mirror of
https://github.com/deepseek-ai/smallpond
synced 2025-06-26 18:27:45 +00:00
reformat code with --line-length=150 (#18)
This commit is contained in:
@@ -5,12 +5,8 @@ import smallpond
|
||||
from smallpond.dataframe import Session
|
||||
|
||||
|
||||
def sort_mock_urls_v2(
|
||||
sp: Session, input_paths: List[str], output_path: str, npartitions: int
|
||||
):
|
||||
dataset = sp.read_csv(
|
||||
input_paths, schema={"urlstr": "varchar", "valstr": "varchar"}, delim=r"\t"
|
||||
).repartition(npartitions)
|
||||
def sort_mock_urls_v2(sp: Session, input_paths: List[str], output_path: str, npartitions: int):
|
||||
dataset = sp.read_csv(input_paths, schema={"urlstr": "varchar", "valstr": "varchar"}, delim=r"\t").repartition(npartitions)
|
||||
urls = dataset.map(
|
||||
"""
|
||||
split_part(urlstr, '/', 1) as host,
|
||||
@@ -25,9 +21,7 @@ def sort_mock_urls_v2(
|
||||
|
||||
if __name__ == "__main__":
|
||||
parser = argparse.ArgumentParser()
|
||||
parser.add_argument(
|
||||
"-i", "--input_paths", nargs="+", default=["tests/data/mock_urls/*.tsv"]
|
||||
)
|
||||
parser.add_argument("-i", "--input_paths", nargs="+", default=["tests/data/mock_urls/*.tsv"])
|
||||
parser.add_argument("-o", "--output_path", type=str, default="sort_mock_urls")
|
||||
parser.add_argument("-n", "--npartitions", type=int, default=10)
|
||||
args = parser.parse_args()
|
||||
|
||||
Reference in New Issue
Block a user