pabloem commented on a change in pull request #13617: URL: https://github.com/apache/beam/pull/13617#discussion_r605987207
########## File path: sdks/python/apache_beam/ml/gcp/recommendations_ai.py ########## @@ -0,0 +1,638 @@ +# +# 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. +# + +"""A connector for sending API requests to the GCP Recommendations AI +API (https://cloud.google.com/recommendations). +""" + +from __future__ import absolute_import + +import itertools +from typing import Sequence +from typing import Tuple + +from google.api_core.retry import Retry + +from apache_beam import TimeDomain +from apache_beam import pvalue +from apache_beam.coders.coders import IterableCoder +from apache_beam.coders.coders import PickleCoder +from apache_beam.metrics import Metrics +from apache_beam.options.pipeline_options import GoogleCloudOptions +from apache_beam.transforms import DoFn +from apache_beam.transforms import ParDo +from apache_beam.transforms import PTransform +from apache_beam.transforms.userstate import BagStateSpec +from apache_beam.transforms.userstate import TimerSpec +from apache_beam.transforms.userstate import on_timer +from cachetools.func import ttl_cache + +# pylint: disable=wrong-import-order, wrong-import-position, ungrouped-imports +try: + from google.cloud import recommendationengine +except ImportError: + raise ImportError( + 'Google Cloud Recommendation AI not supported for this execution ' + 'environment (could not import google.cloud.recommendationengine).') +# pylint: enable=wrong-import-order, wrong-import-position, ungrouped-imports + +__all__ = [ + 'CreateCatalogItem', + 'WriteUserEvent', + 'ImportCatalogItems', + 'ImportUserEvents', + 'PredictUserEvent' +] + + +@ttl_cache(maxsize=128, ttl=3600) +def get_recommendation_prediction_client(): + """Returns a Recommendation AI - Prediction Service client.""" + _client = recommendationengine.PredictionServiceClient() + return _client + + +@ttl_cache(maxsize=128, ttl=3600) +def get_recommendation_catalog_client(): + """Returns a Recommendation AI - Catalog Service client.""" + _client = recommendationengine.CatalogServiceClient() + return _client + + +@ttl_cache(maxsize=128, ttl=3600) +def get_recommendation_user_event_client(): + """Returns a Recommendation AI - UserEvent Service client.""" + _client = recommendationengine.UserEventServiceClient() + return _client + + +class CreateCatalogItem(PTransform): + """Creates catalogitem information. + The ``PTranform`` returns a PCollectionTuple with a PCollections of + successfully and failed created CatalogItems. + + Example usage: + pipeline | CreateCatalogItem(project='example-gcp-project', + catalog_name='my-catalog') + """ + def __init__( + self, + project: str = None, + retry: Retry = None, + timeout: float = 120, + metadata: Sequence[Tuple[str, str]] = None, + catalog_name: str = "default_catalog"): + """Initializes a :class:`CreateCatalogItem` transform. + + Args: + project (str): Optional. GCP project name in which the catalog + data will be imported. + retry (google.api_core.retry.Retry): Optional. Designation of what + errors, if any, should be retried. + timeout (float): Optional. The amount of time, in seconds, to wait + for the request to complete. + metadata (Sequence[Tuple[str, str]]): Optional. Strings which + should be sent along with the request as metadata. + catalog_name (str): Optional. Name of the catalog. + Default: 'default_catalog' + """ + self.project = project + self.retry = retry + self.timeout = timeout + self.metadata = metadata + self.catalog_name = catalog_name + + def expand(self, pcoll): + if self.project is None: + self.project = pcoll.pipeline.options.view_as(GoogleCloudOptions).project + if self.project is None: + raise ValueError( + """GCP project name needs to be specified in "project" pipeline + option""") + return pcoll | ParDo( + _CreateCatalogItemFn( + self.project, + self.retry, + self.timeout, + self.metadata, + self.catalog_name)) + + +class _CreateCatalogItemFn(DoFn): + FAILED_CATALOG_ITEMS = "failed_catalog_items" + + def __init__( + self, + project: str = None, + retry: Retry = None, + timeout: float = 120, + metadata: Sequence[Tuple[str, str]] = None, + catalog_name: str = None): + self._client = None + self.retry = retry + self.timeout = timeout + self.metadata = metadata + self.parent = f"projects/{project}/locations/global/catalogs/{catalog_name}" + self.counter = Metrics.counter(self.__class__, "api_calls") + + def setup(self): + if self._client is None: + self._client = get_recommendation_catalog_client() + + def process(self, element): + catalog_item = recommendationengine.CatalogItem(element) + request = recommendationengine.CreateCatalogItemRequest( + parent=self.parent, catalog_item=catalog_item) + + try: + created_catalog_item = self._client.create_catalog_item( + request=request, + retry=self.retry, + timeout=self.timeout, + metadata=self.metadata) + + self.counter.inc() + yield recommendationengine.CatalogItem.to_dict(created_catalog_item) + except Exception: + yield pvalue.TaggedOutput( + self.FAILED_CATALOG_ITEMS, + recommendationengine.CatalogItem.to_dict(catalog_item)) + + +class _BatchItems(DoFn): Review comment: Like in Java, Python also has a GroupIntoBatches transform that also has autosharding: - https://beam.apache.org/documentation/transforms/python/aggregation/groupintobatches/ - https://beam.apache.org/releases/pydoc/current/apache_beam.transforms.util.html#apache_beam.transforms.util.GroupIntoBatches -- 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. For queries about this service, please contact Infrastructure at: [email protected]
