# DynamoDB Encryption Client for Python
**Note**
Our client-side encryption library was [renamed to AWS Database Encryption SDK](DDBEC-rename.md). The following topic provides information on versions 1.*x*—2.*x* of the DynamoDB Encryption Client for Java and versions 1.*x*—3.*x* of the DynamoDB Encryption Client for Python. For more information, see [AWS Database Encryption SDK for DynamoDB version support](legacy-dynamodb-encryption-client.md#legacy-support).
This topic explains how to install and use the DynamoDB Encryption Client for Python. You can find the code in the [aws-dynamodb-encryption-python](https://github.com/aws/aws-dynamodb-encryption-python/) repository on GitHub, including complete and tested [sample code](https://github.com/aws/aws-dynamodb-encryption-python/tree/master/examples) to help you get started.
**Note**
Versions 1.*x*.*x* and 2.*x*.*x* of the DynamoDB Encryption Client for Python are in [end-of-support phase](what-is-database-encryption-sdk.md#support) effective July 2022. Upgrade to a newer version as soon as possible.
**Topics**
+ [Prerequisites](#python-prerequisites)
+ [Installation](#python-installation)
+ [Using the DynamoDB Encryption Client for Python](python-using.md)
+ [Python examples](python-examples.md)
## Prerequisites
Before you install the Amazon DynamoDB Encryption Client for Python, be sure you have the following prerequisites.
**A supported version of Python**
Python 3.8 or later is required by the Amazon DynamoDB Encryption Client for Python versions 3.3.0 and later. To download Python, see [Python downloads](https://www.python.org/downloads/).
Earlier versions of the Amazon DynamoDB Encryption Client for Python support Python 2.7 and Python 3.4 and later, but we recommend that you use the latest version of the DynamoDB Encryption Client.
**The pip installation tool for Python**
Python 3.6 and later include **pip**, although you might want to upgrade it. For more information about upgrading or installing pip, see [Installation](https://pip.pypa.io/en/latest/installation/) in the **pip** documentation.
## Installation
Use **pip** to install the Amazon DynamoDB Encryption Client for Python, as shown in the following examples.
**To install the latest version**
```
pip install dynamodb-encryption-sdk
```
For more details about using **pip** to install and upgrade packages, see [Installing Packages](https://packaging.python.org/tutorials/installing-packages/).
The DynamoDB Encryption Client requires the [cryptography library](https://cryptography.io/en/latest/) on all platforms. All versions of **pip** install and build the **cryptography** library on Windows. **pip** 8.1 and later installs and builds **cryptography** on Linux. If you are using an earlier version of **pip** and your Linux environment doesn't have the tools needed to build the **cryptography** library, you need to install them. For more information, see [Building cryptography on Linux](https://cryptography.io/en/latest/installation/#building-cryptography-on-linux).
You can get the latest development version of the DynamoDB Encryption Client from the [aws-dynamodb-encryption-python](https://github.com/aws/aws-dynamodb-encryption-python/) repository on GitHub.
After you install the DynamoDB Encryption Client, get started by looking at the example Python code in this guide.
# Using the DynamoDB Encryption Client for Python
**Note**
Our client-side encryption library was [renamed to AWS Database Encryption SDK](DDBEC-rename.md). The following topic provides information on versions 1.*x*—2.*x* of the DynamoDB Encryption Client for Java and versions 1.*x*—3.*x* of the DynamoDB Encryption Client for Python. For more information, see [AWS Database Encryption SDK for DynamoDB version support](legacy-dynamodb-encryption-client.md#legacy-support).
This topic explains some of the features of the DynamoDB Encryption Client for Python that might not be found in other programming language implementations. These features are designed to make it easier to use the DynamoDB Encryption Client in the most secure way. Unless you have an unusual use case, we recommend that you use them.
For details about programming with the DynamoDB Encryption Client, see the [Python examples](python-examples.md) in this guide, the [examples](https://github.com/aws/aws-dynamodb-encryption-python/tree/master/examples) in the aws-dynamodb-encryption-python repository on GitHub, and the [Python documentation](https://aws-dynamodb-encryption-python.readthedocs.io/en/latest/) for the DynamoDB Encryption Client.
**Topics**
+ [Client helper classes](#python-helpers)
+ [TableInfo class](#table-info)
+ [Attribute actions in Python](#python-attribute-actions)
## Client helper classes
The DynamoDB Encryption Client for Python includes several client helper classes that mirror the Boto 3 classes for DynamoDB. These helper classes are designed to make it easier to add encryption and signing to your existing DynamoDB application and avoid the most common problems, as follows:
+ Prevent you from encrypting the primary key in your item, either by adding an override action for the primary key to the [AttributeActions](#python-attribute-actions) object, or by throwing an exception if your `AttributeActions` object explicitly tells the client to encrypt the primary key. If the default action in your `AttributeActions` object is `DO_NOTHING`, the client helper classes use that action for the primary key. Otherwise, they use `SIGN_ONLY`.
+ Create a [TableInfo object](#python-helpers) and populate the [DynamoDB encryption context](concepts.md#encryption-context) based on a call to DynamoDB. This helps to ensure that your DynamoDB encryption context is accurate and the client can identify the primary key.
+ Support methods, such as `put_item` and `get_item`, that transparently encrypt and decrypt your table items when you write to or read from a DynamoDB table. Only the `update_item` method is unsupported.
You can use the client helper classes instead of interacting directly with the lower-level [item encryptor](DDBEC-legacy-concepts.md#item-encryptor). Use these classes unless you need to set advanced options in the item encryptor.
The client helper classes include:
+ [EncryptedTable](https://aws-dynamodb-encryption-python.readthedocs.io/en/latest/lib/encrypted/table.html#module-dynamodb_encryption_sdk.encrypted.table) for applications that use the [Table](https://boto3.amazonaws.com/v1/documentation/api/latest/reference/services/dynamodb.html#table) resource in DynamoDB to process one table at a time.
+ [EncryptedResource](https://aws-dynamodb-encryption-python.readthedocs.io/en/latest/lib/encrypted/resource.html) for applications that use the [Service Resource](https://boto3.amazonaws.com/v1/documentation/api/latest/reference/services/dynamodb.html#service-resource) class in DynamoDB for batch processing.
+ [EncryptedClient](https://aws-dynamodb-encryption-python.readthedocs.io/en/latest/lib/encrypted/client.html) for applications that use the [lower-level client](https://boto3.amazonaws.com/v1/documentation/api/latest/reference/services/dynamodb.html#client) in DynamoDB.
To use the client helper classes, the caller must have permission to call the DynamoDB [DescribeTable](https://docs.aws.amazon.com/amazondynamodb/latest/APIReference/API_DescribeTable.html) operation on the target table.
## TableInfo class
The [TableInfo](https://aws-dynamodb-encryption-python.readthedocs.io/en/latest/lib/tools/structures.html#dynamodb_encryption_sdk.structures.TableInfo) class is a helper class that represents a DynamoDB table, complete with fields for its primary key and secondary indexes. It helps you to get accurate, real-time information about the table.
If you use a [client helper class](#python-helpers), it creates and uses a `TableInfo` object for you. Otherwise, you can create one explicitly. For an example, see [Use the item encryptor](python-examples.md#python-example-item-encryptor).
When you call the `refresh_indexed_attributes` method on a `TableInfo` object, it populates the property values of the object by calling the DynamoDB [DescribeTable](https://docs.aws.amazon.com/amazondynamodb/latest/APIReference/API_DescribeTable.html) operation. Querying the table is much more reliable than hard-coding index names. The `TableInfo` class also includes an `encryption_context_values` property that provides the required values for the [DynamoDB encryption context](concepts.md#encryption-context).
To use the `refresh_indexed_attributes` method, the caller must have permission to call the DynamoDB [DescribeTable](https://docs.aws.amazon.com/amazondynamodb/latest/APIReference/API_DescribeTable.html) operation on the target table.
## Attribute actions in Python
[Attribute actions](DDBEC-legacy-concepts.md#legacy-attribute-actions) tell the item encryptor which actions to perform on each attribute of the item. To specify attribute actions in Python, create an `AttributeActions` object with a default action and any exceptions for particular attributes. The valid values are defined in the `CryptoAction` enumerated type.
**Important**
After you use your attribute actions to encrypt your table items, adding or removing attributes from your data model might cause a signature validation error that prevents you from decrypting your data. For a detailed explanation, see [Changing your data model](data-model.md).
```
DO_NOTHING = 0
SIGN_ONLY = 1
ENCRYPT_AND_SIGN = 2
```
For example, this `AttributeActions` object establishes `ENCRYPT_AND_SIGN` as the default for all attributes, and specifies exceptions for the `ISBN` and `PublicationYear` attributes.
```
actions = AttributeActions(
default_action=CryptoAction.ENCRYPT_AND_SIGN,
attribute_actions={
'ISBN': CryptoAction.DO_NOTHING,
'PublicationYear': CryptoAction.SIGN_ONLY
}
)
```
If you use a [client helper class](#python-helpers), you don't need to specify an attribute action for the primary key attributes. The client helper classes prevent you from encrypting your primary key.
If you do not use a client helper class and the default action is `ENCRYPT_AND_SIGN`, you must specify an action for the primary key. The recommended action for primary keys is `SIGN_ONLY`. To make this easy, use the `set_index_keys` method, which uses SIGN\$1ONLY for primary keys, or DO\$1NOTHING, when that is the default action.
**Warning**
Do not encrypt the primary key attributes. They must remain in plaintext so DynamoDB can find the item without running a full table scan.
```
actions = AttributeActions(
default_action=CryptoAction.ENCRYPT_AND_SIGN,
)
actions.set_index_keys(*table_info.protected_index_keys())
```
# Example code for the DynamoDB Encryption Client for Python
**Note**
Our client-side encryption library was [renamed to AWS Database Encryption SDK](DDBEC-rename.md). The following topic provides information on versions 1.*x*—2.*x* of the DynamoDB Encryption Client for Java and versions 1.*x*—3.*x* of the DynamoDB Encryption Client for Python. For more information, see [AWS Database Encryption SDK for DynamoDB version support](legacy-dynamodb-encryption-client.md#legacy-support).
The following examples show you how to use the DynamoDB Encryption Client for Python to protect DynamoDB data in your application. You can find more examples (and contribute your own) in the [examples](https://github.com/aws/aws-dynamodb-encryption-python/tree/master/examples) directory of the [aws-dynamodb-encryption-python](https://github.com/aws/aws-dynamodb-encryption-python/) repository on GitHub.
**Topics**
+ [Use the EncryptedTable client helper class](#python-example-table)
+ [Use the item encryptor](#python-example-item-encryptor)
## Use the EncryptedTable client helper class
The following example shows you how to use the [Direct KMS Provider](direct-kms-provider.md) with the `EncryptedTable` [client helper class](python-using.md#python-helpers). This example uses the same [cryptographic materials provider](DDBEC-legacy-concepts.md#concept-material-provider) as the [Use the item encryptor](#python-example-item-encryptor) example that follows. However, it uses the `EncryptedTable` class instead of interacting directly with the lower-level [item encryptor](DDBEC-legacy-concepts.md#item-encryptor).
By comparing these examples, you can see the work that the client helper class does for you. This includes creating the [DynamoDB encryption context](concepts.md#encryption-context) and making sure the primary key attributes are always signed, but never encrypted. To create the encryption context and discover the primary key, the client helper classes call the DynamoDB [DescribeTable](https://docs.aws.amazon.com/amazondynamodb/latest/APIReference/API_DescribeTable.html) operation. To run this code, you must have permission to call this operation.
**See the complete code sample**: [aws\$1kms\$1encrypted\$1table.py](https://github.com/aws/aws-dynamodb-encryption-python/blob/master/examples/src/dynamodb_encryption_sdk_examples/aws_kms_encrypted_table.py)
Step 1: Create the table
Start by creating an instance of a standard DynamoDB table with the table name.
```
table_name='test-table'
table = boto3.resource('dynamodb').Table(table_name)
```
Step 2: Create a cryptographic materials provider
Create an instance of the [cryptographic materials provider](crypto-materials-providers.md) (CMP) that you selected.
This example uses the [Direct KMS Provider](direct-kms-provider.md), but you can use any compatible CMP. To create a Direct KMS Provider, specify an [AWS KMS key](https://docs.aws.amazon.com/kms/latest/developerguide/concepts.html#master_keys). This example uses the Amazon Resource Name (ARN) of the AWS KMS key, but you can use any valid key identifier.
```
kms_key_id='arn:aws:kms:us-west-2:111122223333:key/1234abcd-12ab-34cd-56ef-1234567890ab'
kms_cmp = AwsKmsCryptographicMaterialsProvider(key_id=kms_key_id)
```
Step 3: Create the attribute actions object
[Attribute actions](DDBEC-legacy-concepts.md#legacy-attribute-actions) tell the item encryptor which actions to perform on each attribute of the item. The `AttributeActions` object in this example encrypts and signs all items except for the `test` attribute, which is ignored.
Do not specify attribute actions for the primary key attributes when you use a client helper class. The `EncryptedTable` class signs, but never encrypts, the primary key attributes.
```
actions = AttributeActions(
default_action=CryptoAction.ENCRYPT_AND_SIGN,
attribute_actions={'test': CryptoAction.DO_NOTHING}
)
```
Step 4: Create the encrypted table
Create the encrypted table using the standard table, the Direct KMS Provider, and the attribute actions. This step completes the configuration.
```
encrypted_table = EncryptedTable(
table=table,
materials_provider=kms_cmp,
attribute_actions=actions
)
```
Step 5: Put the plaintext item in the table
When you call the `put_item` method on the `encrypted_table`, your table items are transparently encrypted, signed, and added to your DynamoDB table.
First, define the table item.
```
plaintext_item = {
'partition_attribute': 'value1',
'sort_attribute': 55
'example': 'data',
'numbers': 99,
'binary': Binary(b'\x00\x01\x02'),
'test': 'test-value'
}
```
Then, put it in the table.
```
encrypted_table.put_item(Item=plaintext_item)
```
To get the item from the DynamoDB table in its encrypted form, call the `get_item` method on the `table` object. To get the decrypted item, call the `get_item` method on the `encrypted_table` object.
## Use the item encryptor
This example shows you how to interact directly with the [item encryptor](DDBEC-legacy-concepts.md#item-encryptor) in the DynamoDB Encryption Client when encrypting table items, instead of using the [client helper classes](python-using.md#python-helpers) that interact with the item encryptor for you.
When you use this technique, you create the DynamoDB encryption context and configuration object (`CryptoConfig`) manually. Also, you encrypt the items in one call and put them in your DynamoDB table in a separate call. This allows you to customize your `put_item` calls and use the DynamoDB Encryption Client to encrypt and sign structured data that is never sent to DynamoDB.
This example uses the [Direct KMS Provider](direct-kms-provider.md), but you can use any compatible CMP.
**See the complete code sample**: [aws\$1kms\$1encrypted\$1item.py](https://github.com/aws/aws-dynamodb-encryption-python/blob/master/examples/src/dynamodb_encryption_sdk_examples/aws_kms_encrypted_item.py)
Step 1: Create the table
Start by creating an instance of a standard DynamoDB table resource with the table name.
```
table_name='test-table'
table = boto3.resource('dynamodb').Table(table_name)
```
Step 2: Create a cryptographic materials provider
Create an instance of the [cryptographic materials provider](crypto-materials-providers.md) (CMP) that you selected.
This example uses the [Direct KMS Provider](direct-kms-provider.md), but you can use any compatible CMP. To create a Direct KMS Provider, specify an [AWS KMS key](https://docs.aws.amazon.com/kms/latest/developerguide/concepts.html#master_keys). This example uses the Amazon Resource Name (ARN) of the AWS KMS key, but you can use any valid key identifier.
```
kms_key_id='arn:aws:kms:us-west-2:111122223333:key/1234abcd-12ab-34cd-56ef-1234567890ab'
kms_cmp = AwsKmsCryptographicMaterialsProvider(key_id=kms_key_id)
```
Step 3: Use the TableInfo helper class
To get information about the table from DynamoDB, create an instance of the [TableInfo](python-using.md#python-helpers) helper class. When you work directly with the item encryptor, you need to create a `TableInfo` instance and call its methods. The [client helper classes](python-using.md#python-helpers) do this for you.
The `refresh_indexed_attributes` method of `TableInfo` uses the [DescribeTable](https://docs.aws.amazon.com/amazondynamodb/latest/APIReference/API_DescribeTable.html) DynamoDB operation to get real-time, accurate information about the table. This includes its primary key and its local and global secondary indexes. The caller needs to have permission to call `DescribeTable`.
```
table_info = TableInfo(name=table_name)
table_info.refresh_indexed_attributes(table.meta.client)
```
Step 4: Create the DynamoDB encryption context
The [DynamoDB encryption context](concepts.md#encryption-context) contains information about the table structure and how it is encrypted and signed. This example creates a DynamoDB encryption context explicitly, because it interacts with the item encryptor. The [client helper classes](python-using.md#python-helpers) create the DynamoDB encryption context for you.
To get the partition key and sort key, you can use the properties of the [TableInfo](python-using.md#python-helpers) helper class.
```
index_key = {
'partition_attribute': 'value1',
'sort_attribute': 55
}
encryption_context = EncryptionContext(
table_name=table_name,
partition_key_name=table_info.primary_index.partition,
sort_key_name=table_info.primary_index.sort,
attributes=dict_to_ddb(index_key)
)
```
Step 5: Create the attribute actions object
[Attribute actions](DDBEC-legacy-concepts.md#legacy-attribute-actions) tell the item encryptor which actions to perform on each attribute of the item. The `AttributeActions` object in this example encrypts and signs all items except for the primary key attributes, which are signed, but not encrypted, and the `test` attribute, which is ignored.
When you interact directly with the item encryptor and your default action is `ENCRYPT_AND_SIGN`, you must specify an alternative action for the primary key. You can use the `set_index_keys` method, which uses `SIGN_ONLY` for the primary key, or it uses `DO_NOTHING` if it's the default action.
To specify the primary key, this example uses the index keys in the [TableInfo](python-using.md#python-helpers) object, which is populated by a call to DynamoDB. This technique is safer than hard-coding primary key names.
```
actions = AttributeActions(
default_action=CryptoAction.ENCRYPT_AND_SIGN,
attribute_actions={'test': CryptoAction.DO_NOTHING}
)
actions.set_index_keys(*table_info.protected_index_keys())
```
Step 6: Create the configuration for the item
To configure the DynamoDB Encryption Client, use the objects that you just created in a [CryptoConfig](https://aws-dynamodb-encryption-python.readthedocs.io/en/latest/lib/encrypted/config.html) configuration for the table item. The client helper classes create the CryptoConfig for you.
```
crypto_config = CryptoConfig(
materials_provider=kms_cmp,
encryption_context=encryption_context,
attribute_actions=actions
)
```
Step 7: Encrypt the item
This step encrypts and signs the item, but it doesn't put it in the DynamoDB table.
When you use a client helper class, your items are transparently encrypted and signed, and then added to your DynamoDB table when you call the `put_item` method of the helper class. When you use the item encryptor directly, the encrypt and put actions are independent.
First, create a plaintext item.
```
plaintext_item = {
'partition_attribute': 'value1',
'sort_key': 55,
'example': 'data',
'numbers': 99,
'binary': Binary(b'\x00\x01\x02'),
'test': 'test-value'
}
```
Then, encrypt and sign it. The `encrypt_python_item` method requires the `CryptoConfig` configuration object.
```
encrypted_item = encrypt_python_item(plaintext_item, crypto_config)
```
Step 8: Put the item in the table
This step puts the encrypted and signed item in the DynamoDB table.
```
table.put_item(Item=encrypted_item)
```
To view the encrypted item, call the `get_item` method on the original `table` object, instead of the `encrypted_table` object. It gets the item from the DynamoDB table without verifying and decrypting it.
```
encrypted_item = table.get_item(Key=partition_key)['Item']
```
The following image shows part of an example encrypted and signed table item.
The encrypted attribute values are binary data. The names and values of the primary key attributes (`partition_attribute` and `sort_attribute`) and the `test` attribute remain in plaintext. The output also shows the attribute that contains the signature (`*amzn-ddb-map-sig*`) and the [materials description attribute](DDBEC-legacy-concepts.md#legacy-material-description) (`*amzn-ddb-map-desc*`).
![\[An excerpt of an encrypted and signed item\]](http://docs.aws.amazon.com/database-encryption-sdk/latest/devguide/images/encrypted-item-closeup.png)