MySQL Connector/Python

MySQL Connector/Python enables Python programs to access MySQL databases, using an API that is compliant with the Python Database API Specification v2.0 (PEP 249) - We refer to it as the Classic API.

It also contains an implementation of the X DevAPI - An Application Programming Interface for working with the MySQL Document Store.

Features

• Asynchronous Connectivity
• C-extension
• Telemetry
• X DevAPI

Installation

Connector/Python contains the Classic and X DevAPI connector APIs, which are installed separately. Any of these can be installed from a binary or source distribution.

Binaries are distributed in the following package formats:

• RPM
• WHEEL

On the other hand, the source code is distributed as a compressed file from which a wheel package can be built.

The recommended way to install Connector/Python is via pip, which relies on WHEEL packages. For such a reason, it is the installation procedure that is going to be described moving forward.

Please, refer to the official MySQL documentation Connector/Python Installation to know more about installing from an RPM, or building and installing a WHEEL package from a source distribution.

Before installing a package with pip, it is strongly suggested to have the most recent pip version installed on your system. If your system already has pip installed, you might need to update it. Or you can use the standalone pip installer.

The Classic API can be installed via pip as follows:

$ pip install mysql-connector-python

similarly, the X DevAPI can be installed with:

$ pip install mysqlx-connector-python

Please refer to the installation tutorial for installation alternatives of the X DevAPI.

Installation Options

Connector packages included in MySQL Connector/Python allow you to install optional dependencies to unleash certain functionalities.

# 3rd party packages to unleash the telemetry functionality are installed
$ pip install mysql-connector-python[telemetry]

similarly, for the X DevAPI:

# 3rd party packages to unleash the compression functionality are installed
$ pip install mysqlx-connector-python[compression]

This installation option can be seen as a shortcut to install all the dependencies needed by a particular feature. Mind that this is optional and you are free to install the required dependencies by yourself.

Options for the Classic API connector:

• dns-srv
• gssapi
• webauthn
• telemetry

Options for the X DevAPI connector:

• dns-srv
• compression

Classic API Sample Code

import mysql.connector

# Connect to server
cnx = mysql.connector.connect(
    host="127.0.0.1",
    port=3306,
    user="mike",
    password="s3cre3t!")

# Get a cursor
cur = cnx.cursor()

# Execute a query
cur.execute("SELECT CURDATE()")

# Fetch one result
row = cur.fetchone()
print("Current date is: {0}".format(row[0]))

# Close connection
cnx.close()

X DevAPI Sample Code

import mysqlx

# Connect to server
session = mysqlx.get_session(
   host="127.0.0.1",
   port=33060,
   user="mike",
   password="s3cr3t!")
schema = session.get_schema("test")

# Use the collection "my_collection"
collection = schema.get_collection("my_collection")

# Specify which document to find with Collection.find()
result = collection.find("name like :param") \
                   .bind("param", "S%") \
                   .limit(1) \
                   .execute()

# Print document
docs = result.fetch_all()
print(r"Name: {0}".format(docs[0]["name"]))

# Close session
session.close()

HeatWave GenAI and Machine Learning Support

MySQL Connector/Python now includes an optional API for integrating directly with MySQL HeatWave's AI and Machine Learning capabilities. This new SDK is designed to reduce the time required to generate proofs-of-concept (POCs) by providing an intuitive Pythonic interface that automates the management of SQL tables and procedures.

The new mysql.ai module offers two primary components:

• GenAI: Provides implementations of LangChain's abstract LLM, VectorStore, and Embeddings classes (MyLLM, MyVectorStore, MyEmbeddings). This ensures full interoperability with existing LangChain pipelines, allowing developers to easily substitute existing components with HeatWave-backed versions.
• AutoML: Provides Scikit-Learn compatible estimators (MyClassifier, MyRegressor, MyAnomalyDetector, MyGenericTransformer) that inherit from standard Scikit-Learn mixins. These components accept Pandas DataFrames and can be dropped directly into existing Scikit-Learn pipelines and grid searches.

Note on Dependencies: These features introduce dependencies on langchain, pandas, and scikit-learn. To keep existing installations unchanged and the base connector lightweight, these dependencies are not installed by default. You must install them separately to use the mysql.ai features.

Example: GenAI Chatbot with Memory

This example demonstrates how to use MyLLM within a loop to create a simple chatbot that maintains conversation history.

from collections import deque
from mysql import connector
from mysql.ai.genai import MyLLM

def run_chatbot(db_connection, chat_history_size=5):
    # Initialize MyLLM with the database connection
    my_llm = MyLLM(db_connection)

    # Maintain a limited history for context
    chat_history = deque(maxlen=chat_history_size)
    system_msg = "System: You are a helpful AI assistant."

    while True:
        user_input = input("\nUser: ")
        if user_input.lower() in ["exit", "quit"]:
            break

        # Format history and invoke the LLM
        history = [system_msg] + list(chat_history) + [f"User: {user_input}"]
        prompt = "\n".join(history)

        # Invoke HeatWave GenAI
        response = my_llm.invoke(prompt)
        print(f"Bot: {response}")

        # Update history
        chat_history.append(f"User: {user_input}")
        chat_history.append(f"Bot: {response}")

# Usage
with connector.connect(user='root', database='mlcorpus') as db_connection:
    run_chatbot(db_connection)

Example: HeatWave AutoML in a Scikit-Learn Pipeline

This example shows how to use MyClassifier as a drop-in replacement within a standard Scikit-Learn pipeline.

import pandas as pd
from mysql import connector
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import StandardScaler
from mysql.ai.ml import MyClassifier

# 1. Setup Data (Pandas DataFrame)
X = pd.DataFrame([[0.5, 0.1], [1.0, 0.8], [0.1, 0.2]], columns=["feat1", "feat2"])
y = pd.Series([0, 1, 0], name="target")

# 2. Connect and Train
with connector.connect(user='root', database='mlcorpus') as db_connection:
    # Initialize the HeatWave classifier
    clf = MyClassifier(db_connection)

    # Create a standard Scikit-Learn pipeline
    pipe = Pipeline([
        ("scaler", StandardScaler()),
        ("mysql_clf", clf)
    ])

    # Fit the model (automates upload and training on HeatWave)
    pipe.fit(X, y)

    # Predict
    preds = pipe.predict(X)
    print(f"Predictions: {preds}")

    # Score
    score = pipe.score(X, y)
    print(f"Accuracy: {score}")

Additional Resources

• MySQL Connector/Python Developer Guide
• MySQL Connector/Python X DevAPI Reference
• MySQL Connector/Python Forum
• MySQL Public Bug Tracker
• Slack (Sign-up required if you do not have an Oracle account)
• Stack Overflow
• Oracle Blogs

Contributing

There are a few ways to contribute to the Connector/Python code. Please refer to the contributing guidelines for additional information.

License

Please refer to the README.txt and LICENSE.txt files, available in this repository, for further details.