yujunjun/OpenMetadata
1
0
Fork 0
mirror of https://github.com/open-metadata/OpenMetadata.git synced 2025-10-06 14:26:28 +00:00
Code Issues Packages Projects Releases Wiki Activity

MINOR: row sampling error (#21488)

Browse Source 
* fix: row sampling error

* fix: return sample query

(cherry picked from commit 859f24aba7c2b0bcbaf85b149b2a02d60fa4e201)
This commit is contained in:
Teddy 2025-06-02 09:02:17 +02:00 committed by Teddy Crepineau
parent 38724bf2fe
commit 68b0eb34b7
4 changed files with 10 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
ingestion/src/metadata/sampler/sqlalchemy/azuresql/sampler.py
View File

@ -15,7 +15,7 @@ for the profiler
from typing import List, Optional from typing import List, Optional
from sqlalchemy import Column, Table, text from sqlalchemy import Column, Table, text
from sqlalchemy.orm import Query from sqlalchemy.sql.selectable import CTE
from metadata.generated.schema.entity.data.table import TableData, TableType from metadata.generated.schema.entity.data.table import TableData, TableType
from metadata.sampler.sqlalchemy.sampler import ProfileSampleType, SQASampler from metadata.sampler.sqlalchemy.sampler import ProfileSampleType, SQASampler
@ -49,13 +49,12 @@ class AzureSQLSampler(SQASampler):
return selectable return selectable
def get_sample_query(self, *, column=None) -> Query: def get_sample_query(self, *, column=None) -> CTE:
"""get query for sample data""" """Override the base method as ROWS or PERCENT sampling handled through the tablesample clause"""
rnd = self._base_sample_query(column).cte( rnd = self._base_sample_query(column).cte(
f"{self.get_sampler_table_name()}_rnd" f"{self.get_sampler_table_name()}_rnd"
) )
with self.get_client() as client: query = self.get_client().query(rnd)
query = client.query(rnd)
return query.cte(f"{self.get_sampler_table_name()}_sample") return query.cte(f"{self.get_sampler_table_name()}_sample")
def fetch_sample_data(self, columns: Optional[List[Column]] = None) -> TableData: def fetch_sample_data(self, columns: Optional[List[Column]] = None) -> TableData:

5
ingestion/src/metadata/sampler/sqlalchemy/mssql/sampler.py
View File

@ -44,10 +44,9 @@ class MssqlSampler(SQASampler):
return selectable return selectable
def get_sample_query(self, *, column=None) -> CTE: def get_sample_query(self, *, column=None) -> CTE:
"""get query for sample data""" """Override the base method as ROWS or PERCENT sampling handled through the tablesample clause"""
rnd = self._base_sample_query(column).cte( rnd = self._base_sample_query(column).cte(
f"{self.get_sampler_table_name()}_rnd" f"{self.get_sampler_table_name()}_rnd"
) )
with self.get_client() as client: query = self.get_client().query(rnd)
query = client.query(rnd)
return query.cte(f"{self.get_sampler_table_name()}_sample") return query.cte(f"{self.get_sampler_table_name()}_sample")

2
ingestion/src/metadata/sampler/sqlalchemy/sampler.py
View File

@ -136,6 +136,8 @@ class SQASampler(SamplerInterface, SQAInterfaceMixin):
).cte(f"{self.get_sampler_table_name()}_sample") ).cte(f"{self.get_sampler_table_name()}_sample")
table_query = client.query(self.raw_dataset) table_query = client.query(self.raw_dataset)
if self.partition_details:
table_query = self.get_partitioned_query(table_query)
session_query = self._base_sample_query( session_query = self._base_sample_query(
column, column,
(ModuloFn(RandomNumFn(), table_query.count())).label(RANDOM_LABEL) (ModuloFn(RandomNumFn(), table_query.count())).label(RANDOM_LABEL)

5
ingestion/src/metadata/sampler/sqlalchemy/snowflake/sampler.py
View File

@ -87,10 +87,9 @@ class SnowflakeSampler(SQASampler):
) )
def get_sample_query(self, *, column=None) -> CTE: def get_sample_query(self, *, column=None) -> CTE:
"""get query for sample data""" """Override the base method as ROWS or PERCENT sampling handled through the tablesample clause"""
rnd = self._base_sample_query(column).cte( rnd = self._base_sample_query(column).cte(
f"{self.get_sampler_table_name()}_rnd" f"{self.get_sampler_table_name()}_rnd"
) )
with self.get_client() as client: query = self.get_client().query(rnd)
query = client.query(rnd)
return query.cte(f"{self.get_sampler_table_name()}_sample") return query.cte(f"{self.get_sampler_table_name()}_sample")