PyDynamoDB

Warnings

• gotcha PyDynamoDB is a DB API 2.0 client, not an ORM like PynamoDB. They are distinct libraries with different APIs and features. Ensure you are using the correct library and its corresponding import paths.
  Fix: Verify your imports (`from pydynamodb import connect` for PyDynamoDB) and consult the correct documentation for the chosen library.
• gotcha DynamoDB is a NoSQL database and differs significantly from relational databases. Common SQL patterns like normalizing data, simple primary key designs, and over-reliance on filter expressions lead to inefficient and costly operations. Filter expressions are applied *after* scan/query, impacting performance. [19, 22]
  Fix: Design your DynamoDB table schema with access patterns in mind, focusing on denormalization, efficient primary key design (hash and range keys), and utilizing indexes for filtering. Avoid using filter expressions for primary filtering.
• gotcha DynamoDB does not natively support float types for numeric values; it expects decimal types. Passing standard Python floats can lead to unexpected behavior or data loss due to precision issues when interacting with the service. [20]
  Fix: Convert Python float values to `Decimal` objects (from the `decimal` module) before writing them to DynamoDB to ensure accurate representation and storage. For example, `Decimal(str(my_float_value))`.
• gotcha Overusing DynamoDB transactions can lead to higher costs and performance issues, as each transactional write consumes twice the write capacity. Transactions are limited to 25 items and 4MB total size. Crossing updates on items in parallel transactions can also cause `TransactionCanceledException`. [21]
  Fix: Reserve transactions for cases where two or more items absolutely must succeed or fail together. For other scenarios, evaluate if conditional expressions are sufficient. Model data to minimize the need for multi-item transactions.
• gotcha Always wrap DynamoDB operations with try-catch blocks to gracefully handle failures, even though AWS SDKs (like boto3, which PyDynamoDB uses) handle internal retries with exponential backoff. Specific DynamoDB errors (e.g., `ConditionalCheckFailedException`, `ValidationException`, `ResourceNotFoundException`) require explicit handling. [20]
  Fix: Implement robust error handling around your DynamoDB interactions to manage various service-specific exceptions and ensure application stability.

Install

• pip install pydynamodb Install PyDynamoDB

Imports

• connect

  from pydynamodb import connect

  PyDynamoDB is distinct from PynamoDB (an ORM library) despite similar names. Ensure you import from 'pydynamodb'.

Quickstart

This quickstart demonstrates how to connect to DynamoDB, create a table using DDL, insert data using PartiQL, and query it. Ensure you have AWS credentials configured (e.g., via environment variables `AWS_ACCESS_KEY_ID`, `AWS_SECRET_ACCESS_KEY`, `AWS_DEFAULT_REGION`) or replace the placeholders. [6]

import os
from pydynamodb import connect

aws_access_key_id = os.environ.get('AWS_ACCESS_KEY_ID', 'YOUR_ACCESS_KEY')
aws_secret_access_key = os.environ.get('AWS_SECRET_ACCESS_KEY', 'YOUR_SECRET_KEY')
region_name = os.environ.get('AWS_DEFAULT_REGION', 'us-east-1')

try:
    # Establish a connection
    conn = connect(
        aws_access_key_id=aws_access_key_id,
        aws_secret_access_key=aws_secret_access_key,
        region_name=region_name
    )

    # Create a cursor object
    cursor = conn.cursor()

    # Execute a DDL statement (MySQL-like syntax)
    cursor.execute("CREATE TABLE IF NOT EXISTS MyTestTable (id STRING HASH KEY, name STRING)")
    print("Table 'MyTestTable' created or already exists.")

    # Execute a DML statement (PartiQL syntax)
    cursor.execute("INSERT INTO MyTestTable VALUE {'id': ?, 'name': ?}", '1', 'Alice')
    print("Inserted item 'Alice'.")

    # Query data
    cursor.execute("SELECT * FROM MyTestTable WHERE id = ?", '1')
    result = cursor.fetchall()
    print(f"Queried result: {result}")

    # Clean up (optional)
    # cursor.execute("DROP TABLE MyTestTable")
    # print("Table 'MyTestTable' dropped.")

    # Close the connection
    cursor.close()
    conn.close()

except Exception as e:
    print(f"An error occurred: {e}")