mirror of
https://github.com/hanzo-ds/connect.git
synced 2026-10-30 20:31:24 +00:00
zeekay
999d38466d
Imports read `datastore_connect`, distribution `datastore-connect`, so nothing has to import another product's name to talk to Hanzo Datastore. Cut from upstream 0.8.17 — the version this stack already ran — rather than from upstream HEAD, so it is a PURE RENAME with no behaviour change. Forking from HEAD would have dragged versions along with the name (connect alone jumped 0.8.17 -> 1.6.0, a major), which is a different and much riskier change wearing a rename's clothes. X-ClickHouse-* headers are untouched: they are wire protocol the server sends and expects, and a different token shape from the module name.
35 lines
1.7 KiB
Python
35 lines
1.7 KiB
Python
from pathlib import Path
|
|
from typing import Callable
|
|
|
|
from datastore_connect.driver import Client
|
|
|
|
|
|
def test_raw_insert(test_client: Client, table_context: Callable):
|
|
with table_context('test_raw_insert', ["`weir'd` String", 'value String']):
|
|
csv = 'value1\nvalue2'
|
|
test_client.raw_insert('test_raw_insert', ['"weir\'d"'], csv.encode(), fmt='CSV')
|
|
result = test_client.query('SELECT * FROM test_raw_insert')
|
|
assert result.result_set[1][0] == 'value2'
|
|
|
|
test_client.command('TRUNCATE TABLE test_raw_insert')
|
|
tsv = 'weird1\tvalue__`2\nweird2\tvalue77'
|
|
test_client.raw_insert('test_raw_insert', ["`weir'd`", 'value'], tsv, fmt='TSV')
|
|
result = test_client.query('SELECT * FROM test_raw_insert')
|
|
assert result.result_set[0][1] == 'value__`2'
|
|
assert result.result_set[1][1] == 'value77'
|
|
|
|
|
|
def test_raw_insert_compression(test_client: Client, table_context: Callable):
|
|
data_file = f'{Path(__file__).parent}/movies.csv.gz'
|
|
with open(data_file, mode='rb') as movies_file:
|
|
data = movies_file.read()
|
|
with table_context('test_gzip_movies', ['movie String', 'year UInt16', 'rating Decimal32(3)']):
|
|
test_client.raw_insert('test_gzip_movies', None, data, fmt='CSV', compression='gzip',
|
|
settings={'input_format_allow_errors_ratio': .2,
|
|
'input_format_allow_errors_num': 5}
|
|
)
|
|
res = test_client.query(
|
|
'SELECT count() as count, sum(rating) as rating, max(year) as year FROM test_gzip_movies').first_item
|
|
assert res['count'] == 248
|
|
assert res['year'] == 2022
|