from dotenv import load_dotenv load_dotenv() import streamlit as st import google.generativeai as genai import sqlite3 import os genai.configure(api_key=os.getenv("GOOGLE_API_KEY")) model=genai.GenerativeModel('gemini-pro') prompt=[ """ You are an expert in converting English questions to SQL query! The SQL database has the name STUDENT and has the following columns - NAME, CLASS, SECTION \n\nFor example,\nExample 1 - How many entries of records are present?, the SQL command will be something like this SELECT COUNT(*) FROM STUDENT ; \nExample 2 - Tell me all the students studying in Data Science class?, the SQL command will be something like this SELECT * FROM STUDENT where CLASS="Data Science"; \nExample 3-i marks should be greater 40 and atleast 2 person have score above 40 , retrive that class the SQL command will be something like this SELECT CLASS FROM STUDENT GROUP BY CLASS HAVING COUNT(*) >= 2 AND AVG(MARKS) > 50; \nExample 4-Find the names and marks of students in the Science class who have scored more than 60 marks. SQL Command Example: SELECT NAME, MARKS FROM STUDENT WHERE CLASS='Science' AND MARKS > 60; \nExample 5-List the classes with the highest average marks. SQL Command Example: SELECT CLASS FROM STUDENT GROUP BY CLASS HAVING AVG(MARKS) = (SELECT MAX(AVG(MARKS)) FROM STUDENT GROUP BY CLASS); also the sql code should not have ``` in beginning or end and sql word in output and """ ] #llm response def gemini_sql_query(prompt,input): response=model.generate_content([prompt[0],input]) return response.text #dun to retrieve query from the sql database def read_sql_query(sql,db): conn=sqlite3.connect(db) cursor=conn.cursor() cursor.execute(sql) rows=cursor.fetchall() conn.commit() conn.close() for row in rows: print(row) return rows st.set_page_config("DataChat: Explore Your Database") st.header("DataChat: Chat With SQL Database") question=st.text_input("Enter your input/question") table_name = st.text_input("Enter the correct table name") input=f"{question} in {table_name} table" #save uploaded file def save_uploaded_file(uploaded_file): file_path = os.path.join(os.getcwd(), "uploaded.db") with open(file_path, "wb") as f: f.write(uploaded_file.getbuffer()) return file_path # File uploader component st.sidebar.header("Database Upload") uploaded_file = st.sidebar.file_uploader("Upload SQLite Database", type=["db"]) if uploaded_file is not None: # Save the uploaded file db_path = save_uploaded_file(uploaded_file) st.sidebar.success("Database uploaded successfully.") submit=st.button("submit") if submit and uploaded_file and input: query=gemini_sql_query(prompt,input) response=read_sql_query(query,db_path) print(query) col1, col2 = st.columns(2) with col1: st.header("Response:") for row in response: values = [str(value) for value in row] st.write(*values) with col2: st.header("Generated SQL Query:") with st.container(height=300): st.code(query)