native-dumper 0.3.4.4

Library for read and write Native format between Clickhouse and file.

NativeDumper

Library for read and write Native format between Clickhouse and file

Examples

Initialization

from native_dumper import (
    CompressionMethod,
    CHConnector,
    NativeDumper,
)

connector = CHConnector(
    host = <your host>,
    dbname = <your database>,
    user = <your username>,
    password = <your password>,
    port = 8123,
)

dumper = NativeDumper(
    connector=connector,
    compression_method=CompressionMethod.ZSTD,  # or CompressionMethod.LZ4 or CompressionMethod.NONE
)

Read dump from Clickhouse into file

file_name = "native.zstd"
# you need define one of parameter query or table_name
query = "select ..."  # some sql query
table_name = "default.test_table"  # or some table

with open(file_name, "wb") as fileobj:
    dumper.read_dump(
        fileobj,
        query,
        table_name,
    )

Write dump from file into Clickhouse

file_name = "native.zstd"
# you need define one of parameter table_name
table_name = "default.test_table"  # some table

with open(file_name, "rb") as fileobj:
    dumper.write_dump(
        fileobj,
        table_name,
    )

Write from Clickhouse into Clickhouse

Same server

table_dest = "default.test_table_write"  # some table for write
table_src = "default.test_table_read"  # some table for read
query_src = "select ..."  # or some sql query for read

dumper.write_between(
    table_dest,
    table_src,
    query_src,
)

Different servers

connector_src = CHConnector(
    host = <host src>,
    dbname = <database src>,
    user = <username src>,
    password = <password src>,
    port = 8123,
)

dumper_src = NativeDumper(connector=connector_src)

table_dest = "default.test_table_write"  # some table for write
table_src = "default.test_table_read"  # some table for read
query_src = "select ..."  # or some sql query for read

dumper.write_between(
    table_dest,
    table_src,
    query_src,
    dumper_src.cursor,
)

Get NativeReader object from stream

table_name = "default.test_table_read"  # some table for read
query = "select ..."  # or some sql query for read

reader = dumper.to_reader(
    query=query,
    table_name=table_name,
)

NativeReader has three methods available, but only one of the methods is available at a time within a single session.

# read as python generator object
reader.to_rows()
# or read as pandas.DataFrame
reader.to_pandas()
# or read as polars.DataFrame
reader.to_polars()

Write from python objects into target table

# some table for write data
table_name = "default.test_table_write"
dtype_data: Itarable[Any]
pandas_frame: pandas.DataFrame
polars_frame: polars.DataFrame

# write from python object
dumper.from_rows(dtype_data, table_name)
# write from pandas.DataFrame
dumper.from_pandas(pandas_frame, table_name)
# write from polars.DataFrame
dumper.from_polars(polars_frame, table_name)

Open Native file format

Get info from my another repository https://github.com/0xMihalich/nativelib

Installation

From pip

pip install native-dumper

From local directory

pip install .

From git

pip install git+https://github.com/0xMihalich/native_dumper