Skip to main content

Databend Driver Python Binding

Project description

databend-driver

Databend Python Client

image License image

Usage

Local Embedded Connection

The local embedded mode runs a full Databend engine in-process without any server. It is useful for local analytics, testing, and offline workflows.

Install the local extra to pull in the embedded engine:

pip install "databend-driver[local]"

The embedded dependency currently requires Python 3.12 or later.

from databend_driver import connect

# Persistent state stored under ./local-state
conn = connect("databend+local:///./local-state")
conn.exec("CREATE TABLE books(id INT, title STRING)")
conn.exec("INSERT INTO books VALUES (1, 'Databend')")

row = conn.query_row("SELECT title FROM books ORDER BY id LIMIT 1")
print(row.values())  # ('Databend',)

rows = [row.values() for row in conn.query_iter("SELECT * FROM books ORDER BY id")]

Supported local targets:

  • connect(":memory:") — temporary in-memory instance (discarded on close)
  • connect("databend+local:///:memory:") — explicit in-memory instance
  • connect("databend+local:///./local-state") — persistent state under ./local-state
  • connect("databend+local:///./local-state?tenant=default") — persistent state with an explicit tenant
  • connect("databend+local:///./local-state?database=mydb") — open a specific database

You can also use connect_local() directly for more control:

from databend_driver import connect_local

conn = connect_local(database=":memory:")
conn = connect_local(data_path="./local-state", tenant="default")

If the optional databend package is not installed, connect() raises an ImportError with guidance about enabling the local extra and the Python version requirement.

For remote Databend, the same connect() entrypoint accepts standard DSNs:

from databend_driver import connect

conn = connect("databend://root:@localhost:8000/?sslmode=disable")
row = conn.query_row("SELECT 1")

Relation API

The local connection exposes an embedded-specific relation API for working with query results as DataFrames or Arrow tables:

relation = conn.sql("SELECT * FROM books")

df = relation.df()       # pandas DataFrame
pl = relation.pl()       # polars DataFrame
tbl = relation.arrow()   # pyarrow Table

rows = relation.fetchall()   # list[tuple]
row = relation.fetchone()    # tuple | None

Registering External Data

You can register files or in-memory data as virtual tables:

# Register a Parquet file
conn.register("sales", "./data/sales.parquet")
conn.sql("SELECT * FROM sales LIMIT 10").df()

# Register a CSV file
conn.register("events", "./data/events.csv")

# Register a pandas or polars DataFrame
import pandas as pd
df = pd.DataFrame({"id": [1, 2], "name": ["Alice", "Bob"]})
conn.register("users", df)

# Shorthand: register a DataFrame and return a relation immediately
relation = conn.from_df(df)

# Read helpers (register and return relation in one call)
relation = conn.read_parquet("./data/sales.parquet")
relation = conn.read_csv("./data/events.csv")
relation = conn.read_json("./data/logs.ndjson")
relation = conn.read_text("./data/raw.txt")

PEP 249 Cursor Object

from databend_driver import BlockingDatabendClient

client = BlockingDatabendClient('databend://root:root@localhost:8000/?sslmode=disable')
cursor = client.cursor()

cursor.execute(
    """
    CREATE TABLE test (
        i64 Int64,
        u64 UInt64,
        f64 Float64,
        s   String,
        s2  String,
        d   Date,
        t   DateTime
    )
    """
)
cursor.execute("INSERT INTO test VALUES (?, ?, ?, ?, ?, ?, ?)", (1, 1, 1.0, 'hello', 'world', '2021-01-01', '2021-01-01 00:00:00'))
cursor.execute("SELECT * FROM test")
rows = cursor.fetchall()
for row in rows:
    print(row.values())
cursor.close()

Blocking Connection Object

from databend_driver import BlockingDatabendClient

client = BlockingDatabendClient('databend://root:root@localhost:8000/?sslmode=disable')
conn = client.get_conn()
conn.exec(
    """
    CREATE TABLE test (
        i64 Int64,
        u64 UInt64,
        f64 Float64,
        s   String,
        s2  String,
        d   Date,
        t   DateTime
    )
    """
)
rows = conn.query_iter("SELECT * FROM test")
for row in rows:
    print(row.values())
conn.close()

Asyncio Connection Object

import asyncio
from databend_driver import AsyncDatabendClient

async def main():
    client = AsyncDatabendClient('databend://root:root@localhost:8000/?sslmode=disable')
    conn = await client.get_conn()
    await conn.exec(
        """
        CREATE TABLE test (
            i64 Int64,
            u64 UInt64,
            f64 Float64,
            s   String,
            s2  String,
            d   Date,
            t   DateTime
        )
        """
    )
    rows = await conn.query_iter("SELECT * FROM test")
    async for row in rows:
        print(row.values())
    await conn.close()

asyncio.run(main())

Parameter bindings

# Positional parameters using ?
row = await context.conn.query_row("SELECT ?, ?, ?, ?", (3, False, 4, "55"))

# Named parameters using :name
row = await context.conn.query_row(
    "SELECT :a, :b, :c, :d", {"a": 3, "b": False, "c": 4, "d": "55"}
)

# Single value (no tuple needed)
row = await context.conn.query_row("SELECT ?", 3)

# Keyword argument form
row = await context.conn.query_row("SELECT ?, ?, ?, ?", params=(3, False, 4, "55"))

Named parameters use token-aware matching, so :a will not corrupt :ab. For local embedded connections, passing a mismatched number of ? placeholders and values raises a ValueError immediately.

Query ID tracking and query management

# Get the last executed query ID
query_id = conn.last_query_id()
print(f"Last query ID: {query_id}")

# Execute a query and get its ID
await conn.query_row("SELECT 1")
query_id = conn.last_query_id()
print(f"Query ID: {query_id}")

# Kill a running query (if needed)
try:
    await conn.kill_query("some-query-id")
    print("Query killed successfully")
except Exception as e:
    print(f"Failed to kill query: {e}")

Type Mapping

Databend Types

General Data Types

Databend Python
BOOLEAN bool
TINYINT int
SMALLINT int
INT int
BIGINT int
FLOAT float
DOUBLE float
DECIMAL decimal.Decimal
DATE datetime.date
TIMESTAMP datetime.datetime
INTERVAL datetime.timedelta
VARCHAR str
BINARY bytes

Semi-Structured Data Types

Databend Python
ARRAY list
TUPLE tuple
MAP dict
VARIANT str
BITMAP str
GEOMETRY str / bytes
GEOGRAPHY str / bytes

Note: VARIANT is a json encoded string. Example:

CREATE TABLE example (
    data VARIANT
);
INSERT INTO example VALUES ('{"a": 1, "b": "hello"}');
row = await conn.query_row("SELECT * FROM example limit 1;")
data = row.values()[0]
value = json.loads(data)
print(value)

GEOMETRY and GEOGRAPHY follow the current geometry_output_format setting. Text formats such as GeoJSON or WKT return str; binary formats such as WKB or EWKB return bytes.

For example:

row = await conn.query_row("settings(geometry_output_format='WKB') SELECT st_point(60, 37)")
assert isinstance(row.values()[0], bytes)

APIs

Exception Classes (PEP 249 Compliant)

The driver provides a complete set of exception classes that follow the PEP 249 standard for database interfaces:

# Base exceptions
class Warning(Exception): ...
class Error(Exception): ...

# Interface errors
class InterfaceError(Error): ...

# Database errors
class DatabaseError(Error): ...

# Specific database error types
class DataError(DatabaseError): ...
class OperationalError(DatabaseError): ...
class IntegrityError(DatabaseError): ...
class InternalError(DatabaseError): ...
class ProgrammingError(DatabaseError): ...
class NotSupportedError(DatabaseError): ...

These exceptions are automatically mapped from Databend error codes to appropriate PEP 249 exception types based on the nature of the error.

Note: stream_load and load_file support an optional method parameter, it accepts two string values:

  • stage: Data is first uploaded to a temporary stage and then loaded. This is the default behavior.
  • streaming: Data is directly streamed to the Databend server.

AsyncDatabendClient

class AsyncDatabendClient:
    def __init__(self, dsn: str): ...
    async def get_conn(self) -> AsyncDatabendConnection: ...

AsyncDatabendConnection

class AsyncDatabendConnection:
    async def info(self) -> ConnectionInfo: ...
    async def version(self) -> str: ...
    async def close(self) -> None: ...
    def last_query_id(self) -> str | None: ...
    async def kill_query(self, query_id: str) -> None: ...
    async def exec(self, sql: str, params: list[string] | tuple[string] | any = None) -> int: ...
    async def query_row(self, sql: str, params: list[string] | tuple[string] | any = None) -> Row: ...
    async def query_iter(self, sql: str, params: list[string] | tuple[string] | any = None) -> RowIterator: ...
    async def stream_load(self, sql: str, data: list[list[str]], method: str = None) -> ServerStats: ...
    async def load_file(self, sql: str, file: str, method: str = None) -> ServerStats: ...

BlockingDatabendClient

class BlockingDatabendClient:
    def __init__(self, dsn: str): ...
    def get_conn(self) -> BlockingDatabendConnection: ...
    def cursor(self) -> BlockingDatabendCursor: ...

BlockingDatabendConnection

class BlockingDatabendConnection:
    def info(self) -> ConnectionInfo: ...
    def version(self) -> str: ...
    def close(self) -> None: ...
    def last_query_id(self) -> str | None: ...
    def kill_query(self, query_id: str) -> None: ...
    def exec(self, sql: str, params: list[string] | tuple[string] | any = None) -> int: ...
    def query_row(self, sql: str, params: list[string] | tuple[string] | any = None) -> Row: ...
    def query_iter(self, sql: str, params: list[string] | tuple[string] | any = None) -> RowIterator: ...
    def stream_load(self, sql: str, data: list[list[str]], method: str = None) -> ServerStats: ...
    def load_file(self, sql: str, file: str, method: str = None, format_option: dict = None, copy_options: dict = None) -> ServerStats: ...

BlockingDatabendCursor

class BlockingDatabendCursor:
    @property
    def description(self) -> list[tuple[str, str, int | None, int | None, int | None, int | None, bool | None]] | None: ...
    @property
    def rowcount(self) -> int: ...
    def close(self) -> None: ...
    def execute(self, operation: str, params: list[string] | tuple[string] = None) -> None | int: ...
    def executemany(self, operation: str, params: list[string] | tuple[string] = None, values: list[list[string] | tuple[string]]) -> None | int: ...
    def fetchone(self) -> Row | None: ...
    def fetchmany(self, size: int = 1) -> list[Row]: ...
    def fetchall(self) -> list[Row]: ...

    # Optional DB API Extensions
    def next(self) -> Row: ...
    def __next__(self) -> Row: ...
    def __iter__(self) -> BlockingDatabendCursor: ...

Row

class Row:
    def values(self) -> tuple: ...
    def __len__(self) -> int: ...
    def __iter__(self) -> Row: ...
    def __next__(self) -> any: ...
    def __dict__(self) -> dict: ...
    def __getitem__(self, key: int | str) -> any: ...

RowIterator

class RowIterator:
    def schema(self) -> Schema: ...

    def __iter__(self) -> RowIterator: ...
    def __next__(self) -> Row: ...

    def __aiter__(self) -> RowIterator: ...
    async def __anext__(self) -> Row: ...

Field

class Field:
    @property
    def name(self) -> str: ...
    @property
    def data_type(self) -> str: ...

Schema

class Schema:
    def fields(self) -> list[Field]: ...

ServerStats

class ServerStats:
    @property
    def total_rows(self) -> int: ...
    @property
    def total_bytes(self) -> int: ...
    @property
    def read_rows(self) -> int: ...
    @property
    def read_bytes(self) -> int: ...
    @property
    def write_rows(self) -> int: ...
    @property
    def write_bytes(self) -> int: ...
    @property
    def running_time_ms(self) -> float: ...

ConnectionInfo

class ConnectionInfo:
    @property
    def handler(self) -> str: ...
    @property
    def host(self) -> str: ...
    @property
    def port(self) -> int: ...
    @property
    def user(self) -> str: ...
    @property
    def database(self) -> str | None: ...
    @property
    def warehouse(self) -> str | None: ...

connect_local

def connect_local(
    database: str = ":memory:",
    *,
    data_path: str | None = None,
    tenant: str | None = None,
) -> LocalConnection: ...

LocalConnection

class LocalConnection:
    def sql(self, query: str) -> LocalRelation: ...
    def table(self, name: str) -> LocalRelation: ...
    def format_sql(self, sql: str, params: Any = None) -> str: ...
    def execute(self, query: str, params: Any = None) -> None: ...
    def exec(self, sql: str, params: Any = None) -> None: ...
    def query_row(self, sql: str, params: Any = None) -> LocalRow | None: ...
    def query_all(self, sql: str, params: Any = None) -> list[LocalRow]: ...
    def query_iter(self, sql: str, params: Any = None) -> LocalRowIterator: ...
    def close(self) -> None: ...
    def last_query_id(self) -> None: ...  # always None for local mode
    def kill_query(self, query_id: str) -> None: ...  # raises NotImplementedError
    def register(
        self,
        name: str,
        source: Any,           # path str/Path, pandas/polars DataFrame, or pyarrow Table
        *,
        format: str | None = None,
        pattern: str | None = None,
        connection: str | None = None,
    ) -> LocalConnection: ...
    def from_df(self, source: Any, *, name: str | None = None) -> LocalRelation: ...
    def read_parquet(
        self, path: str | Path, *, pattern: str | None = None,
        connection: str | None = None, name: str | None = None,
    ) -> LocalRelation: ...
    def read_csv(
        self, path: str | Path, *, pattern: str | None = None,
        connection: str | None = None, name: str | None = None,
    ) -> LocalRelation: ...
    def read_json(
        self, path: str | Path, *, pattern: str | None = None,
        connection: str | None = None, name: str | None = None,
    ) -> LocalRelation: ...
    def read_text(
        self, path: str | Path, *, pattern: str | None = None,
        connection: str | None = None, name: str | None = None,
    ) -> LocalRelation: ...

LocalRelation

class LocalRelation:
    def df(self) -> Any: ...          # pandas DataFrame
    def pl(self) -> Any: ...          # polars DataFrame
    def arrow(self) -> Any: ...       # pyarrow Table
    def fetchall(self) -> list[tuple]: ...
    def fetchone(self) -> tuple | None: ...

LocalRow

class LocalRow:
    def values(self) -> tuple[Any, ...]: ...
    def __len__(self) -> int: ...
    def __iter__(self) -> LocalRow: ...
    def __next__(self) -> Any: ...
    def __getitem__(self, key: int) -> Any: ...

LocalRowIterator

class LocalRowIterator:
    def schema(self) -> Any: ...  # not yet implemented for local mode
    def close(self) -> None: ...
    def __iter__(self) -> LocalRowIterator: ...
    def __next__(self) -> LocalRow: ...

Development

cd tests
make up
cd bindings/python
uv python install 3.12
uv venv --python 3.12
uv sync --extra local
source .venv/bin/activate
maturin develop

behave tests/asyncio
behave tests/blocking
behave tests/cursor
behave tests/local

Release history Release notifications | RSS feed

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distributions

No source distribution files available for this release.See tutorial on generating distribution archives.

Built Distributions

If you're not sure about the file name format, learn more about wheel file names.

databend_driver-0.34.2-cp39-abi3-manylinux_2_28_aarch64.whl (8.7 MB view details)

Uploaded CPython 3.9+manylinux: glibc 2.28+ ARM64

databend_driver-0.34.2-cp39-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (9.3 MB view details)

Uploaded CPython 3.9+manylinux: glibc 2.17+ x86-64

databend_driver-0.34.2-cp39-abi3-macosx_11_0_arm64.whl (8.3 MB view details)

Uploaded CPython 3.9+macOS 11.0+ ARM64

databend_driver-0.34.2-cp39-abi3-macosx_10_12_x86_64.whl (8.9 MB view details)

Uploaded CPython 3.9+macOS 10.12+ x86-64

databend_driver-0.34.2-cp38-cp38-manylinux_2_28_aarch64.whl (8.9 MB view details)

Uploaded CPython 3.8manylinux: glibc 2.28+ ARM64

databend_driver-0.34.2-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (9.5 MB view details)

Uploaded CPython 3.8manylinux: glibc 2.17+ x86-64

databend_driver-0.34.2-cp38-cp38-macosx_11_0_arm64.whl (8.4 MB view details)

Uploaded CPython 3.8macOS 11.0+ ARM64

databend_driver-0.34.2-cp38-cp38-macosx_10_12_x86_64.whl (9.0 MB view details)

Uploaded CPython 3.8macOS 10.12+ x86-64

File details

Details for the file databend_driver-0.34.2-cp39-abi3-manylinux_2_28_aarch64.whl.

File metadata

File hashes

Hashes for databend_driver-0.34.2-cp39-abi3-manylinux_2_28_aarch64.whl
Algorithm Hash digest
SHA256 18197d37bd5e976fe7aa8c064b8a9ce7ab81e4515568f383f1bd12b1e595f787
MD5 72e0c46b8ee1fdb0651bc450e081a45d
BLAKE2b-256 921dc68eceb1e9fea8a285664045cf9a3139839109ed8a89b5b0321d28adc7b3

See more details on using hashes here.

Provenance

The following attestation bundles were made for databend_driver-0.34.2-cp39-abi3-manylinux_2_28_aarch64.whl:

Publisher: release.yml on databendlabs/bendsql

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file databend_driver-0.34.2-cp39-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.

File metadata

File hashes

Hashes for databend_driver-0.34.2-cp39-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
Algorithm Hash digest
SHA256 4f918bad08ab985473b622c16eaa9bf51a4861dfe8912757ef6cfb347365242c
MD5 7fca70a9b6a3958c63e0d308190fc3c4
BLAKE2b-256 72dab82ae6f161cfeb913b1bae0406e425d5be85cc66b777f17e92bacdc2e9b9

See more details on using hashes here.

Provenance

The following attestation bundles were made for databend_driver-0.34.2-cp39-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl:

Publisher: release.yml on databendlabs/bendsql

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file databend_driver-0.34.2-cp39-abi3-macosx_11_0_arm64.whl.

File metadata

File hashes

Hashes for databend_driver-0.34.2-cp39-abi3-macosx_11_0_arm64.whl
Algorithm Hash digest
SHA256 b6cf832e35e9e273e35d12b872b31f6b5ff7185e2a562454147526c9a86c14a9
MD5 b345f611c8fac660e88976043b5218d7
BLAKE2b-256 814d992c5401ca87f5c501cd43958e9922c98bce0767328971aaecfc1b3ee980

See more details on using hashes here.

Provenance

The following attestation bundles were made for databend_driver-0.34.2-cp39-abi3-macosx_11_0_arm64.whl:

Publisher: release.yml on databendlabs/bendsql

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file databend_driver-0.34.2-cp39-abi3-macosx_10_12_x86_64.whl.

File metadata

File hashes

Hashes for databend_driver-0.34.2-cp39-abi3-macosx_10_12_x86_64.whl
Algorithm Hash digest
SHA256 9607aa13226f4f437c69f6c3c1a34a094909ea5bc5a27ddf0538cdb2e7805ece
MD5 d706e30bd1b97934eee4c5cda260c312
BLAKE2b-256 e97a8f6905a6a608e870227408cae922b69ee3e156c8a90cbe4159d59798fee4

See more details on using hashes here.

Provenance

The following attestation bundles were made for databend_driver-0.34.2-cp39-abi3-macosx_10_12_x86_64.whl:

Publisher: release.yml on databendlabs/bendsql

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file databend_driver-0.34.2-cp38-cp38-manylinux_2_28_aarch64.whl.

File metadata

File hashes

Hashes for databend_driver-0.34.2-cp38-cp38-manylinux_2_28_aarch64.whl
Algorithm Hash digest
SHA256 ca6b4ccad9b930554cf157170e3bc4549c0a5bae3f8fad3a695fe3a4d2cd2ec9
MD5 97810636e437ecad54db4f83e3e90dfe
BLAKE2b-256 3a2273b9c74490cc28bba8a373fcb8b8a248dbfe3e7f6b0b18de2d48c7026d53

See more details on using hashes here.

Provenance

The following attestation bundles were made for databend_driver-0.34.2-cp38-cp38-manylinux_2_28_aarch64.whl:

Publisher: release.yml on databendlabs/bendsql

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file databend_driver-0.34.2-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.

File metadata

File hashes

Hashes for databend_driver-0.34.2-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
Algorithm Hash digest
SHA256 17f44719450fe4d614a7be8b9cf60207ed18fba2296f2e125a545674e6d93343
MD5 00032a263be082ecf7cc04273b6eb006
BLAKE2b-256 e7f18dd423829be83580b029e1983b3f7b83fb027728994574b5419d5711ba07

See more details on using hashes here.

Provenance

The following attestation bundles were made for databend_driver-0.34.2-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl:

Publisher: release.yml on databendlabs/bendsql

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file databend_driver-0.34.2-cp38-cp38-macosx_11_0_arm64.whl.

File metadata

File hashes

Hashes for databend_driver-0.34.2-cp38-cp38-macosx_11_0_arm64.whl
Algorithm Hash digest
SHA256 c6ecdab29cb70803eb65f8f1af8f1080788b43e334bc5550fcdb1761fbb31604
MD5 5945e63d7f56de45064738fdd736f95e
BLAKE2b-256 0544ecf9b7a48cf8a5aa826b78f6d5f3916c016368cf8f252a0a56457b852db8

See more details on using hashes here.

Provenance

The following attestation bundles were made for databend_driver-0.34.2-cp38-cp38-macosx_11_0_arm64.whl:

Publisher: release.yml on databendlabs/bendsql

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file databend_driver-0.34.2-cp38-cp38-macosx_10_12_x86_64.whl.

File metadata

File hashes

Hashes for databend_driver-0.34.2-cp38-cp38-macosx_10_12_x86_64.whl
Algorithm Hash digest
SHA256 6feb4fab3c7f89ddd563abad840d3c25dd7cfbab43c00b281ffbd476fb935daf
MD5 f87a42b344d0151e23557af48790341d
BLAKE2b-256 680923fcabf232617008bd5b92a0d65e96eebcd55f216c35588f3760a27f83f1

See more details on using hashes here.

Provenance

The following attestation bundles were made for databend_driver-0.34.2-cp38-cp38-macosx_10_12_x86_64.whl:

Publisher: release.yml on databendlabs/bendsql

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page