itam-assistant 0.1.4__py3-none-any.whl → 0.1.5__py3-none-any.whl

itam_assistant/do_ai.py

@@ -0,0 +1,928 @@
+# -*- coding: utf-8 -*-
+import time
+from itam_assistant.ailyapp_client import AilyLarkClient
+from itam_assistant.lark_client import LarkdocsClient
+from itam_assistant.intent_detail import *
+from itam_assistant.openapi import *
+import datetime
+import copy
+import os
+import csv
+
+
+# 定义一个全局变量Client
+# Testsuitelink = "https://bytedance.larkoffice.com/sheets/ZVzfsw4rMhkMF6tjtxmc4BdSnMb"
+
+
+def do_ai_auto(Testk_data, clientinfo):
+    """
+    自动化执行AI测试用例
+    """
+    startAt = 0
+    try:
+        # 获取租户访问令牌
+        tenant_access_token = AilyLarkClient(clientinfo).get_tenant_access_token()
+        if not tenant_access_token:
+            raise ValueError("未能获取到有效的租户访问令牌")
+        # 判断Testsuitelink中是否包含https://
+        if "https://" in Testk_data:
+
+            # 通过文档链接获取spreadsheet_token
+            spreadsheet_token = Testk_data.split("/")[-1]
+            if not spreadsheet_token:
+                raise ValueError("未能从文档链接中提取到有效的spreadsheet_token")
+            # 读取表格用户输入
+            spreadsheet = LarkdocsClient().get_the_worksheet(spreadsheet_token)
+            if not spreadsheet:
+                raise ValueError("未能获取到有效的工作表数据")
+            for i in spreadsheet.sheets:
+                column_count = i.grid_properties.column_count
+                row_count = i.grid_properties.row_count
+                sheet_id = i.sheet_id
+                title = i.title
+                if title == "测试集":
+                    # 构建JSON字符串
+                    json_str = {"ranges": [sheet_id + "!A1:A" + str(row_count)]}
+                    # 获取纯文本内容
+                    test = LarkdocsClient().get_plaintextcontent(json_str, spreadsheet_token, sheet_id)
+                    test = json.loads(test)
+                    userinput = test['data']['value_ranges'][0]['values']
+                    print(f"表头为{userinput[0]}")
+                    for i in range(1, row_count):
+                        if userinput[i][0]:
+                            if startAt == 0:
+                                startAt = int(time.time())
+                            # 创建会话
+                            seseion_id = AilyLarkClient(clientinfo).create_ailysession(tenant_access_token)
+                            if not seseion_id:
+                                raise ValueError("未能成功创建会话")
+                            # 创建消息
+                            message_id = AilyLarkClient(clientinfo).create_ailysessionaily_message(tenant_access_token,
+                                                                                                   seseion_id,
+                                                                                                   userinput[i][0])
+                            if not message_id:
+                                raise ValueError("未能成功创建消息")
+                            # 创建运行实例
+                            runs = AilyLarkClient(clientinfo).create_ailysession_run(tenant_access_token, seseion_id)
+                            # 可不需等待运行实例创建完成
+                            # if not runs:
+                            #    raise ValueError("未能成功创建运行实例")
+                            time.sleep(5)
+                        else:
+                            return startAt, i
+                            break
+                    return startAt, row_count
+                    break
+        elif Testk_data[0].get('ext'):
+            num = 0
+            for i in Testk_data:
+                aa = i['ext']['input']
+                if startAt == 0:
+                    startAt = int(time.time())
+                # 创建会话
+                seseion_id = AilyLarkClient(clientinfo).create_ailysession(tenant_access_token)
+                time.sleep(10)
+                if not seseion_id:
+                    raise ValueError("未能成功创建会话")
+                # 创建消息
+                message_id = AilyLarkClient(clientinfo).create_ailysessionaily_message(tenant_access_token, seseion_id,
+                                                                                       aa)
+                time.sleep(5)
+                if not message_id:
+                    raise ValueError("未能成功创建消息")
+                # 创建运行实例
+                runs = AilyLarkClient(clientinfo).create_ailysession_run(tenant_access_token, seseion_id)
+                time.sleep(5)
+                num = num + 1
+            return startAt, num
+    except KeyError as ke:
+        print(f"KeyError 发生: 数据中缺少必要的键，错误详情: {ke}")
+        return None, None
+    except json.JSONDecodeError as jde:
+        print(f"JSON 解析错误: {jde}")
+        return None, None
+    except ValueError as ve:
+        print(f"值错误: {ve}")
+        return None, None
+    except Exception as e:
+        print(f"发生未知错误: {e}")
+        return None, None
+
+
+def get_results(num,res):
+    """
+    获取结果
+    """
+    labels = []
+    a = len(res["body"]["Results"])
+    #判断res["body"]["Results"]的个数，如果大于num，则取num个值，小于num，则取res["body"]["Results"]的个数
+    if a > num:
+        num = num
+    else:
+        num = a
+    #提取res["body"]["Results"]的前num个值
+    res["body"]["Results"] = res["body"]["Results"][:num]
+    #遍历res["body"]["Results"]提取 name_zh、brand_zh、model_zh、specification_zh 并拼接
+    for i in range(0, num):
+        res["body"]["Results"][i]['Item']['sku_zh'] = res["body"]["Results"][i]['Item']['name_zh'] + " " + res["body"]["Results"][i]['Item']['brand_zh'] + " " + res["body"]["Results"][i]['Item']['model_zh'] + " " + res["body"]["Results"][i]['Item']['specification_zh']
+        label = {"label" :res["body"]["Results"][i]['Item']['sku_zh'], "score" : res["body"]["Results"][i]['Score']}
+        labels.append(copy.deepcopy(label))
+    return labels
+
+
+
+def do_waterlevellineres_list(res, info):
+    """
+    获取结果，并组装水位线info
+    """
+    if res == '':
+        info['label'] = [{'label': '', 'score': 0}, {'label': '', 'score': 0}]
+        info['rt'] = False
+        return info
+
+    # 判断res["body"]["Results"]不为空，空则：label0 label1 为空，label默认2级
+    if res["body"]["Results"]:
+        info['output']['用户输入/output']='log_id:'+res['log_id']
+        # 取["Results"]下前2个结果，若只有1个结果，label1为空
+        if len(res["body"]["Results"]) > 0:
+            for j in range(len(res["body"]["Results"])):
+                aaa = {'label': res["body"]["Results"][j]['Item']['name_zh']+"&"+res["body"]["Results"][j]['Item']['brand_zh']+"&"+res["body"]["Results"][j]['Item']['model_zh']+"&"+res["body"]["Results"][j]['Item']['specification_zh'],
+                       'score': res["body"]["Results"][j]['Score']}
+                info['label'].append(copy.deepcopy(aaa))
+        # 判断label0和label1是否为空，为空则：label默认2级
+        # 判断exp和label是否一致，一致则：rt=True，不一致则：rt=False
+        if info['exp'][0]['label'] == info['label'][0]['label'] and info['exp'][1]['label'] == info['label'][1][
+            'label'] and info['exp'][0]['score'] <= info['label'][0]['score'] and info['exp'][1]['score'] <= \
+                info['label'][1]['score']:
+            info['rt'] = True
+        else:
+            info['rt'] = False
+
+    else:
+        info['label'] = info['label'] = [{'label': '', 'score': 0}, {'label': '', 'score': 0}]
+        info['rt'] = False
+
+    return info
+
+def do_waterlevellineres_listv2(res, info):
+    """
+    获取结果，并组装水位线info
+    """
+    if res == '':
+        info['label'] = [{'label': '', 'score': 0}, {'label': '', 'score': 0}]
+        info['rt'] = False
+        return info
+    reslist = res["body"]["Results"]
+    info['output']['用户输入/output'] = 'log_id:' + res.get('log_id') or res.get('requestId')
+    if reslist:
+        #取所有结果并追加到info['label']
+        info['label'] = []
+        if len(reslist) > 0:
+            for j in range(len(reslist)):
+                aaa = {'label': reslist[j]['Item']['name_zh'],
+                       'score': reslist[j]['Score'],
+                       'info':reslist[j]['Item']['name_zh']+"&"+reslist[j]['Item']['brand_zh']+"&"+reslist[j]['Item']['model_zh']+"&"+reslist[j]['Item']['specification_zh'],}
+                info['label'].append(copy.deepcopy(aaa))
+        # 判断exp和label是否一致，一致则：rt=True，不一致则：rt=False
+        for a in range(len(info['exp'])):
+            if info['exp'][a]['label']== info['label'][a]['label']:
+                info['rt'] = True
+            else:
+                info['rt'] = False
+                break
+    else:
+        info['label'] = info['label'] = [{'label': '', 'score': 0}, {'label': '', 'score': 0}]
+        info['rt'] = False
+
+    return info
+
+def do_waterlevellineres_listassetspu(res, info,hardtype):
+    """
+    获取结果，并组装水位线info
+    """
+    if res == '':
+        info['label'] = [{'label': '', 'score': 0}]
+        info['rt'] = False
+        return info
+    if res["data"].get("AiBorrowAndUseResponseList") == None:
+        info['label'] = [{'label': '', 'score': 0},{'label': '', 'score': 0.8}]
+        info['rt'] = False
+        if info['exp']==[]:
+            info['rt'] = True
+        return info
+    reslist=res["data"]["AiBorrowAndUseResponseList"]
+    info['output']['用户输入/output']='requestId:'+res['requestId']
+    aaa ={}
+    # 判断res["body"]["Results"]不为空，空则：label0 label1 为空，label默认2级
+    if reslist:
+        # 取["Results"]下前2个结果，若只有1个结果，label1为空
+        if len(reslist) > 0:
+           for j in range(len(reslist)):
+               if hardtype==2:
+                   aaa = {'label': reslist[j]['AccessoryModelScope']['AccessoryModelInfo']['Name']['ValueZh'],
+                          'score': reslist[j]['Score']}
+                   info['label'].append(copy.deepcopy(aaa))
+               if hardtype == 1:
+                   aaa = {'label': reslist[j]['AssetModelScope']['SpuNameZh'] or reslist[j]['AssetModelScope']['NameZh']+reslist[j]['AssetModelScope']['ModelZh']+reslist[j]['AssetModelScope']['SpecificationZh'],
+                          'score': reslist[j]['Score']}
+                   info['label'].append(copy.deepcopy(aaa))
+        # 判断label0和label1是否为空，为空则：label默认2级
+        for a in range(len(info['exp'])):
+            if info['exp'][a]['label'] == info['label'][a]['label']:
+                info['rt'] = True
+            else:
+                info['rt'] = False
+                break
+    else:
+        info['label'] = info['label'] = [{'label': '', 'score': 0}, {'label': '', 'score': 0}]
+        info['rt'] = False
+    return info
+
+def do_waterlevellineres_listassetspu_pre(res, info,hardtype):
+    """
+    获取结果，并组装水位线info
+    """
+    if res == '':
+        info['label'] = [{'label': '', 'score': 0}]
+        info['rt'] = False
+        return info
+    if res.get("body") == None:
+        info['label'] = [{'label': '', 'score': 0},{'label': '', 'score': 0.8}]
+        info['rt'] = False
+        if info['exp'][0]['label']=='':
+            info['rt'] = True
+        return info
+    reslist=res["body"]
+    aaa ={}
+    # 判断res["body"]["Results"]不为空，空则：label0 label1 为空，label默认2级
+    if reslist:
+        # 取["Results"]下前2个结果，若只有1个结果，label1为空
+        if len(reslist) > 0:
+           for j in range(len(reslist)):
+               if hardtype==2:
+                   aaa = {'label': reslist[j]['AccessoryModelScope']['AccessoryModelInfo']['Name']['ValueZh'],
+                          'score': reslist[j]['Score']}
+                   info['label'].append(copy.deepcopy(aaa))
+               if hardtype == 1:
+                   aaa = {'label': reslist[j]['Item']['name_zh']+reslist[j]['Item']['brand_zh']+reslist[j]['Item']['model_zh']+reslist[j]['Item']['specification_zh'],
+                          'score': reslist[j]['Score']}
+                   info['label'].append(copy.deepcopy(aaa))
+        # 判断label0和label1是否为空，为空则：label默认2级
+        # 判断exp和label是否一致，一致则：rt=True，不一致则：rt=False
+        if info['exp'][0]['label'] == info['label'][0]['label'] and info['exp'][1]['label'] == info['label'][1][
+            'label'] and info['exp'][0]['score'] <= info['label'][0]['score'] and info['exp'][1]['score'] <= \
+                info['label'][1]['score']:
+            info['rt'] = True
+        else:
+            info['rt'] = False
+    else:
+        info['label'] = info['label'] = [{'label': '', 'score': 0}, {'label': '', 'score': 0}]
+        info['rt'] = False
+    return info
+
+def do_waterlevellineres_sr(res, info,hardtype):
+    """
+    获取结果，并组装水位线info
+    """
+    if res == '':
+        info['label'] = [{'label': '', 'score': 0}]
+        info['rt'] = False
+        return info
+    if res["data"].get("AiBorrowAndUseResponseList") == None:
+        info['label'] = [{'label': '', 'score': 0},{'label': '', 'score': 0.8}]
+        info['rt'] = False
+        if info['exp'][0]['label']=='':
+            info['rt'] = True
+        return info
+    reslist=res["data"]["AssetModels"]
+    aaa ={}
+    # 判断res["body"]["Results"]不为空，空则：label0 label1 为空，label默认2级
+    if reslist:
+        # 取["Results"]下前2个结果，若只有1个结果，label1为空
+        if len(reslist) > 0:
+           for j in range(len(reslist)):
+               if hardtype==2:
+                   aaa = {'label': reslist[j]['AccessoryModelScope']['AccessoryModelInfo']['Name']['ValueZh'],
+                          'score': reslist[j]['Score']}
+                   info['label'].append(copy.deepcopy(aaa))
+               if hardtype == 1:
+                   aaa = {'label': reslist[j]['SpuNameZh'] or reslist[j]['NameZh']+reslist[j]['ModelZh']+reslist[j]['SpecificationZh'],
+                          'score': 0.5}
+                   info['label'].append(copy.deepcopy(aaa))
+        # 判断label0和label1是否为空，为空则：label默认2级
+        # 判断exp和label是否一致，一致则：rt=True，不一致则：rt=False
+        if info['exp'][0]['label'] == info['label'][0]['label'] and info['exp'][1]['label'] == info['label'][1][
+            'label'] and info['exp'][0]['score'] <= info['label'][0]['score'] and info['exp'][1]['score'] <= \
+                info['label'][1]['score']:
+            info['rt'] = True
+        else:
+            info['rt'] = False
+    else:
+        info['label'] = info['label'] = [{'label': '', 'score': 0}, {'label': '', 'score': 0}]
+        info['rt'] = False
+    return info
+
+
+def do_metricsevaluation_list(collections,data,score_threshold):
+    """
+    指标 评测
+    """
+    info = {
+        "input": {
+            "用户输入/userInput": "我要申请软件，名字叫：ai_xzh_all_restricted_software完全受限软件"
+        },
+        "output": {
+            "用户输入/output": "我要申请软件，名字叫：ai_xzh_all_restricted_software完全受限软件"
+        },
+        "rt": True,
+        "label": [{"label": "测试"}, {"label": "测试"}],
+        "exp": [{"label": "测试"}, {"label": "测试" }],
+        "artificial": []
+    }
+    info_list = []
+    businessscenario = []
+    for i in collections:
+        for j in data:
+            if i['content'] == j['用户输入/userInput']:
+                info = {
+                    "input": {},
+                    "output": {},
+                    "rt": False,
+                    "label": [{"label": "测试"}, {"label": "测试"}],
+                    "exp": [{"label": "测试",}, {"label": "测试"}],
+                    "artificial": []
+                }
+                info['input']['用户输入/userInput'] = i['ext']['output']
+                info['output']['用户输入/output'] = "对话id："+j['对话日志/intentID']+"    对话内容："+j['用户输入/userInput']
+                if i['ext']['BPO标注-AP-分发技能'] != '' and i['ext']['BPO标注-AP-分发技能'] in j['分发技能/skill'][0]:
+                    info['rt'] = True
+                    info['label'] = [{'label': j['分发技能/skill'][0]}]
+                    info['exp'] = [{'label': j['分发技能/skill'][0]}]
+                    info['artificial'] = info['exp']
+                else:
+                    info['rt'] = False
+                    info['label'] = [{'label': j['分发技能/skill'][0]}]
+                    info['exp'] = [{'label': i['ext']['BPO标注-AP-分发技能']}]
+                    info['artificial'] = info['exp']
+                if i['ext']['BPO标注-AP-分发技能'] not in businessscenario:
+                    businessscenario.append(i['ext']['BPO标注-AP-分发技能'])
+                #将data中的j删除
+                data.remove(j)
+                break
+        info_list.append(copy.deepcopy(info))
+    return info_list,businessscenario
+
+
+def do_scenereview_list(collections,data,score_threshold):
+    """
+    场景 评测  提取关键词
+    """
+    score_threshold=0.8
+    info = {
+        "input": {
+            "用户输入/userInput": "我要申请软件，名字叫：ai_xzh_all_restricted_software完全受限软件"
+        },
+        "output": {
+            "用户输入/output": "我要申请软件，名字叫：ai_xzh_all_restricted_software完全受限软件"
+        },
+        "rt": True,
+        "label": [{"label": "GUI 软件申请", "score": 0.6}, {"label": "软件申请", "score": 0.5}],
+        "exp": [{"label": "GUI 软件申请", " score": 0.9}, {"label": "软件申请", "score": 0.8 }],
+        "artificial": []
+    }
+    info_list = []
+    for i in collections:
+        for j in data:
+            if i['content'] == j['用户输入/userInput']:
+                info['input']['用户输入/userInput'] = i['ext']['output']
+                info['output']['用户输入/output'] = j['用户输入/userInput']
+                if i['ext']['BPO标注-AP-资产名称'] == j['llm关键词']:
+                    info['rt'] = True
+                    info['label'] = [{'label': str(j['llm关键词']), 'score': score_threshold}]
+                else:
+                    info['rt'] = False
+                    info['label'] = [{'label': str(i['ext']['BPO标注-AP-资产名称']), 'score': score_threshold}]
+                info['exp'] = [{'label': str(i['ext']['BPO标注-AP-资产名称']), 'score': score_threshold}]
+                info['artificial'] = info['exp']
+    info_list.append(copy.deepcopy(info))
+    return info_list
+
+def do_waterlevelline_autotest(collections, clientinfo, score_threshold):
+    """
+    水位线评测- 返回 符合报告模式的结果
+    """
+    keywprd= []
+    info_list = []
+    info = {
+        "input": {
+            "用户输入/userInput": "我要申请软件，名字叫：ai_xzh_all_restricted_software完全受限软件"
+        },
+        "output": {
+            "用户输入/output": "我要申请软件，名字叫：ai_xzh_all_restricted_software完全受限软件"
+        },
+        "rt": True,
+        "label": [{"label": "GUI 软件申请", "score": 0.6}, {"label": "软件申请", "score": 0.5}],
+        "exp": [{"label": "GUI 软件申请", " score": 0.9}, {"label": "软件申请", "score": 0.8
+                                                       }],
+        "artificial": []
+    }
+    a =0
+    for i in collections:
+        info['input']['用户输入/userInput'] = i['ext']['BPO标注-AP-资产名称']
+        info['output']['用户输入/output'] = i['ext']['BPO标注-AP-资产名称']
+        info['exp'] = []
+        for j in [i['ext']['BPO标注-AP-资产型号'], i['ext'].get('BPO标注-AP-资产型号1'),i['ext'].get('BPO标注-AP-资产型号2'),i['ext'].get('BPO标注-AP-资产型号3'),i['ext'].get('BPO标注-AP-资产型号4'),i['ext'].get('BPO标注-AP-资产型号5'),i['ext'].get('BPO标注-AP-资产型号6'),i['ext'].get('BPO标注-AP-资产型号7'),i['ext'].get('BPO标注-AP-资产型号8'),i['ext'].get('BPO标注-AP-资产型号9'),i['ext'].get('BPO标注-AP-资产型号10')]:
+            if j:
+                info['exp'].append({'label': j, 'score': score_threshold})
+        #判断i['ext']['BPO标注-AP-资产型号']是否为空，为空就不用读取，不为空就读取
+        if i['ext']['BPO标注-AP-资产型号']:
+            asset_name = i['ext']['BPO标注-AP-资产名称']
+            try:
+                if isinstance(asset_name, str):
+                    asset_name = json.loads(asset_name)
+            except json.JSONDecodeError:
+                # 若解析失败，说明不是 JSON 格式，保持原样
+                pass
+        info['artificial'] = info['exp']
+        if i['ext']['BPO标注-AP-资产名称']:
+            asset_name = i['ext']['BPO标注-AP-资产名称']
+            try:
+                if isinstance(asset_name, str):
+                    asset_name = json.loads(asset_name)
+            except json.JSONDecodeError:
+                # 若解析失败，说明不是 JSON 格式，保持原样
+                pass
+            if "软件申请" in i['ext']['BPO标注-AP-分发技能']:
+                keywprd = software_asset_sku_structure(asset_name)
+                res = json.loads(get_query_vector(keywprd, clientinfo))
+            if "设备/配件申请" in i['ext']['BPO标注-AP-分发技能']:
+                #keywprd = equipmentrequest_structure(asset_name, i['ext']['asset_type'])
+                keywprd = {
+    "From": 0,
+    "Size": 10,
+    "MinScore": 0.7,
+    "AssetModelFieldsWithAnd": [
+        {
+            "FieldName": "vec_search",
+            "FieldType": "knn",
+            "QueryValue": [
+                asset_name
+            ]
+        }
+    ],
+    "SPUIDs": None,
+    "AssetModelBizTypes": [
+        "asset_sku"
+    ]
+}
+                res = json.loads(get_query_vector(keywprd, clientinfo))
+                #res = get_by_AssetModelBizTypes(keywprd,res0)
+            if "设备/配件退还" in i['ext']['BPO标注-AP-分发技能']:
+                keywprd = equipmentreturn_structure0(asset_name, i['ext']['asset_type'])
+                res0 = json.loads(get_query_vector(keywprd, clientinfo))
+                res = get_by_AssetModelBizTypes(keywprd, res0)
+        else:
+            res = ""
+        infoout = do_waterlevellineres_listv2(res, info)
+        info_list.append(copy.deepcopy(infoout))
+        a = a+1
+        print("这是"+str(a))
+        bbb = a
+    return info_list
+
+def do_waterlevelline_autotest_aseetspu(collections, clientinfo, score_threshold):
+    """
+    水位线评测- 返回 符合报告模式的结果
+    """
+    keywprd= []
+    info_list = []
+    info = {
+        "input": {
+            "用户输入/userInput": "我要申请软件，名字叫：ai_xzh_all_restricted_software完全受限软件"
+        },
+        "output": {
+            "用户输入/output": "我要申请软件，名字叫：ai_xzh_all_restricted_software完全受限软件"
+        },
+        "rt": True,
+        "label": [{"label": "GUI 软件申请", "score": 0.6}, {"label": "软件申请", "score": 0.5}],
+        "exp": [{"label": "GUI 软件申请", " score": 0.9}, {"label": "软件申请", "score": 0.8
+                                                       }],
+        "artificial": []
+    }
+    a =0
+    for i in collections:
+        info['input']['用户输入/userInput'] = i['ext']['BPO标注-AP-资产名称']
+        info['output']['用户输入/output'] = i['ext']['BPO标注-AP-资产名称']
+        info['exp'] = []
+        for j in [i['ext']['BPO标注-AP-资产型号'], i['ext'].get('BPO标注-AP-资产型号1'),
+                  i['ext'].get('BPO标注-AP-资产型号2'), i['ext'].get('BPO标注-AP-资产型号3'),
+                  i['ext'].get('BPO标注-AP-资产型号4'), i['ext'].get('BPO标注-AP-资产型号5'),
+                  i['ext'].get('BPO标注-AP-资产型号6'), i['ext'].get('BPO标注-AP-资产型号7'),
+                  i['ext'].get('BPO标注-AP-资产型号8'), i['ext'].get('BPO标注-AP-资产型号9'),
+                  i['ext'].get('BPO标注-AP-资产型号10')]:
+            if j:
+                info['exp'].append({'label': j, 'score': score_threshold})
+        info['artificial'] = info['exp']
+        info['label']=[]
+        if i['ext']['BPO标注-AP-资产名称']:
+            if "设备" in i['ext']['BPO标注-AP-分发技能'] or "配件" in i['ext']['BPO标注-AP-分发技能']:
+                asset_name = i['ext']['BPO标注-AP-资产名称']
+                try:
+                    # 尝试将其解析为 JSON 对象
+                    if isinstance(asset_name, str):
+                        asset_name0 = json.loads(asset_name)
+                        asset_name = asset_name0['asset_name']
+                except json.JSONDecodeError:
+                    # 若解析失败，说明不是 JSON 格式，保持原样
+                    pass
+                key = asset_name.get('asset_name') if isinstance(asset_name, dict) else asset_name
+                if i['ext']['BPO标注-AP-资产类型'] == "设备":
+                    hardtype = 1
+                elif  i['ext']['BPO标注-AP-资产类型'] == "配件" or  i['ext']['BPO标注-AP-资产类型'] == '2.0' :
+                    hardtype = 2
+                res = GetBestMatchItemonline(key,hardtype,clientinfo)
+        else:
+            res = ""
+        infoout = do_waterlevellineres_listassetspu(res, info,hardtype)
+        info_list.append(copy.deepcopy(infoout))
+        a = a+1
+        print("这是"+str(a))
+        bbb = a
+    return info_list
+
+def do_waterlevelline_autotest_aseetspu_old(collections, clientinfo, score_threshold):
+    """
+    水位线评测- 返回 符合报告模式的结果
+    """
+    keywprd= []
+    info_list = []
+    info = {
+        "input": {
+            "用户输入/userInput": "我要申请软件，名字叫：ai_xzh_all_restricted_software完全受限软件"
+        },
+        "output": {
+            "用户输入/output": "我要申请软件，名字叫：ai_xzh_all_restricted_software完全受限软件"
+        },
+        "rt": True,
+        "label": [{"label": "GUI 软件申请", "score": 0.6}, {"label": "软件申请", "score": 0.5}],
+        "exp": [{"label": "GUI 软件申请", " score": 0.9}, {"label": "软件申请", "score": 0.8
+                                                       }],
+        "artificial": []
+    }
+    a =0
+    for i in collections:
+        info['input']['用户输入/userInput'] = i['ext']['BPO标注-AP-资产名称']
+        info['output']['用户输入/output'] = i['ext']['BPO标注-AP-资产名称']
+        info['exp'] = [
+            {'label': i['ext']['BPO标注-AP-资产型号'], 'score': score_threshold},
+            {'label': i['ext'].get('BPO标注-AP-资产型号1', ''), 'score': score_threshold}]
+        info['artificial'] = info['exp']
+        info['label']=[]
+        if i['ext']['BPO标注-AP-资产名称']:
+            if "设备" in i['ext']['BPO标注-AP-分发技能'] or "配件" in i['ext']['BPO标注-AP-分发技能']:
+                asset_name = i['ext']['BPO标注-AP-资产名称']
+                asset_name = i['ext']['BPO标注-AP-资产类型']
+                try:
+                    # 尝试将其解析为 JSON 对象
+                    if isinstance(asset_name, str):
+                        asset_name = json.loads(asset_name)
+                except json.JSONDecodeError:
+                    # 若解析失败，说明不是 JSON 格式，保持原样
+                    pass
+                key = asset_name.get('asset_name') if isinstance(asset_name, dict) else asset_name
+                if i['ext']['BPO标注-AP-资产型号'] == "设备":
+                    hardtype = 1
+                elif  i['ext']['BPO标注-AP-资产型号'] == "配件" :
+                    hardtype = 2
+                res = GetBestMatchItemonline_old(key,hardtype,clientinfo)
+        else:
+            res = ""
+        infoout = do_waterlevellineres_listassetspu(res, info,hardtype)
+        info_list.append(copy.deepcopy(infoout))
+        a = a+1
+        print("这是"+str(a))
+        bbb = a
+    return info_list
+
+
+def do_waterlevelline_autotest_aseetspu_pre(collections, clientinfo, score_threshold):
+    """
+    水位线评测- 返回 符合报告模式的结果
+    """
+    keywprd= []
+    info_list = []
+    info = {
+        "input": {
+            "用户输入/userInput": "我要申请软件，名字叫：ai_xzh_all_restricted_software完全受限软件"
+        },
+        "output": {
+            "用户输入/output": "我要申请软件，名字叫：ai_xzh_all_restricted_software完全受限软件"
+        },
+        "rt": True,
+        "label": [{"label": "GUI 软件申请", "score": 0.6}, {"label": "软件申请", "score": 0.5}],
+        "exp": [{"label": "GUI 软件申请", " score": 0.9}, {"label": "软件申请", "score": 0.8
+                                                       }],
+        "artificial": []
+    }
+    a =0
+    for i in collections:
+        info['input']['用户输入/userInput'] = i['ext']['BPO标注-AP-资产名称']
+        info['output']['用户输入/output'] = i['ext']['BPO标注-AP-资产名称']
+        info['exp'] = [
+            {'label': i['ext']['BPO标注-AP-资产型号'], 'score': score_threshold},
+            {'label': i['ext'].get('BPO标注-AP-资产型号1', ''), 'score': score_threshold}]
+        info['artificial'] = info['exp']
+        info['label']=[]
+        if i['ext']['BPO标注-AP-资产名称']:
+            if "设备" in i['ext']['BPO标注-AP-分发技能'] or "配件" in i['ext']['BPO标注-AP-分发技能']:
+                asset_name = i['ext']['BPO标注-AP-资产名称']
+                asset_name = i['ext']['BPO标注-AP-资产类型']
+                try:
+                    # 尝试将其解析为 JSON 对象
+                    if isinstance(asset_name, str):
+                        asset_name = json.loads(asset_name)
+                except json.JSONDecodeError:
+                    # 若解析失败，说明不是 JSON 格式，保持原样
+                    pass
+                key = asset_name.get('asset_name') if isinstance(asset_name, dict) else asset_name
+                if i['ext']['BPO标注-AP-资产型号'] == "设备":
+                    hardtype = 1
+                elif  i['ext']['BPO标注-AP-资产型号'] == "配件" :
+                    hardtype = 2
+                res = test_hardware_match(key,hardtype,clientinfo)
+        else:
+            res = ""
+        infoout = do_waterlevellineres_listassetspu_pre(res, info,hardtype)
+        info_list.append(copy.deepcopy(infoout))
+        a = a+1
+        print("这是"+str(a))
+        bbb = a
+    return info_list
+
+def do_waterlevelline_autotest_search(collections, clientinfo, score_threshold):
+    """
+    水位线评测- 返回 符合报告模式的结果
+    """
+    keywprd= []
+    info_list = []
+    info = {
+        "input": {
+            "用户输入/userInput": "我要申请软件，名字叫：ai_xzh_all_restricted_software完全受限软件"
+        },
+        "output": {
+            "用户输入/output": "我要申请软件，名字叫：ai_xzh_all_restricted_software完全受限软件"
+        },
+        "rt": True,
+        "label": [{"label": "GUI 软件申请", "score": 0.6}, {"label": "软件申请", "score": 0.5}],
+        "exp": [{"label": "GUI 软件申请", " score": 0.9}, {"label": "软件申请", "score": 0.8
+                                                       }],
+        "artificial": []
+    }
+    a =0
+    for i in collections:
+        info['input']['用户输入/userInput'] = i['ext']['BPO标注-AP-资产名称']
+        info['output']['用户输入/output'] = i['ext']['BPO标注-AP-资产名称']
+        info['exp'] = [
+            {'label': i['ext']['BPO标注-AP-资产型号'], 'score': score_threshold},
+            {'label': i['ext'].get('BPO标注-AP-资产型号1', ''), 'score': score_threshold}]
+        info['artificial'] = info['exp']
+        info['label']=[]
+        if i['ext']['BPO标注-AP-资产名称']:
+            if "设备" in i['ext']['BPO标注-AP-分发技能'] or "配件" in i['ext']['BPO标注-AP-分发技能']:
+                asset_name = i['ext']['BPO标注-AP-资产名称']
+                asset_name = i['ext']['BPO标注-AP-资产类型']
+                try:
+                    # 尝试将其解析为 JSON 对象
+                    if isinstance(asset_name, str):
+                        asset_name = json.loads(asset_name)
+                except json.JSONDecodeError:
+                    # 若解析失败，说明不是 JSON 格式，保持原样
+                    pass
+                key = asset_name.get('asset_name') if isinstance(asset_name, dict) else asset_name
+                if i['ext']['BPO标注-AP-资产型号'] == "设备":
+                    hardtype = 1
+                elif  i['ext']['BPO标注-AP-资产型号'] == "配件" :
+                    hardtype = 2
+                res = searchListAssetModelScope(key,hardtype,clientinfo)
+        else:
+            res = ""
+        infoout = do_waterlevellineres_sr(res, info,hardtype)
+        info_list.append(copy.deepcopy(infoout))
+        a = a+1
+        print("这是"+str(a))
+        bbb = a
+    return info_list
+
+
+def do_waterlevelline_autotest_SoftwareApplyRecommendList(collections, clientinfo, score_threshold):
+    """
+    水位线评测- 返回 符合报告模式的结果
+    """
+    keywprd= []
+    info_list = []
+    info = {
+        "input": {
+            "用户输入/userInput": "我要申请软件，名字叫：ai_xzh_all_restricted_software完全受限软件"
+        },
+        "output": {
+            "用户输入/output": "我要申请软件，名字叫：ai_xzh_all_restricted_software完全受限软件"
+        },
+        "rt": True,
+        "label": [{"label": "GUI 软件申请", "score": 0.6}, {"label": "软件申请", "score": 0.5}],
+        "exp": [{"label": "GUI 软件申请", " score": 0.9}, {"label": "软件申请", "score": 0.8
+                                                       }],
+        "artificial": []
+    }
+    a =0
+    for i in collections:
+        info['input']['用户输入/userInput'] = i['ext'].get('BP标注-AP-资产名称') or i['ext'].get('BPO标注-AP-资产名称')
+        info['output']['用户输入/output'] = i['ext']['BP标注-AP-资产名称']
+        info['exp'] = [
+            {'label': i['ext']['BP标注-AP-资产名称'], 'score': score_threshold},
+            {'label': i['ext'].get('BP标注-AP-资产型号1', ''), 'score': score_threshold}]
+        info['artificial'] = info['exp']
+        info['label']=[]
+        if i['ext']['BP标注-AP-资产名称']:
+            if "软件" in i['ext']['BP标注-AP-分发技能']:
+                asset_name = i['ext']['BP标注-AP-资产名称']
+                try:
+                    # 尝试将其解析为 JSON 对象
+                    if isinstance(asset_name, str):
+                        asset_name = json.loads(asset_name)
+                except json.JSONDecodeError:
+                    # 若解析失败，说明不是 JSON 格式，保持原样
+                    pass
+                key = asset_name.get('asset_name') if isinstance(asset_name, dict) else asset_name
+                res = SoftwareApplyGetBestMatchItem(key,clientinfo)
+        else:
+            res = ""
+        infoout = do_waterlevellineres_software(res, info)
+        info_list.append(copy.deepcopy(infoout))
+        a = a+1
+        print("这是"+str(a))
+        bbb = a
+    return info_list
+
+
+def do_waterlevellineres_software(res, info):
+    """
+    获取结果，并组装水位线info
+    """
+    if res == '':
+        info['label'] = [{'label': '', 'score': 0}]
+        info['rt'] = False
+        return info
+    if res["data"].get("SoftwareApplyRecommendList") == None:
+        info['label'] = [{'label': '', 'score': 0},{'label': '', 'score': 0.8}]
+        info['rt'] = False
+        if info['exp'][0]['label']=='':
+            info['rt'] = True
+        return info
+    reslist=res["data"]["SoftwareApplyRecommendList"]
+    aaa ={}
+    # 判断res["body"]["Results"]不为空，空则：label0 label1 为空，label默认2级
+    if reslist:
+        # 取["Results"]下前2个结果，若只有1个结果，label1为空
+        if len(reslist) > 0:
+           for j in range(len(reslist)):
+               aaa = {'label': reslist[j]['Data']['Name']['ValueZh'],
+                      'score': reslist[j]['Score']}
+               info['label'].append(copy.deepcopy(aaa))
+        if info['exp'][0]['label'] == info['label'][0]['label'] and info['exp'][0]['score'] <= info['label'][0]['score']:
+            info['rt'] = True
+        else:
+            info['rt'] = False
+    else:
+        info['label'] = info['label'] = [{'label': '', 'score': 0}, {'label': '', 'score': 0}]
+        info['rt'] = False
+    return info
+
+def get_conversationlogs1(startAt):
+    """
+    对话ID 技能分发 用户输入
+    res_data = {
+            'intentID': 7485259579248705537,
+            'skillLabels': ["GUI 设备/配件申请"],
+            'userInput': "我要申请一个鼠标",
+
+         }
+         """
+    data = webapiClient().get_intent_detail_list(startAt)
+
+
+def get_conversationlogs(startAt, pageSize=10):
+    """
+    对话ID 技能分发 用户输入
+    res_data = {
+            'intentID': 7485259579248705537,
+            'skillLabels': ["GUI 设备/配件申请"],
+            'userInput': "我要申请一个鼠标",
+
+         }
+    """
+    try:
+        # 之前提到形参 'pageSize' 未填，这里假设默认值为 10，你可按需修改
+        data = webapiClient().get_intent_detail_list(startAt, pageSize=10)
+        return data
+    except KeyError as ke:
+        print(f"KeyError 发生: 数据中缺少必要的键，错误详情: {ke}")
+        return None
+    except IndexError as ie:
+        print(f"IndexError 发生: 索引超出范围，错误详情: {ie}")
+        return None
+    except Exception as e:
+        print(f"发生未知错误: {e}")
+        return None
+
+
+def write_reslut(data, Testsuitelink, title):
+    """
+    写入表格
+    """
+    try:
+        # 解析 spreadsheet_token
+        spreadsheet_token = Testsuitelink.split("/")[-1]
+
+        # 生成新工作表名称
+        new_sheet_title = f"{title}{datetime.datetime.now().strftime('%Y_%m_%d_%H_%M_%S')}"
+        sheetinfo = {"index": 0, "title": new_sheet_title}
+
+        # 创建新工作表
+        spreadsheet0 = LarkdocsClient().createsheets(spreadsheet_token, sheetinfo)
+        sheet_id = spreadsheet0['sheet_id']
+
+        # 准备表头数据
+        headers = list(data[0].keys())
+        header_data = [
+            {
+                "range": f"{sheet_id}!{chr(ord('A') + col)}1:{chr(ord('A') + col)}1",
+                "values": [[[{"text": {"text": header}, "type": "text"}]]]
+            }
+            for col, header in enumerate(headers)
+        ]
+
+        # 写入表头
+        LarkdocsClient().writesheets(spreadsheet_token, sheet_id, {"value_ranges": header_data})
+
+        # 写入数据
+        for row, row_data in enumerate(data, start=1):
+            row_values = [
+                {
+                    "range": f"{sheet_id}!{chr(ord('A') + col)}{row + 1}:{chr(ord('A') + col)}{row + 1}",
+                    "values": [[[{"text": {"text": str(row_data[header])}, "type": "text"}]]]
+                }
+                for col, header in enumerate(headers)
+            ]
+            LarkdocsClient().writesheets(spreadsheet_token, sheet_id, {"value_ranges": row_values})
+
+        return True
+    except KeyError as ke:
+        print(f"KeyError 发生: 数据中缺少必要的键，错误详情: {ke}")
+        return False
+    except IndexError as ie:
+        print(f"IndexError 发生: 索引超出范围，错误详情: {ie}")
+        return False
+    except Exception as e:
+        print(f"发生未知错误: {e}")
+        return False
+
+
+def write_excletolist(data_name):
+    """
+    1. 读取本地表格
+    2. 将表格内容拼接为text
+    """
+    try:
+        # 查看当前工作目录
+        print(f"当前工作目录: {os.getcwd()}")
+        # /Users/bytedance/itam_assistant/itam_assistant/accessory.csv
+        # 构建文件路径
+        file_path = f'data/{data_name}.csv'
+        Candidates = []
+        Candidate = {
+            "Score": 0,
+            "Text": "IOS手机",
+            "Attrs": {"id": "", "type": ""}}
+        text = ""
+        with open(file_path, 'r', encoding='utf-8', newline='') as file:
+            reader = csv.reader(file)
+            headers = next(reader)  # 读取表头
+            for header in headers:
+                text += f"{header}: "
+            text = text.rstrip(': ') + '\n'
+
+            for row in reader:
+                textout = ""
+                textout += ', '.join(row)
+                Candidate['Text'] = textout
+                Candidates.append(copy.deepcopy(Candidate))
+        return Candidates
+    except FileNotFoundError:
+        print(f"未找到文件: {file_path}")
+        return None
+    except Exception as e:
+        print(f"发生未知错误: {e}")
+        return None
+
+
+