Spaces:

DevashishNagpal
/

nlp-to-sql-chat-assistant

Sleeping

App Files Files Community

Devashish-Nagpal commited on 13 days ago

Commit

ddf9ea0

1 Parent(s): 345c7d2

Updated the approach and shifted the flask app to gradio app for deployment on HuggingFace

Browse files

Files changed (5) hide show

app/db.py +4 -1
app/main.py +100 -45
app/nlp.py +68 -3
dockerfile +7 -2
requirements.txt +0 -0

app/db.py CHANGED Viewed

@@ -1,12 +1,14 @@
 import sqlite3
 from sqlite3 import Error
-from app.utils import validate_sql
 class Database:
     def __init__(self, db_path='data/database.sqlite'):
         self.db_path = db_path
     def execute_query(self, query):
         if not validate_sql(query):
             return {"error": "Invalid SQL query. Only SELECT queries are allowed at this point."}
         try:
@@ -18,4 +20,5 @@ class Database:
             conn.close()
             return {"columns": columns, "data": results}
         except Error as e:
             return {"error": str(e)}

 import sqlite3
 from sqlite3 import Error
+from .utils import validate_sql
+import logging
 class Database:
     def __init__(self, db_path='data/database.sqlite'):
         self.db_path = db_path
     def execute_query(self, query):
+        logging.info(f"Executing SQL: {query}")
         if not validate_sql(query):
             return {"error": "Invalid SQL query. Only SELECT queries are allowed at this point."}
         try:
             conn.close()
             return {"columns": columns, "data": results}
         except Error as e:
+            logging.error(f"SQL Error: {str(e)}")
             return {"error": str(e)}

app/main.py CHANGED Viewed

@@ -1,53 +1,108 @@
-from flask import Flask, request, render_template_string
-from app.nlp import NLPToSQL
-from app.db import Database
-app = Flask(__name__)
 nlp = NLPToSQL()
 db = Database()
-HTML_TEMPLATE = """
-!DOCTYPE html>
-<html>
-<head><title> Chat Assistant </title></head>
-<body>
-    <h1> Database Chat Assistant</h1>
-    <form method="POST">
-        <input type="text" name="query" placeholder= "Enter your query..." size="50">
-        <button type="submit">Ask</button>
-    </form>
-    {% if response %}
-        <h3> Response: </h3>
-        <pre>{{ response }}</pre>
-    {% endif %}
-    {% if error %}
-        <p style = "color:red;">{{ error }} </p>
-    {% endif %}
-</body>
-</html>
-"""
-@app.route("/", methods=["GET", "POST"])
-def index():
-    if request.method == 'POST':
-        user_query = request.form['query']
-        try:
-            sql = nlp.query_to_sql(user_query)
-            result = db.execute_query(sql)
-            if 'error' in result:
-                return render_template_string(HTML_TEMPLATE, error = result['error'])
-            if not result['data']:
-                return render_template_string(HTML_TEMPLATE, error = "No data found")
-            response = " | ".join(result['columns']) + "\n"
-            response += "-"*50 + "\n"
-            for row in result['data']:
-                response += " | ".join(str(cell) for cell in row) + "\n"
-                return render_template_string(HTML_TEMPLATE, response = response)
-        except Exception as e:
-            return render_template_string(HTML_TEMPLATE, error = f"Error: {str(e)}")
-    return render_template_string(HTML_TEMPLATE)

+import logging
+import gradio as gr
+from .nlp import NLPToSQL
+from .db import Database
 nlp = NLPToSQL()
 db = Database()
+logging.basicConfig(level=logging.INFO)
+def chatbot(user_query):
+    try:
+        sql = nlp.query_to_sql(user_query)
+        logging.info(f"Generated SQL: {sql}")
+        result = db.execute_query(sql)
+        if 'error' in result:
+            return f"❌ Error: {result['error']}"
+        if not result['data']:
+            return "⚠️ No data found."
+        # Formatting SQL output
+        response = "**Query:**\n" + sql + "\n\n"
+        response += "**Result:**\n"
+        response += " | ".join(result['columns']) + "\n"
+        response += "-"*50 + "\n"
+        for row in result['data']:
+            response += " | ".join(str(cell) for cell in row) + "\n"
+        return response
+    except Exception as e:
+        logging.error(f"Error: {str(e)}")
+        return f"❌ Error: {str(e)}"
+demo = gr.Interface(
+    fn=chatbot,
+    inputs=gr.Textbox(lines=2, placeholder="Ask your database anything..."),
+    outputs="markdown",
+    title="SQL Chat Assistant",
+    description="Enter a natural language question, and get the corresponding SQL query & result."
+)
+if __name__ == "__main__":
+    demo.launch(server_name="0.0.0.0", server_port=7860)
+# Flask-based application can also be used
+# import logging
+# from flask import Flask, request, render_template_string
+# from .nlp import NLPToSQL
+# from .db import Database
+# app = Flask(__name__)
+# nlp = NLPToSQL()
+# db = Database()
+# logging.basicConfig(level=logging.INFO)
+# HTML_TEMPLATE = """
+# <!DOCTYPE html>
+# <html>
+# <head><title> Chat Assistant </title></head>
+# <body>
+#     <h1> Database Chat Assistant</h1>
+#     <form method="POST">
+#         <input type="text" name="query" placeholder= "Enter your query..." size="50">
+#         <button type="submit">Ask</button>
+#     </form>
+#     {% if response %}
+#         <h3> Response: </h3>
+#         <pre>{{ response }}</pre>
+#     {% endif %}
+#     {% if error %}
+#         <p style = "color:red;">{{ error }} </p>
+#     {% endif %}
+# </body>
+# </html>
+# """
+# @app.route("/", methods=["GET", "POST"])
+# def index():
+#     if request.method == 'POST':
+#         user_query = request.form['query']
+#         try:
+#             sql = nlp.query_to_sql(user_query)
+#             logging.info(f"Generated SQL: {sql}")
+#             result = db.execute_query(sql)
+#             if 'error' in result:
+#                 return render_template_string(HTML_TEMPLATE, error=result['error'])
+#             if not result['data']:
+#                 return render_template_string(HTML_TEMPLATE, error="No data found")
+#             response = " | ".join(result['columns']) + "\n"
+#             response += "-"*50 + "\n"
+#             for row in result['data']:
+#                 response += " | ".join(str(cell) for cell in row) + "\n"
+#             return render_template_string(HTML_TEMPLATE, response=response)
+#         except Exception as e:
+#             logging.error(f"Error: {str(e)}")
+#             return render_template_string(HTML_TEMPLATE, error=f"Error: {str(e)}")
+#     return render_template_string(HTML_TEMPLATE)
+# if __name__ == "__main__":
+#     app.run(debug=True)

app/nlp.py CHANGED Viewed

@@ -1,6 +1,8 @@
 from transformers import pipeline
-class NLPToSQL:
     def __init__(self):
         self.model = pipeline(
             "text2text-generation",
@@ -9,6 +11,69 @@ class NLPToSQL:
         )
     def query_to_sql(self, user_query):
-        prompt = f"translate English to SQL: {user_query}"
         result = self.model(prompt, max_length=200)
-        return result[0]['generated_text']

 from transformers import pipeline
+import re
+from typing import Dict
+class NLPToSQL2:
     def __init__(self):
         self.model = pipeline(
             "text2text-generation",
         )
     def query_to_sql(self, user_query):
+        prompt = (f"Generate a valid SQL query in the correct format based on the following schema:\n"
+            f"Table1: Employees\n"
+            f"Columns: ID, Name, Department, Salary\n"
+            f"Table2: Departments\n"
+            f"Columns: Name, Manager\n"
+            f"Natural Language: {user_query}"
+            f"SQL query:"
+                  )
         result = self.model(prompt, max_length=200)
+        sql = result[0]['generated_text']
+        return sql
+class NLPToSQL:
+    def __init__(self):
+        self.query_patterns: Dict[str, str] = {
+            r"show\s+(?:me\s+)?all\s+employees?\s+in\s+(?:the\s+)?(\w+)\s+department":
+                "SELECT * FROM Employees WHERE LOWER(Department) = LOWER('{}')",
+            r"who\s+is\s+(?:the\s+)?manager\s+of\s+(?:the\s+)?(\w+)\s+department":
+                "SELECT Manager FROM Departments WHERE LOWER(Name) = LOWER('{}')",
+            r"list\s+(?:all\s+)?employees?\s+hired\s+after\s+(\d{4}-\d{2}-\d{2})":
+                "SELECT * FROM Employees WHERE Hire_Date > '{}'",
+            r"what\s+is\s+(?:the\s+)?total\s+salary\s+(?:expense\s+)?for\s+(?:the\s+)?(\w+)\s+department":
+                "SELECT SUM(Salary) as Total_Salary FROM Employees WHERE LOWER(Department) = LOWER('{}')",
+            r"show\s+(?:me\s+)?(?:the\s+)?salary\s+of\s+(\w+)":
+                "SELECT Salary FROM Employees WHERE LOWER(Name) = LOWER('{}')",
+            r"list\s+(?:all\s+)?employees?\s+with\s+salary\s+(?:greater|more)\s+than\s+(\d+)":
+                "SELECT * FROM Employees WHERE Salary > {}",
+            r"(?:show|list)\s+(?:me\s+)?all\s+departments":
+                "SELECT * FROM Departments",
+            r"(?:show|list)\s+(?:me\s+)?all\s+employees":
+                "SELECT * FROM Employees"
+        }
+    def query_to_sql(self, user_query: str) -> str:
+        normalized_query = " ".join(user_query.lower().split())
+        for pattern, sql_template in self.query_patterns.items():
+            match = re.search(pattern, normalized_query, re.IGNORECASE)
+            if match:
+                if match.groups():
+                    return sql_template.format(*match.groups())
+                return sql_template
+        return self._generate_fallback_query(normalized_query)
+    def _generate_fallback_query(self, query: str) -> str:
+        if any(word in query for word in ['department', 'manager']):
+            return "SELECT * FROM Departments"
+        return "SELECT * FROM Employees"
+    def sanitize_sql(self, sql: str) -> str:
+        sql = re.sub(r'[;"]', '', sql)
+        sql = sql.replace("'", "''")
+        if not sql.strip().endswith(';'):
+            sql = f"{sql};"
+        return sql

dockerfile CHANGED Viewed

@@ -10,6 +10,11 @@ COPY . .
 RUN pip install --no-cache-dir -r requirements.txt
-EXPOSE 8000
-CMD ["gunicorn", "--bind", "0.0.0.0:8000", "app.main:app"]

 RUN pip install --no-cache-dir -r requirements.txt
+EXPOSE 7860
+# EXPOSE 8000
+#(for flask)
+CMD ["gunicorn", "--bind", "0.0.0.0:7860", "app.main:app"]
+# (for flask)
+# CMD ["gunicorn", "--bind", "0.0.0.0:8000", "app.main:app"]

requirements.txt CHANGED Viewed

Binary files a/requirements.txt and b/requirements.txt differ