templates.randomized_encryption_scheme module

Generic classes for creating an EncryptionScheme that allows for precomputed, or stored randomness.

class templates.randomized_encryption_scheme.RandomizableCiphertext(raw_value, scheme, *, fresh=False)

Bases: Ciphertext[KM, PT, RP, CV], ABC

Ciphertext that can be rerandomized. Subclass of Ciphertext.

__init__(raw_value, scheme, *, fresh=False)

Construct a RandomizableCiphertext, with the given value for the given EncryptionScheme.

Parameters:

• raw_value (Any) – Ciphertext value.

• scheme (RandomizedEncryptionScheme[TypeVar(KM), TypeVar(PT), TypeVar(RP), TypeVar(CV), TypeVar(RC, bound= RandomizableCiphertext[Any, Any, Any, Any])]) – RandomizedEncryptionScheme that is used to encrypt this ciphertext.

• fresh (bool) – Indicates whether fresh randomness is already applied to the raw_value.

Raises:

TypeError – When scheme has the incorrect type.

__str__()

Return type:

str

Returns:

String representation of RandomizedCiphertext.

abstract apply_randomness(randomization_value)

Apply a random value to rerandomize this ciphertext.

Parameters:

randomization_value (Any) – Random value used to rerandomize this ciphertext.

Raises:

NotImplementedError – When scheme does not support rerandomization.

Return type:

None

property fresh: bool

Indicate whether the ciphertest has fresh randomness.

Ciphertexts that are send to other parties should generally be fresh. This can be achieved by calling self.randomize().

Returns:

True if the randomness is fresh, False otherwise.

get_value()

Get the raw value of the ciphertext.

Accessing this value marks the ciphertext as not fresh. If this is not desired, call the peek_value method.

Return type:

TypeVar(CV)

Returns:

Value of the ciphertext

peek_value()

Peek at the raw value of the ciphertext.

Accessing this value does not change the freshness of the ciphertext. If this is not desired, call the get_value method.

Return type:

TypeVar(CV)

Returns:

Value of the ciphertext

randomize()

Rerandomize this ciphertext object.

Return type:

TypeVar(RC, bound= RandomizableCiphertext[Any, Any, Any, Any])

Returns:

The rerandomized object (self).

property value: CV

Raw value of the ciphertext.

Returns:

Value of the ciphertext

class templates.randomized_encryption_scheme.RandomizedEncryptionScheme(randomizations=None, max_size=100, total=None, nr_of_threads=1, path=None, separator=',', start_generation=True, debug=False)

Bases: EncryptionScheme[KM, PT, RP, CV, RC], ABC

Abstract base class for a RandomizedEncryptionScheme. Subclass of EncryptionScheme

__init__(randomizations=None, max_size=100, total=None, nr_of_threads=1, path=None, separator=',', start_generation=True, debug=False)

Initiate a Randomness variable with the given parameters to be used for randomizing ciphertexts.

Parameters:

• randomizations (Optional[Queue[int]]) – queue with randomizations. If no queue is given, it creates a fresh one.

• max_size (int) – maximum size of the queue.

• total (Optional[int]) – upper bound on the total amount of randomizations to generate.

• nr_of_threads (int) – number of generation worker threads that should be started.

• path (Optional[str]) – path (including filename) to the file that contains randomizations. By default no path is given and no randomness is extracted from any files.

• separator (str) – separator for the random values in the given file.

• start_generation (bool) – flag that determines whether the scheme starts generating randomness immediately.

• debug (bool) – flag to determine whether debug information should be displayed.

boot_generation(nr_of_threads=None, path=None, start_generation=True)

calls the boot_generation method of the internal randomness

Parameters:

• nr_of_threads (Optional[int]) – number of generation threads. (Default: None, the nr_of_threads parameter is taken from the original __init__)

• path (Optional[str]) – path to the file containing randomizations. (Default: None, the path parameter is taken from the original __init__)

• start_generation (bool) – flag that determines whether the threads start generating immediately.

Return type:

None

abstract generate_randomness()

Method to generate randomness for this particular scheme.

Raises:

NotImplementedError – When scheme does not support randomness generation.

Return type:

int

Returns:

A single random element with respect to the scheme.

get_randomness()

Get new randomness from the randomness source.

Return type:

int

Returns:

One random value.

initialize_randomness(**kwargs)

Initializes a randomness class

Parameters:

**kwargs (Any) – keyword arguments to pass on to Randomness constructor

Return type:

None

shut_down()

Give the shut down signal to the scheme’s randomness. This tells the threads to gracefully shut down.

Return type:

None

unsafe_encrypt(plaintext, apply_encoding=True)

Encrypts the entered (encoded) Plaintext, but does not apply randomness. Also encodes the Plaintext when this is required.

Parameters:

• plaintext (Union[TypeVar(PT), EncodedPlaintext[TypeVar(RP)]]) – Plaintext or EncodedPlaintext to be encrypted.

• apply_encoding (bool) – Boolean indicating whether a non-encoded plaintext should be encoded. If False, the plaintext is encrypted in raw form.

Return type:

TypeVar(RC, bound= RandomizableCiphertext[Any, Any, Any, Any])

Returns:

Non-randomized RandomizableCiphertext object containing the encrypted value of the plaintext.

class templates.randomized_encryption_scheme.Randomness(generation_function, randomizations=None, max_size=100, total=None, nr_of_threads=1, path=None, separator=',', start_generation=True, debug=False)

Bases: object

Object containing precomputed randomness. This randomness is either stored internally or in a specified file. Allows for functionality to create randomness on-the-fly.

__init__(generation_function, randomizations=None, max_size=100, total=None, nr_of_threads=1, path=None, separator=',', start_generation=True, debug=False)

Construct a Randomness object. This construction starts generation workers and a file worker that generate new randomness using the given generation function and abstract random from a file respectively. This happens in separate threads and the random elements are placed in a (thread-safe) queue. This queue can then be used to request random elements at encryption time.

Parameters:

• debug (bool) – flag to determine whether debug information should be displayed.

• start_generation (bool) – flag to determine whether all threads should immediately start generating randomness.

• max_size (int) – maximum size of the buffer of randomizations.

• total (Optional[int]) – upper bound on the total amount of randomizations to generate.

• nr_of_threads (int) – number of threads that generate randomizations in parallel.

• generation_function (Callable[[], int]) – Function that generates one random value.

• randomizations (Optional[Queue[int]]) – Precomputed in-memory randomness to be stored in this object for usage in an encryption scheme.

• path (Optional[str]) – Optional path to a file containing randomness.

• separator (str) – Separator between random numbers in the file.

__len__()

Determine number of randomizations currently present in the queue.

Return type:

int

Returns:

Number of randomizations currently present in the queue.

add_generation_worker()

Add an extra thread that generates randomness.

Return type:

None

boot_generation(nr_of_threads=None, path=None, start_generation=True)

Shut down the generation threads and file thread if they are still running. Then, initialize new generation threads and a file thread.

Parameters:

• nr_of_threads (Optional[int]) – number of generation threads. (Default: None, the nr_of_threads parameter is taken from the original __init__)

• path (Optional[str]) – path to the file containing randomizations. (Default: None, the path parameter is taken from the original __init__)

• start_generation (bool) – flag that determines whether the threads start generating immediately.

Return type:

None

default_shutdown_timeout = 0.01

file_worker()

Function to be run by the file worker. The file is read in chunks and these chunks are then used to extract random elements.

Return type:

None

generation_function()

Wrapper around generation function

Return type:

int

Returns:

random element

generation_worker(identifier)

Function to be run by each generation worker. These workers keep generating random elements and try to add them to the queue until a stop sign is given.

Parameters:

identifier (int) – identifier used for debugging

Return type:

None

get_one()

Checks the buffer for random value. If the buffer is empty, it waits for a random value from the generation thread.

Return type:

int

Returns:

One random element.

safe_print(message)

Atomic print. It does nothing if the debug flag is False

Parameters:

message (str) – message to be printed.

Return type:

None

shut_down()

Turn on the signal that tells the threads to drop what they’re doing and shut down.

Return type:

None

start_generating()

Turn on the signal that tells the threads to start generating randomness.

Return type:

None

stop_generating()

Tell the threads that they can stop generating randomness.

Return type:

None