add AWS_REDSHIFT_DIALECT_PROMPT_CLAUDE3

application/.env.cntemplate

@@ -18,6 +18,8 @@ BEDROCK_REGION=cn-north-1
 RDS_REGION_NAME=cn-north-1
 AWS_DEFAULT_REGION=cn-north-1
 
+DYNAMODB_AWS_REGION=cn-north-1
+
 SAGEMAKER_ENDPOINT_EMBEDDING=embedding-bge-m3-3ab71
 SAGEMAKER_ENDPOINT_INTENT=llm-internlm2-chat-7b-3ab71
 SAGEMAKER_ENDPOINT_SQL=sql-sqlcoder-7b-2-7e5b6

application/nlq/business/connection.py

@@ -4,6 +4,7 @@
 
 logger = logging.getLogger(__name__)
 
+
 class ConnectionManagement:
     connection_config_dao = ConnectConfigDao()
 
@@ -23,7 +24,8 @@ def get_conn_config_by_name(cls, conn_name):
 
     @classmethod
     def update_connection(cls, conn_name, db_type, db_host, db_port, db_user, db_pwd, db_name, comment):
-        cls.connection_config_dao.update_db_info(conn_name, db_type, db_host, db_port, db_user, db_pwd, db_name, comment)
+        cls.connection_config_dao.update_db_info(conn_name, db_type, db_host, db_port, db_user, db_pwd, db_name,
+                                                 comment)
         logger.info(f"Connection {conn_name} updated")
 
     @classmethod
@@ -48,4 +50,9 @@ def get_table_definition_by_config(cls, conn_config: ConnectConfigEntity, schema
     @classmethod
     def get_db_url_by_name(cls, conn_name):
         conn_config = cls.get_conn_config_by_name(conn_name)
-        return RelationDatabase.get_db_url_by_connection(conn_config)
+        return RelationDatabase.get_db_url_by_connection(conn_config)
+
+    @classmethod
+    def get_db_type_by_name(cls, conn_name):
+        conn_config = cls.get_conn_config_by_name(conn_name)
+        return conn_config.db_type

application/nlq/data_access/dynamo_connection.py

@@ -11,7 +11,7 @@
 
 # DynamoDB table name
 CONNECT_CONFIG_TABLE_NAME = 'NlqConnectConfig'
-DYNAMODB_AWS_REGION = os.environ.get('DYNAMODB_AWS_REGION', 'us-west-2')
+DYNAMODB_AWS_REGION = os.environ.get('DYNAMODB_AWS_REGION')
 
 class ConnectConfigEntity:
     """Connect config entity mapped to DynamoDB item"""

application/pages/1_🌍_Natural_Language_Querying.py

@@ -356,6 +356,7 @@ def main():
                             conn_name = database_profile['conn_name']
                             db_url = ConnectionManagement.get_db_url_by_name(conn_name)
                             database_profile['db_url'] = db_url
+                            database_profile['db_type'] = ConnectionManagement.get_db_type_by_name(conn_name)
 
                         if intent_ner_recognition:
                             intent_response = get_query_intent(model_type, search_box)
@@ -384,8 +385,7 @@ def main():
                                                                           model_id=model_type,
                                                                           sql_examples=retrieve_result,
                                                                           ner_example=entity_slot_retrieve,
-                                                                          dialect=get_db_url_dialect(
-                                                                              database_profile['db_url']),
+                                                                          dialect=database_profile['db_type'],
                                                                           model_provider=model_provider)
                                         sql_str = get_response_sql(each_task_sql_query)
                                         each_res_dict["sql"] = sql_str
@@ -440,7 +440,7 @@ def main():
                                                    model_id=model_type,
                                                    sql_examples=retrieve_result,
                                                    ner_example=entity_slot_retrieve,
-                                                   dialect=get_db_url_dialect(database_profile['db_url']),
+                                                   dialect=database_profile['db_type'],
                                                    model_provider=model_provider)
 
                             logger.info(f'got llm response: {response}')

application/utils/llm.py

@@ -7,7 +7,7 @@
 from utils import opensearch
 from utils.prompt import POSTGRES_DIALECT_PROMPT_CLAUDE3, MYSQL_DIALECT_PROMPT_CLAUDE3, \
     DEFAULT_DIALECT_PROMPT, SEARCH_INTENT_PROMPT_CLAUDE3, CLAUDE3_DATA_ANALYSE_SYSTEM_PROMPT, \
-    CLAUDE3_DATA_ANALYSE_USER_PROMPT
+    CLAUDE3_DATA_ANALYSE_USER_PROMPT, AWS_REDSHIFT_DIALECT_PROMPT_CLAUDE3
 import os
 import logging
 from langchain_core.output_parsers import JsonOutputParser
@@ -182,9 +182,7 @@ def generate_prompt(ddl, hints, search_box, sql_examples=None, ner_example=None,
     elif dialect == 'mysql':
         dialect_prompt = MYSQL_DIALECT_PROMPT_CLAUDE3
     elif dialect == 'redshift':
-        dialect_prompt = '''You are a Amazon Redshift expert. Given an input question, first create a syntactically 
-        correct Redshift query to run, then look at the results of the query and return the answer to the input 
-        question.'''
+        dialect_prompt = AWS_REDSHIFT_DIALECT_PROMPT_CLAUDE3
     else:
         dialect_prompt = DEFAULT_DIALECT_PROMPT
 

application/utils/prompt.py

@@ -16,6 +16,11 @@
 Pay attention to use CURDATE() function to get the current date, if the question involves "today". Aside from giving the SQL answer, concisely explain yourself after giving the answer
 in the same language as the question.""".format(top_k=TOP_K)
 
+
+AWS_REDSHIFT_DIALECT_PROMPT_CLAUDE3 = """You are a Amazon Redshift expert. Given an input question, first create a syntactically correct Redshift query to run, then look at the results of the query and return the answer to the input 
+question.When generating SQL, do not add double quotes or single quotes around table names. Unless the user specifies in the question a specific number of examples to obtain, query for at most {top_k} results using the LIMIT clause as per MySQL. 
+Never query for all columns from a table.""".format(top_k=TOP_K)
+
 SEARCH_INTENT_PROMPT_CLAUDE3 = """You are an intent classifier and entity extractor, and you need to perform intent classification and entity extraction on search queries.
 Background: I want to query data in the database, and you need to help me determine the user's relevant intent and extract the keywords from the query statement. Finally, return a JSON structure.
 

application/utils/prompts/generate_prompt.py

@@ -1,5 +1,5 @@
 from utils.prompt import POSTGRES_DIALECT_PROMPT_CLAUDE3, MYSQL_DIALECT_PROMPT_CLAUDE3, \
-    DEFAULT_DIALECT_PROMPT, AGENT_COT_SYSTEM_PROMPT, AGENT_COT_EXAMPLE
+    DEFAULT_DIALECT_PROMPT, AGENT_COT_SYSTEM_PROMPT, AGENT_COT_EXAMPLE, AWS_REDSHIFT_DIALECT_PROMPT_CLAUDE3
 from utils.prompts import guidance_prompt
 from utils.prompts import table_prompt
 import logging
@@ -285,9 +285,7 @@ def generate_llm_prompt(ddl, hints, search_box, sql_examples=None, ner_example=N
     elif dialect == 'mysql':
         dialect_prompt = MYSQL_DIALECT_PROMPT_CLAUDE3
     elif dialect == 'redshift':
-        dialect_prompt = '''You are a Amazon Redshift expert. Given an input question, first create a syntactically 
-        correct Redshift query to run, then look at the results of the query and return the answer to the input 
-        question. query for at most 100 results using the LIMIT. '''
+        dialect_prompt = AWS_REDSHIFT_DIALECT_PROMPT_CLAUDE3
     else:
         dialect_prompt = DEFAULT_DIALECT_PROMPT