scan directly
This commit is contained in:
@@ -1,3 +1,4 @@
|
|||||||
|
import os
|
||||||
from collections.abc import Iterator
|
from collections.abc import Iterator
|
||||||
from datetime import datetime
|
from datetime import datetime
|
||||||
from functools import partial
|
from functools import partial
|
||||||
@@ -13,7 +14,7 @@ from models import Deal
|
|||||||
from partitions import daily_partitions_def, multi_partitions_def
|
from partitions import daily_partitions_def, multi_partitions_def
|
||||||
from plato.fetch import scrape_plato
|
from plato.fetch import scrape_plato
|
||||||
from plato.parse import parse as parse_plato
|
from plato.parse import parse as parse_plato
|
||||||
from shared.utils import get_partition_keys, load_partitions, parse_partition_keys
|
from shared.utils import get_partition_keys, parse_partition_keys
|
||||||
from sounds.fetch import fetch_deals
|
from sounds.fetch import fetch_deals
|
||||||
from sounds.parse import parse as parse_sounds
|
from sounds.parse import parse as parse_sounds
|
||||||
from utils.email import EmailService
|
from utils.email import EmailService
|
||||||
@@ -134,11 +135,13 @@ def works(context: dg.AssetExecutionContext) -> Iterator[dg.Output[pl.DataFrame]
|
|||||||
partitions = context.instance.get_materialized_partitions(cleaned_deals.key)
|
partitions = context.instance.get_materialized_partitions(cleaned_deals.key)
|
||||||
ic(partitions)
|
ic(partitions)
|
||||||
logger.info("Works", partitions=partitions)
|
logger.info("Works", partitions=partitions)
|
||||||
dfs = list(load_partitions(context, cleaned_deals.key, partitions))
|
path = os.path.join(
|
||||||
ic(len(dfs))
|
context.resources.polars_parquet_io_manager.base_dir, *cleaned_deals.key.path
|
||||||
if dfs:
|
)
|
||||||
columns = ["artist", "title", "release"]
|
ic(path)
|
||||||
yield dg.Output(pl.concat(dfs, how="vertical_relaxed").select(columns).unique())
|
yield dg.Output(
|
||||||
|
pl.scan_parquet(path).select(["artist", "title", "release"]).unique().collect()
|
||||||
|
)
|
||||||
|
|
||||||
|
|
||||||
@asset(
|
@asset(
|
||||||
|
|||||||
Reference in New Issue
Block a user