apsi 0.1.2

Python wrapper for labeled and unlabeled asynchronous private set intersection (APSI).

PyAPSI

Python wrapper for labeled and unlabeled asynchronous private set intersection (APSI).

Setup

For manylinux_2_31_x86_64 compatible platforms you can install PyAPSI from PyPi with

pip install apsi

You can check the system library versions that are required to be manylinux_2_31_x86_64 compatible in the auditwheel policy.

In case you feel like contributing a build setup for Windows and OSX compatible wheels or extend the "From Source" section below, I would be happy to review your pull request.

Example

Example usage of the labeled APSI server and client. The unlabeled variant can be used analogous to this.

from apsi import LabeledServer, LabeledClient

apsi_params = """
{
    "table_params": {
        "hash_func_count": 3,
        "table_size": 512,
        "max_items_per_bin": 92
    },
    "item_params": {"felts_per_item": 8},
    "query_params": {
        "ps_low_degree": 0,
        "query_powers": [1, 3, 4, 5, 8, 14, 20, 26, 32, 38, 41, 42, 43, 45, 46]
    },
    "seal_params": {
        "plain_modulus": 40961,
        "poly_modulus_degree": 4096,
        "coeff_modulus_bits": [40, 32, 32]
    }
}
"""

server = LabeledServer()
server.init_db(apsi_params, max_label_length=10)
server.add_items([("item", "1234567890"), ("abc", "123"), ("other", "my label")])

client = LabeledClient(apsi_params)

oprf_request = client.oprf_request(["item", "abc"])
oprf_response = server.handle_oprf_request(oprf_request)
query = client.build_query(oprf_response)
response = server.handle_query(query)
result = client.extract_result(response)

assert result == {"item": "1234567890", "abc": "123"}

To control multi threading and logging in APSI see apsi.utils.

Building & Testing

Docker

Before you start, make sure that Taskfile, Docker and Poetry are installed.

You can then run a full build with tests that will generate a wheel file in dist/ as follows:

task wheel PYTHON_VERSION=3.10.4

Note: Only Python 3.8, 3.9, 3.10, and their patch versions for which official Python Docker images exist are supported.

From Source

Please have a look at the files inside docker/ for the required vcpkg setup and apsi AVX2 patch, in case you'd like to build from source in a custom environment.