rusackas commented on code in PR #37120:
URL: https://github.com/apache/superset/pull/37120#discussion_r3653911295
##########
superset/commands/database/export.py:
##########
@@ -104,8 +104,12 @@ def _file_content(model: Database) -> str:
@staticmethod
def _export(
- model: Database, export_related: bool = True
+ model: Database, export_related: bool = True, seen: set[str] | None =
None
) -> Iterator[tuple[str, Callable[[], str]]]:
+ # Initialize seen set if not provided
+ if seen is None:
+ seen = set()
Review Comment:
Not dead code, `_export` gets called directly (bypassing `run()`) in a few
unit tests without passing `seen`, so the None-check is actually exercised.
##########
superset/commands/dataset/export.py:
##########
@@ -103,29 +107,33 @@ def _export(
)
file_path = f"databases/{db_file_name}.yaml"
- payload = model.database.export_to_dict(
- recursive=False,
- include_parent_ref=False,
- include_defaults=True,
- export_uuids=True,
- )
- # TODO (betodealmeida): move this logic to export_to_dict once this
- # becomes the default export endpoint
- if payload.get("extra"):
- try:
- payload["extra"] = json.loads(payload["extra"])
- except json.JSONDecodeError:
- logger.info("Unable to decode `extra` field: %s",
payload["extra"])
-
- if ssh_tunnel := model.database.ssh_tunnel:
- ssh_tunnel_payload = ssh_tunnel.export_to_dict(
+ # Only yield the database file if not already seen
+ # This is critical to fix the issue where databases were being
duplicated
+ # and potentially overwritten when charts from different databases
were exported
+ if file_path not in seen:
+ payload = model.database.export_to_dict(
recursive=False,
include_parent_ref=False,
include_defaults=True,
- export_uuids=False,
+ export_uuids=True,
)
- payload["ssh_tunnel"] = mask_password_info(ssh_tunnel_payload)
+ # TODO (betodealmeida): move this logic to export_to_dict once
this
+ # becomes the default export endpoint
+ if payload.get("extra"):
+ try:
+ payload["extra"] = json.loads(payload["extra"])
+ except json.JSONDecodeError:
+ logger.info("Unable to decode `extra` field: %s",
payload["extra"])
+
+ if ssh_tunnel := model.database.ssh_tunnel:
+ ssh_tunnel_payload = ssh_tunnel.export_to_dict(
+ recursive=False,
+ include_parent_ref=False,
+ include_defaults=True,
+ export_uuids=False,
+ )
+ payload["ssh_tunnel"] =
mask_password_info(ssh_tunnel_payload)
- payload["version"] = EXPORT_VERSION
+ payload["version"] = EXPORT_VERSION
- yield file_path, lambda: yaml.safe_dump(payload, sort_keys=False)
+ yield file_path, lambda: yaml.safe_dump(payload,
sort_keys=False)
Review Comment:
Same as the CodeAnt comment above on this file, it's a single yield outside
a loop so there's no late-binding bug here.
##########
superset/commands/query/export.py:
##########
@@ -79,21 +83,23 @@ def _export(
database_slug = secure_filename(model.database.database_name)
file_name = f"databases/{database_slug}.yaml"
- payload = model.database.export_to_dict(
- recursive=False,
- include_parent_ref=False,
- include_defaults=True,
- export_uuids=True,
- )
- # TODO (betodealmeida): move this logic to export_to_dict once this
- # becomes the default export endpoint
- if "extra" in payload:
- try:
- payload["extra"] = json.loads(payload["extra"])
- except json.JSONDecodeError:
- logger.info("Unable to decode `extra` field: %s",
payload["extra"])
+ # Only yield if not already seen (similar to dataset export)
+ if file_name not in seen:
+ payload = model.database.export_to_dict(
+ recursive=False,
+ include_parent_ref=False,
+ include_defaults=True,
+ export_uuids=True,
+ )
+ # TODO (betodealmeida): move this logic to export_to_dict once
this
+ # becomes the default export endpoint
+ if "extra" in payload:
+ try:
+ payload["extra"] = json.loads(payload["extra"])
+ except json.JSONDecodeError:
+ logger.info("Unable to decode `extra` field: %s",
payload["extra"])
- payload["version"] = EXPORT_VERSION
+ payload["version"] = EXPORT_VERSION
- file_content = yaml.safe_dump(payload, sort_keys=False)
- yield file_name, lambda: file_content
+ file_content = yaml.safe_dump(payload, sort_keys=False)
+ yield file_name, lambda: file_content
Review Comment:
Same story, single yield outside a loop, so `file_content` can't get stomped
by a later iteration.
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
