kacpermuda commented on code in PR #47168:
URL: https://github.com/apache/airflow/pull/47168#discussion_r1979203842
##########
providers/snowflake/src/airflow/providers/snowflake/transfers/copy_into_snowflake.py:
##########
@@ -175,53 +175,79 @@ def _extract_openlineage_unique_dataset_paths(
>>> results = [{"file":
"azure://my_account.blob.core.windows.net/azure_container/dir3/file.csv"}]
>>> method(results)
- ([('wasbs://azure_container@my_account', 'dir3')], [])
+ ([('wasbs://azure_container@my_account', 'dir3/file.csv')], [])
>>> results = [{"file":
"azure://my_account.blob.core.windows.net/azure_container"}]
>>> method(results)
([('wasbs://azure_container@my_account', '/')], [])
>>> results = [{"file": "s3://bucket"}, {"file": "gcs://bucket/"},
{"file": "s3://bucket/a.csv"}]
>>> method(results)
- ([('gcs://bucket', '/'), ('s3://bucket', '/')], [])
+ ([('gcs://bucket', '/'), ('s3://bucket', '/'), ('s3://bucket',
'a.csv')], [])
>>> results = [{"file": "s3://bucket/dir/file.csv"}, {"file":
"gcs://bucket/dir/dir2/a.txt"}]
>>> method(results)
- ([('gcs://bucket', 'dir/dir2'), ('s3://bucket', 'dir')], [])
+ ([('gcs://bucket', 'dir/dir2/a.txt'), ('s3://bucket',
'dir/file.csv')], [])
>>> results = [
... {"file": "s3://bucket/dir/file.csv"},
... {"file":
"azure://my_account.something_new.windows.net/azure_container"},
... ]
>>> method(results)
- ([('s3://bucket', 'dir')],
['azure://my_account.something_new.windows.net/azure_container'])
+ ([('s3://bucket', 'dir/file.csv')],
['azure://my_account.something_new.windows.net/azure_container'])
+
+ >>> results = [
+ ... {"file": "s3://bucket/dir/file.csv"},
+ ... {"file": "s3:/invalid-s3-uri"},
+ ... {"file": "gcs:invalid-gcs-uri"},
+ ... ]
+ >>> method(results)
+ ([('s3://bucket', 'dir/file.csv')], ['gcs:invalid-gcs-uri',
's3:/invalid-s3-uri'])
"""
import re
- from pathlib import Path
from urllib.parse import urlparse
azure_regex = r"azure:\/\/(\w+)?\.blob.core.windows.net\/(\w+)\/?(.*)?"
extraction_error_files = []
unique_dataset_paths = set()
for row in query_result:
- uri = urlparse(row["file"])
- if uri.scheme == "azure":
- match = re.fullmatch(azure_regex, row["file"])
- if not match:
- extraction_error_files.append(row["file"])
- continue
- account_name, container_name, name = match.groups()
- namespace = f"wasbs://{container_name}@{account_name}"
- else:
- namespace = f"{uri.scheme}://{uri.netloc}"
- name = uri.path.lstrip("/")
-
- name = Path(name).parent.as_posix()
- if name in ("", "."):
- name = "/"
-
- unique_dataset_paths.add((namespace, name))
+ try:
+ uri = urlparse(row["file"])
+
+ # Check for valid URI structure
+ if not uri.scheme or not uri.netloc:
+ # Handle invalid URIs
+ if uri.scheme == "azure" and "azure://" in row["file"]:
+ # Try to process Azure URI
+ match = re.fullmatch(azure_regex, row["file"])
+ if not match:
+ extraction_error_files.append(row["file"])
+ continue
+ account_name, container_name, name = match.groups()
+ namespace = f"wasbs://{container_name}@{account_name}"
Review Comment:
Not sure if this check makes sense. If there is no scheme or netloc, then
how it can be a valid azure uri? We should probably add this to extraction
errors straight away, correct? Do you have a case that falls into this if ?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
