feat: new Columnar upload form and API (#28192)

This commit is contained in:
Daniel Vaz Gaspar
2024-05-06 15:51:42 +01:00
committed by GitHub
parent f5843fe588
commit 9a339f08a7
29 changed files with 2267 additions and 1232 deletions

View File

@@ -14,34 +14,26 @@
# KIND, either express or implied. See the License for the
# specific language governing permissions and limitations
# under the License.
import io
import zipfile
from typing import Any, TYPE_CHECKING
import pandas as pd
from flask import flash, g, redirect
from flask import redirect
from flask_appbuilder import expose, SimpleFormView
from flask_appbuilder.models.sqla.interface import SQLAInterface
from flask_appbuilder.security.decorators import has_access
from flask_babel import lazy_gettext as _
from werkzeug.wrappers import Response
from wtforms.fields import StringField
from wtforms.validators import ValidationError
import superset.models.core as models
from superset import app, db
from superset.connectors.sqla.models import SqlaTable
from superset import app
from superset.constants import MODEL_VIEW_RW_METHOD_PERMISSION_MAP, RouteMethod
from superset.exceptions import CertificateException
from superset.extensions import event_logger
from superset.sql_parse import Table
from superset.superset_typing import FlaskResponse
from superset.utils import core as utils
from superset.views.base import DeleteMixin, SupersetModelView, YamlExportMixin
from .forms import ColumnarToDatabaseForm
from .mixins import DatabaseMixin
from .validators import schema_allows_file_upload, sqlalchemy_uri_validator
from .validators import sqlalchemy_uri_validator
if TYPE_CHECKING:
from werkzeug.datastructures import FileStorage
@@ -148,145 +140,3 @@ class CustomFormView(SimpleFormView):
form=form,
appbuilder=self.appbuilder,
)
class ColumnarToDatabaseView(SimpleFormView):
form = ColumnarToDatabaseForm
form_template = "superset/form_view/columnar_to_database_view/edit.html"
form_title = _("Columnar to Database configuration")
add_columns = ["database", "schema", "table_name"]
def form_get(self, form: ColumnarToDatabaseForm) -> None:
form.if_exists.data = "fail"
def form_post( # pylint: disable=too-many-locals
self, form: ColumnarToDatabaseForm
) -> Response:
database = form.database.data
columnar_table = Table(table=form.name.data, schema=form.schema.data)
files = form.columnar_file.data
file_type = {file.filename.split(".")[-1] for file in files}
if file_type == {"zip"}:
zipfile_ob = zipfile.ZipFile( # pylint: disable=consider-using-with
form.columnar_file.data[0]
)
file_type = {filename.split(".")[-1] for filename in zipfile_ob.namelist()}
files = [
# pylint: disable=consider-using-with
io.BytesIO((zipfile_ob.open(filename).read(), filename)[0])
for filename in zipfile_ob.namelist()
]
if len(file_type) > 1:
message = _(
"Multiple file extensions are not allowed for columnar uploads."
" Please make sure all files are of the same extension.",
)
flash(message, "danger")
return redirect("/columnartodatabaseview/form")
read = pd.read_parquet
kwargs = {
"columns": form.usecols.data if form.usecols.data else None,
}
if not schema_allows_file_upload(database, columnar_table.schema):
message = _(
'Database "%(database_name)s" schema "%(schema_name)s" '
"is not allowed for columnar uploads. "
"Please contact your Superset Admin.",
database_name=database.database_name,
schema_name=columnar_table.schema,
)
flash(message, "danger")
return redirect("/columnartodatabaseview/form")
try:
chunks = [read(file, **kwargs) for file in files]
df = pd.concat(chunks)
database = (
db.session.query(models.Database)
.filter_by(id=form.data.get("database").data.get("id"))
.one()
)
database.db_engine_spec.df_to_sql(
database,
columnar_table,
df,
to_sql_kwargs={
"chunksize": 1000,
"if_exists": form.if_exists.data,
"index": form.index.data,
"index_label": form.index_label.data,
},
)
# Connect table to the database that should be used for exploration.
# E.g. if hive was used to upload a csv, presto will be a better option
# to explore the table.
explore_database = database
explore_database_id = database.explore_database_id
if explore_database_id:
explore_database = (
db.session.query(models.Database)
.filter_by(id=explore_database_id)
.one_or_none()
or database
)
sqla_table = (
db.session.query(SqlaTable)
.filter_by(
table_name=columnar_table.table,
schema=columnar_table.schema,
database_id=explore_database.id,
)
.one_or_none()
)
if sqla_table:
sqla_table.fetch_metadata()
if not sqla_table:
sqla_table = SqlaTable(table_name=columnar_table.table)
sqla_table.database = explore_database
sqla_table.database_id = database.id
sqla_table.owners = [g.user]
sqla_table.schema = columnar_table.schema
sqla_table.fetch_metadata()
db.session.add(sqla_table)
db.session.commit()
except Exception as ex: # pylint: disable=broad-except
db.session.rollback()
message = _(
'Unable to upload Columnar file "%(filename)s" to table '
'"%(table_name)s" in database "%(db_name)s". '
"Error message: %(error_msg)s",
filename=[file.filename for file in form.columnar_file.data],
table_name=form.name.data,
db_name=database.database_name,
error_msg=str(ex),
)
flash(message, "danger")
stats_logger.incr("failed_columnar_upload")
return redirect("/columnartodatabaseview/form")
# Go back to welcome page / splash screen
message = _(
'Columnar file "%(columnar_filename)s" uploaded to table "%(table_name)s" '
'in database "%(db_name)s"',
columnar_filename=[file.filename for file in form.columnar_file.data],
table_name=str(columnar_table),
db_name=sqla_table.database.database_name,
)
flash(message, "info")
event_logger.log_with_context(
action="successful_columnar_upload",
database=form.database.data.name,
schema=form.schema.data,
table=form.name.data,
)
return redirect("/tablemodelview/list/")