search

datastax / astrapy

AstraPy is a Pythonic interface for DataStax Astra DB and the Data API
https://github.com/datastax/astrapy
Apache License 2.0
18 stars 19 forks source link
astradb datastax python stargate

readme

AstraPy

A pythonic client for DataStax Astra DB.

This README targets AstraPy version 1.0.0+, which introduces a whole new API. Click here for the pre-existing API (fully compatible with newer versions).

Quickstart

Install with pip install astrapy.

Get the API Endpoint and the Token to your Astra DB instance at astra.datastax.com.

Try the following code after replacing the connection parameters:

import astrapy

ASTRA_DB_APPLICATION_TOKEN = "AstraCS:..."
ASTRA_DB_API_ENDPOINT = "https://01234567-....apps.astra.datastax.com"

my_client = astrapy.DataAPIClient()
my_database = my_client.get_database(
    ASTRA_DB_API_ENDPOINT,
    token=ASTRA_DB_APPLICATION_TOKEN,
)

my_collection = my_database.create_collection(
    "dreams",
    dimension=3,
    metric=astrapy.constants.VectorMetric.COSINE,
)

my_collection.insert_one({"summary": "I was flying", "$vector": [-0.4, 0.7, 0]})

my_collection.insert_many(
    [
        {
            "_id": astrapy.ids.UUID("018e65c9-e33d-749b-9386-e848739582f0"),
            "summary": "A dinner on the Moon",
            "$vector": [0.2, -0.3, -0.5],
        },
        {
            "summary": "Riding the waves",
            "tags": ["sport"],
            "$vector": [0, 0.2, 1],
        },
        {
            "summary": "Friendly aliens in town",
            "tags": ["scifi"],
            "$vector": [-0.3, 0, 0.8],
        },
        {
            "summary": "Meeting Beethoven at the dentist",
            "$vector": [0.2, 0.6, 0],
        },
    ],
)

my_collection.update_one(
    {"tags": "sport"},
    {"$set": {"summary": "Surfers' paradise"}},
)

cursor = my_collection.find(
    {},
    sort={"$vector": [0, 0.2, 0.4]},
    limit=2,
    include_similarity=True,
)

for result in cursor:
    print(f"{result['summary']}: {result['$similarity']}")

# This would print:
#   Surfers' paradise: 0.98238194
#   Friendly aliens in town: 0.91873914

Next steps:

Usage with HCD and other non-Astra installations

The main difference to target e.g. a Hyper-Converged Database (HCD) installation is how the client is initialized. Here is a short example showing just how to get to a Database (what comes next is unchaged compared to using Astra DB).

from astrapy import DataAPIClient
from astrapy.constants import Environment
from astrapy.authentication import UsernamePasswordTokenProvider

# Build a token
tp = UsernamePasswordTokenProvider("username", "password")

# Initialize the client and get a "Database" object
client = DataAPIClient(token=tp, environment=Environment.HCD)
database = client.get_database("http://localhost:8181", token=tp)

For more on this case, please consult the dedicated reference.

AstraPy's API

Abstraction diagram

AstraPy's abstractions for working at the data and admin layers are structured as depicted by this diagram:

AstraPy, abstractions chart

Here's a small admin-oriented example:

import astrapy

# this must have "Database Administrator" permissions:
ASTRA_DB_APPLICATION_TOKEN = "AstraCS:..."

my_client = astrapy.DataAPIClient(ASTRA_DB_APPLICATION_TOKEN)

my_astra_admin = my_client.get_admin()

database_list = list(my_astra_admin.list_databases())

db_info = database_list[0].info
print(db_info.name, db_info.id, db_info.region)

my_database_admin = my_astra_admin.get_database_admin(db_info.id)

my_database_admin.list_namespaces()
my_database_admin.create_namespace("my_dreamspace")

Exceptions

The package comes with its own set of exceptions, arranged in this hierarchy:

AstraPy, exception hierarchy

For more information, and code examples, check out the docstrings and consult the API reference linked above.

Working with dates

Date and datetime objects, i.e. instances of the standard library datetime.datetime and datetime.date classes, can be used anywhere in documents:

import datetime
import astrapy

ASTRA_DB_APPLICATION_TOKEN = "AstraCS:..."
ASTRA_DB_API_ENDPOINT = "https://01234567-....apps.astra.datastax.com"

my_client = astrapy.DataAPIClient()
my_database = my_client.get_database(
    ASTRA_DB_API_ENDPOINT,
    token=ASTRA_DB_APPLICATION_TOKEN,
)
my_collection = my_database.dreams

my_collection.insert_one({"when": datetime.datetime.now()})
my_collection.insert_one({"date_of_birth": datetime.date(2000, 1, 1)})

my_collection.update_one(
    {"registered_at": datetime.date(1999, 11, 14)},
    {"$set": {"message": "happy Sunday!"}},
)

print(
    my_collection.find_one(
        {"date_of_birth": {"$lt": datetime.date(2001, 1, 1)}},
        projection={"_id": False},
    )
)
# This would print:
#    {'date_of_birth': datetime.datetime(2000, 1, 1, 0, 0)}

Note: reads from a collection will always return the datetime class regardless of wheter a date or a datetime was provided in the insertion.

Working with ObjectIds and UUIDs

Astrapy repackages the ObjectId from bson and the UUID class and utilities from the uuid package and its uuidv6 extension. You can also use them directly.

Even when setting a default ID type for a collection, you still retain the freedom to use any ID type for any document:

import astrapy
import bson

ASTRA_DB_APPLICATION_TOKEN = "AstraCS:..."
ASTRA_DB_API_ENDPOINT = "https://01234567-....apps.astra.datastax.com"

my_client = astrapy.DataAPIClient()
my_database = my_client.get_database(
    ASTRA_DB_API_ENDPOINT,
    token=ASTRA_DB_APPLICATION_TOKEN,
)

my_collection = my_database.create_collection(
    "ecommerce",
    default_id_type=astrapy.constants.DefaultIdType.UUIDV6,
)

my_collection.insert_one({"_id": astrapy.ids.ObjectId("65fd9b52d7fabba03349d013")})
my_collection.find({
    "_id": astrapy.ids.UUID("018e65c9-e33d-749b-9386-e848739582f0"),
})

my_collection.update_one(
    {"tag": "in_stock"},
    {"$set": {"inventory_id": bson.objectid.ObjectId()}},
    upsert=True,
)

my_collection.insert_one({"_id": astrapy.ids.uuid8()})

For contributors

First install poetry with pip install poetry and then the project dependencies with poetry install --with dev.

Linter, style and typecheck should all pass for a PR:

make format

With make format-fix the style and imports are autofixed (by black and isort resp.)

Features must be thoroughly covered in tests (see tests/idiomatic/* for naming convention and module structure).

Running tests

Tests are grouped in three blocks (in as many subdirs of tests/):

Actually, for convenience, sub-blocks of tests are considered:

Tests can be run on three types of Data API targets (with slight differences in what is applicable):

Depending on the (sub-block, target) combination, some environment variables may be needed. Templates for the environment variables are to be found in tests/env_templates.

The general expectation is that idiomatic non-Admin tests, and vectorize in-depth tests, are part of the main CI flow; conversely, core, admin and vectorize all-providers are kept as a manual task to run (locally in most cases) when circumstances require it (use your judgement).

Required environment variables

Below is a detail of the reference template files needed for the various types of testing:

Sample testing commands

For the DockerCompose case, prepend all of the following with DOCKER_COMPOSE_LOCAL_DATA_API="yes".

All the usual pytest ways of restricting the test selection hold in addition (e.g. poetry run pytest tests/idiomatic/unit or [...] -k <test_name_selector>).

core regular:
poetry run pytest tests/core
core ops:

Note the special variable needed to actually run this. You will have to manually clean up afterwards.

TEST_ASTRADBOPS="1" poetry run pytest tests/core/test_ops.py
idiomatic regular:

Warning: this will also trigger the very long-running idiomatic admin Astra if the vars as in env.astra.admin.template are also detected. Likewise, the idiomatic admin nonAstra may start (if DO_IDIOMATIC_ADMIN_TESTS is set), which however takes few seconds.

poetry run pytest tests/idiomatic
idiomatic admin Astra:
poetry run pytest tests/idiomatic/integration/test_admin.py
idiomatic admin nonAstra:
DO_IDIOMATIC_ADMIN_TESTS="1" poetry run pytest tests/idiomatic/integration/test_nonastra_admin.py
vectorize in-depth:
poetry run pytest tests/vectorize_idiomatic/integration/test_vectorize_methods*.py

or just:

poetry run pytest tests/vectorize_idiomatic/integration/test_vectorize_methods_sync.py
vectorize all-providers:

This generates all possible test cases and runs them:

poetry run pytest tests/vectorize_idiomatic

For a spot test, you may restrict to one case, e.g.

EMBEDDING_MODEL_TAGS="openai/text-embedding-3-large/HEADER/0" poetry run pytest tests/vectorize_idiomatic/integration/test_vectorize_providers.py -k test_vectorize_usage_auth_type_header_sync

Useful flags for testing

Remove logging noise with:

poetry run pytest [...] -o log_cli=0

Increase logging level to TRACE (i.e. level 5):

poetry run pytest [...] -o log_cli=1 --log-cli-level=5

Do not drop collections (valid for core):

TEST_SKIP_COLLECTION_DELETE=1 poetry run pytest [...]

Appendices

Appendix A: quick reference for imports

Client, data and admin abstractions:

from astrapy import (
    DataAPIClient,
    Database,
    AsyncDatabase,
    Collection,
    AsyncCollection,
    AstraDBAdmin,
    AstraDBDatabaseAdmin,
    DataAPIDatabaseAdmin,
)

Constants for data-related use:

from astrapy.constants import (
    ReturnDocument,
    SortDocuments,
    VectorMetric,
    DefaultIdType,
    Environment,
)

ObjectIds and UUIDs:

from astrapy.ids import (
    ObjectId,
    uuid1,
    uuid3,
    uuid4,
    uuid5,
    uuid6,
    uuid7,
    uuid8,
    UUID,
)

Operations (for bulk_write collection method):

from astrapy.operations import (
    BaseOperation,
    InsertOne,
    InsertMany,
    UpdateOne,
    UpdateMany,
    ReplaceOne,
    DeleteOne,
    DeleteMany,
    AsyncBaseOperation,
    AsyncInsertOne,
    AsyncInsertMany,
    AsyncUpdateOne,
    AsyncUpdateMany,
    AsyncReplaceOne,
    AsyncDeleteOne,
    AsyncDeleteMany,
)

Result classes:

from astrapy.results import (
    OperationResult,
    DeleteResult,
    InsertOneResult,
    InsertManyResult,
    UpdateResult,
    BulkWriteResult,
)

Exceptions:

from astrapy.exceptions import (
    BulkWriteException,
    CollectionAlreadyExistsException,
    CollectionNotFoundException,
    CumulativeOperationException,
    CursorIsStartedException,
    DataAPIDetailedErrorDescriptor,
    DataAPIErrorDescriptor,
    DataAPIException,
    DataAPIFaultyResponseException,
    DataAPIHttpException,
    DataAPIResponseException,
    DataAPITimeoutException,
    DeleteManyException,
    DevOpsAPIErrorDescriptor,
    DevOpsAPIException,
    DevOpsAPIResponseException,
    InsertManyException,
    TooManyDocumentsToCountException,
    UpdateManyException,
)

Info/metadata classes:

from astrapy.info import (
    AdminDatabaseInfo,
    DatabaseInfo,
    CollectionInfo,
    CollectionVectorServiceOptions,
    CollectionDefaultIDOptions,
    CollectionVectorOptions,
    CollectionOptions,
    CollectionDescriptor,
    EmbeddingProviderParameter,
    EmbeddingProviderModel,
    EmbeddingProviderToken,
    EmbeddingProviderAuthentication,
    EmbeddingProvider,
    FindEmbeddingProvidersResult,
)

Admin-related classes and constants:

from astrapy.admin import (
    ParsedAPIEndpoint,
)

Cursors:

from astrapy.cursors import (
    BaseCursor,
    Cursor,
    AsyncCursor,
    CommandCursor,
    AsyncCommandCursor,
)

Appendix B: compatibility with pre-1.0.0 library

If your code uses the pre-1.0.0 astrapy (i.e. from astrapy.db import Database, Collection and so on) you are strongly advised to migrate to the current API.

That being said, there are no known breakings of backward compatibility: legacy code would run with a newest astrapy version just as well. Here is a recap of the minor changes that came to the old API with 1.0.0 (and beyond):

Keep in mind that the pre-1.0.0 library, now dubbed "core", is what the current 1.0.0 API ("idiomatic") builds on.