diff --git a/pilot/openapi/api_v1/api_v1.py b/pilot/openapi/api_v1/api_v1.py index 23cba7948..f552b8dc0 100644 --- a/pilot/openapi/api_v1/api_v1.py +++ b/pilot/openapi/api_v1/api_v1.py @@ -229,7 +229,6 @@ async def dialogue_scenes(): new_modes: List[ChatScene] = [ ChatScene.ChatWithDbExecute, ChatScene.ChatExcel, - ChatScene.ChatWithDbQA, ChatScene.ChatKnowledge, ChatScene.ChatDashboard, ChatScene.ChatAgent, diff --git a/pilot/out_parser/base.py b/pilot/out_parser/base.py index 0dfe82d57..ca7caab31 100644 --- a/pilot/out_parser/base.py +++ b/pilot/out_parser/base.py @@ -215,7 +215,7 @@ class BaseOutputParser(ABC): cleaned_output = self.__illegal_json_ends(cleaned_output) return cleaned_output - def parse_view_response(self, ai_text, data) -> str: + def parse_view_response(self, ai_text, data, parse_prompt_response:Any=None) -> str: """ parse the ai response info to user view Args: diff --git a/pilot/scene/base.py b/pilot/scene/base.py index b56176991..3c1312996 100644 --- a/pilot/scene/base.py +++ b/pilot/scene/base.py @@ -25,8 +25,8 @@ class Scene: class ChatScene(Enum): ChatWithDbExecute = Scene( code="chat_with_db_execute", - name="Chat Data", - describe="Dialogue with your private data through natural language.", + name="Chat DB", + describe="Dialogue with your private databse data through natural language.", param_types=["DB Select"], ) ExcelLearning = Scene( diff --git a/pilot/scene/base_chat.py b/pilot/scene/base_chat.py index d22a0a85f..e23446bdc 100644 --- a/pilot/scene/base_chat.py +++ b/pilot/scene/base_chat.py @@ -227,7 +227,6 @@ class BaseChat(ABC): ) ) ### run - # result = self.do_action(prompt_define_response) result = await blocking_func_to_async( self._executor, self.do_action, prompt_define_response ) @@ -243,6 +242,7 @@ class BaseChat(ABC): self.prompt_template.output_parser.parse_view_response, speak_to_user, result, + prompt_define_response ) view_message = view_message.replace("\n", "\\n") diff --git a/pilot/scene/chat_data/chat_excel/excel_analyze/chat.py b/pilot/scene/chat_data/chat_excel/excel_analyze/chat.py index ff12ec803..410c5dbbd 100644 --- a/pilot/scene/chat_data/chat_excel/excel_analyze/chat.py +++ b/pilot/scene/chat_data/chat_excel/excel_analyze/chat.py @@ -50,17 +50,28 @@ class ChatExcel(BaseChat): super().__init__(chat_param=chat_param) def _generate_numbered_list(self) -> str: - command_strings = [] - if CFG.command_disply: - for name, item in CFG.command_disply.commands.items(): - if item.enabled: - command_strings.append(f"{name}:{item.description}") + antv_charts = [{"line_chart":"used to display comparative trend analysis data"}, + {"pie_chart":"suitable for scenarios such as proportion and distribution statistics"}, + {"response_table":"suitable for display with many display columns or non-numeric columns"}, + {"data_text":" the default display method, suitable for single-line or simple content display"}, + {"scatter_plot":"Suitable for exploring relationships between variables, detecting outliers, etc."}, + {"bubble_chart":"Suitable for relationships between multiple variables, highlighting outliers or special situations, etc."}, + {"donut_chart":"Suitable for hierarchical structure representation, category proportion display and highlighting key categories, etc."}, + {"area_chart":"Suitable for visualization of time series data, comparison of multiple groups of data, analysis of data change trends, etc."}, + {"heatmap":"Suitable for visual analysis of time series data, large-scale data sets, distribution of classified data, etc."} + ] + + # command_strings = [] + # if CFG.command_disply: + # for name, item in CFG.command_disply.commands.items(): + # if item.enabled: + # command_strings.append(f"{name}:{item.description}") # command_strings += [ # str(item) # for item in CFG.command_disply.commands.values() # if item.enabled # ] - return "\n".join(f"{i+1}. {item}" for i, item in enumerate(command_strings)) + return "\n".join(f"{key}:{value}" for dict_item in antv_charts for key, value in dict_item.items()) async def generate_input_values(self) -> Dict: input_values = { diff --git a/pilot/scene/chat_data/chat_excel/excel_analyze/prompt.py b/pilot/scene/chat_data/chat_excel/excel_analyze/prompt.py index 140c098d3..75fd26420 100644 --- a/pilot/scene/chat_data/chat_excel/excel_analyze/prompt.py +++ b/pilot/scene/chat_data/chat_excel/excel_analyze/prompt.py @@ -22,7 +22,7 @@ Constraint: 5.The part of the required output format needs to be parsed by the code. Please ensure that this part of the content is output as required. Please respond in the following format: - Summary of your analytical thinking.[Data display method][Correct duckdb data analysis sql] + thoughts summary to say to user.[Data display method][Correct duckdb data analysis sql] User Questions: {user_input} @@ -38,7 +38,7 @@ _DEFAULT_TEMPLATE_ZH = """ 4.优先使用数据分析的方式回答，如果用户问题不涉及数据分析内容，你可以按你的理解进行回答 5.要求的输出格式中部分需要被代码解析执行，请确保这部分内容按要求输出 请确保你的输出格式如下: - 分析思路总结.[数据展示方式][正确的duckdb数据分析sql] + 对用户说的想法摘要.[数据展示方式][正确的duckdb数据分析sql] 用户问题：{user_input} """ diff --git a/pilot/scene/chat_data/chat_excel/excel_reader.py b/pilot/scene/chat_data/chat_excel/excel_reader.py index 175b3bf4e..2bc882527 100644 --- a/pilot/scene/chat_data/chat_excel/excel_reader.py +++ b/pilot/scene/chat_data/chat_excel/excel_reader.py @@ -247,17 +247,12 @@ if __name__ == "__main__": # sql = add_quotes_to_chinese_columns(sql_3) # print(f"excute sql:{sql}") - data = { - 'name': ['John', 'Alice', 'Bob'], - 'age': [30, 25, 35], - 'timestamp': [pd.Timestamp('2022-01-01'), pd.Timestamp('2022-01-02'), pd.Timestamp('2022-01-03')], - 'category': pd.Categorical(['A', 'B', 'C']) - } + my_list = [{'name': 'John', 'age': 30}, {'name': 'Alice', 'age': 25}, {'name': 'Bob', 'age': 35}] - df = pd.DataFrame(data) + for dict_item in my_list: + for key, value in dict_item.items(): + print(key, value) - json_data = df.to_json(orient='records', date_format='iso', date_unit='s') - print(json_data) class ExcelReader: def __init__(self, file_path): diff --git a/pilot/scene/chat_db/auto_execute/chat.py b/pilot/scene/chat_db/auto_execute/chat.py index 6fd3c5d7e..7fa74ffa7 100644 --- a/pilot/scene/chat_db/auto_execute/chat.py +++ b/pilot/scene/chat_db/auto_execute/chat.py @@ -37,7 +37,7 @@ class ChatWithDbAutoExecute(BaseChat): ) self.database = CFG.LOCAL_DB_MANAGE.get_connect(self.db_name) - self.top_k: int = 200 + self.top_k: int = 50 self.api_call = ApiCall(display_registry=CFG.command_disply) async def generate_input_values(self): @@ -64,7 +64,7 @@ class ChatWithDbAutoExecute(BaseChat): # table_infos = self.database.table_simple_info() input_values = { - "input": self.current_user_input, + # "input": self.current_user_input, "top_k": str(self.top_k), "dialect": self.database.dialect, "table_info": table_infos, @@ -79,4 +79,4 @@ class ChatWithDbAutoExecute(BaseChat): def do_action(self, prompt_response): print(f"do_action:{prompt_response}") - return self.database.run(prompt_response.sql) + return self.database.run_to_df diff --git a/pilot/scene/chat_db/auto_execute/out_parser.py b/pilot/scene/chat_db/auto_execute/out_parser.py index 577cac1ef..382bda3da 100644 --- a/pilot/scene/chat_db/auto_execute/out_parser.py +++ b/pilot/scene/chat_db/auto_execute/out_parser.py @@ -1,6 +1,8 @@ import json from typing import Dict, NamedTuple import logging +import xml.etree.ElementTree as ET +from pilot.common.json_utils import serialize from pilot.out_parser.base import BaseOutputParser, T from pilot.configs.config import Config from pilot.scene.chat_db.data_loader import DbDataLoader @@ -31,21 +33,28 @@ class DbChatOutputParser(BaseOutputParser): thoughts = response[key] return SqlAction(sql, thoughts) - def parse_view_response(self, speak, data) -> str: - import pandas as pd + def parse_view_response(self, speak, data, prompt_response) -> str: + + param = {} + api_call_element = ET.Element("chart-view") + try: + df = data(prompt_response.sql) + param["type"] = "response_table" + param["sql"] = prompt_response.sql + param["data"] = json.loads(df.to_json(orient='records', date_format='iso', date_unit='s')) + view_json_str = json.dumps(param, default=serialize) + except Exception as e: + err_param ={} + param["sql"] = prompt_response.sql + err_param["type"] = "response_table" + err_param["err_msg"] = str(e) + view_json_str = json.dumps(err_param, default=serialize) + + api_call_element.text = view_json_str + result = ET.tostring(api_call_element, encoding="utf-8") + + return speak + "\n" + result.decode("utf-8") + + + - ### tool out data to table view - data_loader = DbDataLoader() - if len(data) < 1: - data.insert(0, []) - df = pd.DataFrame(data[1:], columns=data[0]) - if not CFG.NEW_SERVER_MODE and not CFG.SERVER_LIGHT_MODE: - table_style = """""" - html_table = df.to_html(index=False, escape=False) - html = f"{table_style}{html_table}" - view_text = f"##### {str(speak)}" + "\n" + html.replace("\n", " ") - return view_text - else: - return data_loader.get_table_view_by_conn(data, speak) diff --git a/pilot/scene/chat_db/auto_execute/prompt.py b/pilot/scene/chat_db/auto_execute/prompt.py index 585146fbb..a77bfd2da 100644 --- a/pilot/scene/chat_db/auto_execute/prompt.py +++ b/pilot/scene/chat_db/auto_execute/prompt.py @@ -13,36 +13,35 @@ _PROMPT_SCENE_DEFINE_EN = "You are a database expert. " _PROMPT_SCENE_DEFINE_ZH = "你是一个数据库专家. " _DEFAULT_TEMPLATE_EN = """ -Given an input question, create a syntactically correct {dialect} sql. -Table structure information: - {table_info} -Constraint: -1. You can only use the table provided in the table structure information to generate sql. If you cannot generate sql based on the provided table structure, please say: "The table structure information provided is not enough to generate sql query." It is prohibited to fabricate information at will. -2. Do not query columns that do not exist. Pay attention to which column is in which table. -3. Unless the user specifies in the question a specific number of examples he wishes to obtain, always limit the query to a maximum of {top_k} results. -4. Please ensure that the output result contains: response_tableSQL Query to run，and replace the generated sql into the parameter sql. +Please create a syntactically correct {dialect} sql based on the user question, use the following tables schema to generate sql: + {table_info} -Please make sure to respond as following format: - thoughts summary to say to user.response_tableSQL Query to run +Constraint: + 1.Unless the user specifies in his question a specific number of examples he wishes to obtain, always limit your query to at most {top_k} results. + 2.Please do not use columns that do not appear in the tables schema. Also be careful not to misunderstand the relationship between fields and tables in SQL. + 3.Use as few tables as possible when querying. + 4.Please check the correctness of the SQL and ensure that the query performance is optimized under correct conditions. -Question: {input} +Please think step by step and respond according to the following JSON format: + {response} +Ensure the response is correct json and can be parsed by Python json.loads. + """ _DEFAULT_TEMPLATE_ZH = """ -给定一个输入问题，创建一个语法正确的 {dialect} sql。 -已知表结构信息: +请根据用户输入问题，使用如下的表结构定义创建一个语法正确的 {dialect} sql: {table_info} 约束: -1. 只能使用表结构信息中提供的表来生成 sql，如果无法根据提供的表结构中生成 sql ，请说：“提供的表结构信息不足以生成 sql 查询。” 禁止随意捏造信息。 -2. 不要查询不存在的列，注意哪一列位于哪张表中。 -3. 请确保输出结果包含response_table要运行的SQL， 并将对应的sql替换到sql参数中 -4. 除非用户在问题中指定了他希望获得的具体示例数量，否则始终将查询限制为最多 {top_k} 个结果。 + 1. 除非用户在问题中指定了他希望获得的具体数据行数，否则始终将查询限制为最多 {top_k} 个结果。 + 2. 只能使用表结构信息中提供的表来生成 sql，如果无法根据提供的表结构中生成 sql ，请说：“提供的表结构信息不足以生成 sql 查询。” 禁止随意捏造信息。 + 3. 请注意生成SQL时不要弄错表和列的关系 + 4. 请检查SQL的正确性，并保证正确的情况下优化查询性能 + +请一步步思考并按照以下JSON格式回复： + {response} +确保返回正确的json并且可以被Python json.loads方法解析. -请务必按照以下格式回复： - 对用户说的想法摘要。response_table要运行的 SQL - -问题:{input} """ _DEFAULT_TEMPLATE = ( @@ -60,7 +59,7 @@ RESPONSE_FORMAT_SIMPLE = { PROMPT_SEP = SeparatorStyle.SINGLE.value -PROMPT_NEED_NEED_STREAM_OUT = True +PROMPT_NEED_NEED_STREAM_OUT = False # Temperature is a configuration hyperparameter that controls the randomness of language model output. # A high temperature produces more unpredictable and creative results, while a low temperature produces more common and conservative output. @@ -69,8 +68,8 @@ PROMPT_TEMPERATURE = 0.5 prompt = PromptTemplate( template_scene=ChatScene.ChatWithDbExecute.value(), - input_variables=["input", "table_info", "dialect", "top_k", "response"], - # response_format=json.dumps(RESPONSE_FORMAT_SIMPLE, ensure_ascii=False, indent=4), + input_variables=["table_info", "dialect", "top_k", "response"], + response_format=json.dumps(RESPONSE_FORMAT_SIMPLE, ensure_ascii=False, indent=4), template_define=PROMPT_SCENE_DEFINE, template=_DEFAULT_TEMPLATE, stream_out=PROMPT_NEED_NEED_STREAM_OUT, @@ -79,6 +78,7 @@ prompt = PromptTemplate( ), # example_selector=sql_data_example, temperature=PROMPT_TEMPERATURE, + need_historical_messages=True ) CFG.prompt_template_registry.register(prompt, is_default=True) from . import prompt_baichuan