已接种",
    }

    url = f"/v1/conversation/set"
    resp = client.post(
        url,
        json=json_data,
        headers={
            "Content-type": "application/json",
            "Authorization": "Bearer ragflow-UxOGYzZjUwYjMwOTExZWZiODc0MDI0Mm"
        }
    )
    if not 200 <= resp.status_code < 300:
        raise Exception(f"GET {url} status_code {resp.status_code}.")

    # received_data = []
    for chunk in resp.iter_encoded():
        answer = chunk.decode('utf-8').strip()
        log.info(f"\n\n answer: {answer} \n\n")     

注：运行下试试。

二、历史记录脏数据（三级）

描述：现在查看历史记录有脏数据，截图如下：

图7b-1

注：这个问题需要处理一下。

开工：

第一步：排查（四级）

20250509周五时间段：16:48-17:00

可能是专家诊断里，没有去除系统消息，打断点看一下。

这个已经处理，是由于专家诊断里，没有删除系统消息，导致存到数据库里了。

第二步：测试

测试还可以

三、配置审计日志（三级）

描述：现在集群莫名的丢失索引，需要配置审计日志，以便下次索引被删除知道原因。

开工：

第一步：配置（四级）

20250511周日时间段：23:27-01:00

Docker环境下Elasticsearch审计日志配置

在Docker环境中配置Elasticsearch审计日志需要注意一些特殊事项。以下是针对Docker部署的修改建议和完整配置方案：

关键修改点

1. 日志路径配置

Docker容器内的路径需要特别注意，建议使用环境变量覆盖：

xpack.security.audit.logfile.path: /usr/share/elasticsearch/logs/audit.log

2. 日志持久化

必须将日志目录挂载到宿主机，否则容器重启后日志会丢失：

docker run -d \
  -v /path/on/host/logs:/usr/share/elasticsearch/logs \
  -v /path/on/host/config/elasticsearch.yml:/usr/share/elasticsearch/config/elasticsearch.yml \
  elasticsearch:8.12.0

注：接下来，看下完整的Docker配置方案。

第二步：完整的Docker配置方案（四级）

方案1：通过elasticsearch.yml挂载

1.准备elasticsearch.yml（宿主机文件）：

xpack.security.enabled: true
xpack.security.audit.enabled: true
xpack.security.audit.logfile.events.include: index_event,create_index,delete_index
xpack.security.audit.logfile.path: /usr/share/elasticsearch/logs/audit.log

2.启动命令：

docker run -d \
  --name elasticsearch \
  -p 9200:9200 -p 9300:9300 \
  -v ./elasticsearch.yml:/usr/share/elasticsearch/config/elasticsearch.yml \
  -v ./es-logs:/usr/share/elasticsearch/logs \
  -e "discovery.type=single-node" \
  elasticsearch:8.12.0

注：现在还不行，先把配置审计日志注释掉，索引恢复后再打开。

第三步：注释掉审计日志配置（四级）

20250512周一时间段：10:03-11:00

现在日志如下：

ess.thread.name":"main","log.logger":"org.elasticsearch.plugins.PluginsService","elasticsearch.node.name":"zero4","elasticsearch.cluster.name":"ragflow-cluster"}
{"@timestamp":"2025-05-12T02:21:47.771Z", "log.level": "INFO", "message":"loaded module [ingest-geoip]", "ecs.version": "1.2.0","service.name":"ES_ECS","event.dataset":"elasticsearch.server","process.thread.name":"main","log.logger":"org.elasticsearch.plugins.PluginsService","elasticsearch.node.name":"zero4","elasticsearch.cluster.name":"ragflow-cluster"}
{"@timestamp":"2025-05-12T02:21:47.772Z", "log.level": "INFO", "message":"loaded module [x-pack-write-load-forecaster]", "ecs.version": "1.2.0","service.name":"ES_ECS","event.dataset":"elasticsearch.server","process.thread.name":"main","log.logger":"org.elasticsearch.plugins.PluginsService","elasticsearch.node.name":"zero4","elasticsearch.cluster.name":"ragflow-cluster"}
{"@timestamp":"2025-05-12T02:21:47.772Z", "log.level": "INFO", "message":"loaded module [inference]", "ecs.version": "1.2.0","service.name":"ES_ECS","event.dataset":"elasticsearch.server","process.thread.name":"main","log.logger":"org.elasticsearch.plugins.PluginsService","elasticsearch.node.name":"zero4","elasticsearch.cluster.name":"ragflow-cluster"}
{"@timestamp":"2025-05-12T02:21:47.772Z", "log.level": "INFO", "message":"loaded module [search-business-rules]", "ecs.version": "1.2.0","service.name":"ES_ECS","event.dataset":"elasticsearch.server","process.thread.name":"main","log.logger":"org.elasticsearch.plugins.PluginsService","elasticsearch.node.name":"zero4","elasticsearch.cluster.name":"ragflow-cluster"}
{"@timestamp":"2025-05-12T02:21:47.772Z", "log.level": "INFO", "message":"loaded module [wildcard]", "ecs.version": "1.2.0","service.name":"ES_ECS","event.dataset":"elasticsearch.server","process.thread.name":"main","log.logger":"org.elasticsearch.plugins.PluginsService","elasticsearch.node.name":"zero4","elasticsearch.cluster.name":"ragflow-cluster"}
{"@timestamp":"2025-05-12T02:21:47.772Z", "log.level": "INFO", "message":"loaded module [ingest-attachment]", "ecs.version": "1.2.0","service.name":"ES_ECS","event.dataset":"elasticsearch.server","process.thread.name":"main","log.logger":"org.elasticsearch.plugins.PluginsService","elasticsearch.node.name":"zero4","elasticsearch.cluster.name":"ragflow-cluster"}
{"@timestamp":"2025-05-12T02:21:47.772Z", "log.level": "INFO", "message":"loaded module [x-pack-sql]", "ecs.version": "1.2.0","service.name":"ES_ECS","event.dataset":"elasticsearch.server","process.thread.name":"main","log.logger":"org.elasticsearch.plugins.PluginsService","elasticsearch.node.name":"zero4","elasticsearch.cluster.name":"ragflow-cluster"}
{"@timestamp":"2025-05-12T02:21:47.772Z", "log.level": "INFO", "message":"loaded module [unsigned-long]", "ecs.version": "1.2.0","service.name":"ES_ECS","event.dataset":"elasticsearch.server","process.thread.name":"main","log.logger":"org.elasticsearch.plugins.PluginsService","elasticsearch.node.name":"zero4","elasticsearch.cluster.name":"ragflow-cluster"}
{"@timestamp":"2025-05-12T02:21:47.772Z", "log.level": "INFO", "message":"loaded module [x-pack-async]", "ecs.version": "1.2.0","service.name":"ES_ECS","event.dataset":"elasticsearch.server","process.thread.name":"main","log.logger":"org.elasticsearch.plugins.PluginsService","elasticsearch.node.name":"zero4","elasticsearch.cluster.name":"ragflow-cluster"}
{"@timestamp":"2025-05-12T02:21:47.773Z", "log.level": "INFO", "message":"loaded module [runtime-fields-common]", "ecs.version": "1.2.0","service.name":"ES_ECS","event.dataset":"elasticsearch.server","process.thread.name":"main","log.logger":"org.elasticsearch.plugins.PluginsService","elasticsearch.node.name":"zero4","elasticsearch.cluster.name":"ragflow-cluster"}
{"@timestamp":"2025-05-12T02:21:47.773Z", "log.level": "INFO", "message":"loaded module [vector-tile]", "ecs.version": "1.2.0","service.name":"ES_ECS","event.dataset":"elasticsearch.server","process.thread.name":"main","log.logger":"org.elasticsearch.plugins.PluginsService","elasticsearch.node.name":"zero4","elasticsearch.cluster.name":"ragflow-cluster"}
{"@timestamp":"2025-05-12T02:21:47.773Z", "log.level": "INFO", "message":"loaded module [lang-expression]", "ecs.version": "1.2.0","service.name":"ES_ECS","event.dataset":"elasticsearch.server","process.thread.name":"main","log.logger":"org.elasticsearch.plugins.PluginsService","elasticsearch.node.name":"zero4","elasticsearch.cluster.name":"ragflow-cluster"}
{"@timestamp":"2025-05-12T02:21:47.773Z", "log.level": "INFO", "message":"loaded module [x-pack-eql]", "ecs.version": "1.2.0","service.name":"ES_ECS","event.dataset":"elasticsearch.server","process.thread.name":"main","log.logger":"org.elasticsearch.plugins.PluginsService","elasticsearch.node.name":"zero4","elasticsearch.cluster.name":"ragflow-cluster"}

注：这个配置还是有点问题，先把保存宠物档案做一下，以免影响用户使用。之后，晚上的时候再做这个。

四、头条战果汇报

昨日数据来啦，昨日总收入:807.24，昨日总播放:911.8万，软件截图如下：

图7d-1

注：想要全脱管运营头条号的联系我，你出账号，我来运营，保你天天有钱花，咨询电话： 17701328814（微信同号），也可以加群先了解一下。

图7d-2

注：个人微信如下，欢迎骚扰。

图7d-3

五、生活照片

拍摄于‎2026‎年‎2‎月‎1‎日，‏‎10:49:12，带二宝去看花，二宝当时三岁四个月。把那个女孩拉黑后，我在脑补那个女孩的出生环境，应该是家里比较富裕，独生子女，被惯着长大的，对别人没礼貌，不在乎，但社会不会惯着啊，别人又不是女孩爸妈，对家人不客气，对社会人需要客气吧。就像之前在六道口上班时，有一天去图书部，一个黄头发的库管和一个女销售吵起来了，库管：我TM在地铁上，你就说要医学综合，到工位上还一直催催催，我说好，我来找，我把大综合，小综合还有附加卷给你配好了，问你要哪一种，你就不回了，把我晾在仓库，前后也就五分钟，等了你一天你都没回，你是被车撞死了吗，站在我面前的是鬼魂吗，你有礼貌吗，你父母教你做人了吗。销售姑娘把书一摔，愤愤离去。我说：怎么了，兄弟，吃枪药了。库管：你不知道，那天，我没睡醒就催，在地铁上还催，到楼上了，我说我现在给你找书，大概五分钟，我全扒拉出来了，问她要大综合，小缩合，要不要附加卷，她没音了，等了她一上午也没回，领导来了，说书那么乱，还把我熊一顿。库管做的不对，不应该对小姑娘那样，缺少怜香惜玉之心，但小姑娘也有错，你好歹回一句啊，害得库管被骂，都有问题，各打五十大板吧。

图7e-1

《本文完》