Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[migration] metadata for dashboard filters #9109

Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions UPDATING.md
Original file line number Diff line number Diff line change
Expand Up @@ -22,6 +22,9 @@ This file documents any backwards-incompatible changes in Superset and
assists people when migrating to a new version.

## Next
* [9109](https://github.com/apache/incubator-superset/pull/9109): Expire `filter_immune_slices` and
`filter_immune_filter_fields` to favor dashboard scoped filter metadata `filter_scopes`.

* [9046](https://github.com/apache/incubator-superset/pull/9046): Replaces `can_only_access_owned_queries` by
`all_query_access` favoring a white list approach. Since a new permission is introduced use `superset init`
to create and associate it by default to the `Admin` role. Note that, by default, all non `Admin` users will
Expand Down
Original file line number Diff line number Diff line change
@@ -0,0 +1,112 @@
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements. See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership. The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied. See the License for the
# specific language governing permissions and limitations
# under the License.
"""empty message

Revision ID: 3325d4caccc8
Revises: e96dbf2cfef0
Create Date: 2020-02-07 14:13:51.714678

"""

# revision identifiers, used by Alembic.
import json
import logging

from alembic import op
from sqlalchemy import and_, Column, ForeignKey, Integer, String, Table, Text
from sqlalchemy.ext.declarative import declarative_base
from sqlalchemy.orm import relationship

from superset import db
from superset.utils.dashboard_filter_scopes_converter import convert_filter_scopes

revision = "3325d4caccc8"
down_revision = "e96dbf2cfef0"

Base = declarative_base()


class Slice(Base):
"""Declarative class to do query in upgrade"""

__tablename__ = "slices"
id = Column(Integer, primary_key=True)
slice_name = Column(String(250))
params = Column(Text)
viz_type = Column(String(250))


dashboard_slices = Table(
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Just for consistency, can you make this the same format as Slice and Dashboard?

class DashboardSlices(Base):
    __tablename__ = "dashboard_slices"
    id = Column(Integer, primary_key = True)
    ...and so on

Copy link
Author

@graceguo-supercat graceguo-supercat Feb 11, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I found a few places in our code base that use variable (instead of class) for relationship table.

"dashboard_slices",
Base.metadata,
Column("id", Integer, primary_key=True),
Column("dashboard_id", Integer, ForeignKey("dashboards.id")),
Column("slice_id", Integer, ForeignKey("slices.id")),
)


class Dashboard(Base):
__tablename__ = "dashboards"
id = Column(Integer, primary_key=True)
json_metadata = Column(Text)
slices = relationship("Slice", secondary=dashboard_slices, backref="dashboards")


def upgrade():
bind = op.get_bind()
session = db.Session(bind=bind)

dashboards = session.query(Dashboard).all()
for i, dashboard in enumerate(dashboards):
print(f"scanning dashboard ({i + 1}/{len(dashboards)}) >>>>")
try:
json_metadata = json.loads(dashboard.json_metadata or "{}")
if "filter_scopes" in json_metadata:
continue

filters = [
slice for slice in dashboard.slices if slice.viz_type == "filter_box"
]

# if dashboard has filter_box
if filters:
filter_scopes = convert_filter_scopes(json_metadata, filters)
json_metadata["filter_scopes"] = filter_scopes
logging.info(
f"Adding filter_scopes for dashboard {dashboard.id}: {json.dumps(filter_scopes)}"
)

json_metadata.pop("filter_immune_slices", None)
json_metadata.pop("filter_immune_slice_fields", None)

if json_metadata:
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Don’t we also need to update the JSON metadata of the record if it’s falsey?

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

correct! I fixed it.
some old dashboards have no json_metadata. so after this migration i want to keep their json_metadata attribute be None instead of {}

dashboard.json_metadata = json.dumps(
json_metadata, indent=None, separators=(",", ":"), sort_keys=True
)
else:
dashboard.json_metadata = None

session.merge(dashboard)
except Exception as e:
logging.exception(f"dashboard {dashboard.id} has error: {e}")
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

When you tested this, were there ever any errors? I'm curious what kind of errors might occur.

Copy link
Author

@graceguo-supercat graceguo-supercat Feb 11, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

i mean to catch any user data caused null pointer exception. for example, user may have added invalid json_metada that I can't parse ? but should be very rare.


session.commit()
session.close()


def downgrade():
pass
72 changes: 72 additions & 0 deletions superset/utils/dashboard_filter_scopes_converter.py
Original file line number Diff line number Diff line change
@@ -0,0 +1,72 @@
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements. See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership. The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied. See the License for the
# specific language governing permissions and limitations
# under the License.
import json
import logging
from collections import defaultdict
from typing import Dict, List

from superset.models.slice import Slice

logger = logging.getLogger(__name__)


def convert_filter_scopes(json_metadata: Dict, filters: List[Slice]):
filter_scopes = {}
immuned_by_id: List[int] = json_metadata.get("filter_immune_slices") or []
immuned_by_column: Dict = defaultdict(list)
for slice_id, columns in json_metadata.get(
"filter_immune_slice_fields", {}
).items():
for column in columns:
immuned_by_column[column].append(int(slice_id))

def add_filter_scope(filter_field, filter_id):
# in case filter field is invalid
if isinstance(filter_field, str):
current_filter_immune = list(
set(immuned_by_id + immuned_by_column.get(filter_field, []))
)
filter_fields[filter_field] = {
"scope": ["ROOT_ID"],
"immune": current_filter_immune,
}
else:
logging.info(f"slice [{filter_id}] has invalid field: {filter_field}")

for filter_slice in filters:
filter_fields: Dict = {}
filter_id = filter_slice.id
slice_params = json.loads(filter_slice.params or "{}")
configs = slice_params.get("filter_configs") or []

if slice_params.get("date_filter"):
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

How do we know this is the exhaustive list of slice parameters?

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Screen_Shot_2020-02-14_at_10_29_54_AM

I was thinking about to add an iteration for all the possible date filter related keys. But think this function will be used by:

  • db migration, only run once,
  • dashboard import, to convert filter_immune metadata in old dashboard to use new filter_scopes
    Given the usage is kind of backward compatible not future compatible, I feel enumeration probably is good enough?

add_filter_scope("__time_range", filter_id)
if slice_params.get("show_sqla_time_column"):
add_filter_scope("__time_col", filter_id)
if slice_params.get("show_sqla_time_granularity"):
add_filter_scope("__time_grain", filter_id)
if slice_params.get("show_druid_time_granularity"):
add_filter_scope("__granularity", filter_id)
if slice_params.get("show_druid_time_origin"):
add_filter_scope("druid_time_origin", filter_id)
for config in configs:
add_filter_scope(config.get("column"), filter_id)

if filter_fields:
filter_scopes[filter_id] = filter_fields

return filter_scopes