garf-executors 0.0.11__py3-none-any.whl → 0.2.3__py3-none-any.whl

garf_executors/entrypoints/utils.py

@@ -93,6 +93,7 @@ class GarfParamsException(Exception):
 class LoggerEnum(str, enum.Enum):
   local = 'local'
   rich = 'rich'
+  gcloud = 'gcloud'
 
 
 def init_logging(
@@ -100,6 +101,7 @@ def init_logging(
   logger_type: str | LoggerEnum = 'local',
   name: str = __name__,
 ) -> logging.Logger:
+  loglevel = getattr(logging, loglevel)
   if logger_type == 'rich':
     logging.basicConfig(
       format='%(message)s',
@@ -109,6 +111,23 @@ def init_logging(
         rich_logging.RichHandler(rich_tracebacks=True),
       ],
     )
+  elif logger_type == 'gcloud':
+    try:
+      import google.cloud.logging as glogging
+    except ImportError as e:
+      raise ImportError(
+        'Please install garf-executors with Cloud logging support - '
+        '`pip install garf-executors[bq]`'
+      ) from e
+
+    client = glogging.Client()
+    handler = glogging.handlers.CloudLoggingHandler(client, name=name)
+    handler.close()
+    glogging.handlers.setup_logging(handler, log_level=loglevel)
+    logging.basicConfig(
+      level=loglevel,
+      handlers=[handler],
+    )
   else:
     logging.basicConfig(
       format='[%(asctime)s][%(name)s][%(levelname)s] %(message)s',

garf_executors/execution_context.py

@@ -35,17 +35,17 @@ class ExecutionContext(pydantic.BaseModel):
   Attributes:
     query_parameters: Parameters to dynamically change query text.
     fetcher_parameters: Parameters to specify fetching setup.
-    writer: Type of writer to use.
+    writer: Type of writer to use. Can be a single writer string or list of writers.
     writer_parameters: Optional parameters to setup writer.
   """
 
   query_parameters: query_editor.GarfQueryParameters | None = pydantic.Field(
     default_factory=dict
   )
-  fetcher_parameters: dict[str, str | list[str | int]] | None = pydantic.Field(
-    default_factory=dict
+  fetcher_parameters: dict[str, str | bool | int | list[str | int]] | None = (
+    pydantic.Field(default_factory=dict)
   )
-  writer: str | None = None
+  writer: str | list[str] | None = None
   writer_parameters: dict[str, str] | None = pydantic.Field(
     default_factory=dict
   )
@@ -75,9 +75,42 @@ class ExecutionContext(pydantic.BaseModel):
 
   @property
   def writer_client(self) -> abs_writer.AbsWriter:
-    writer_client = writer.create_writer(self.writer, **self.writer_parameters)
-    if self.writer == 'bq':
+    """Returns single writer client."""
+    if isinstance(self.writer, list) and len(self.writer) > 0:
+      writer_type = self.writer[0]
+    else:
+      writer_type = self.writer
+
+    writer_params = self.writer_parameters or {}
+
+    if not writer_type:
+      raise ValueError('No writer specified')
+
+    writer_client = writer.create_writer(writer_type, **writer_params)
+    if writer_type == 'bq':
       _ = writer_client.create_or_get_dataset()
-    if self.writer == 'sheet':
+    if writer_type == 'sheet':
       writer_client.init_client()
     return writer_client
+
+  @property
+  def writer_clients(self) -> list[abs_writer.AbsWriter]:
+    """Returns list of writer clients."""
+    if not self.writer:
+      return []
+
+    # Convert single writer to list for uniform processing
+    writers_to_use = (
+      self.writer if isinstance(self.writer, list) else [self.writer]
+    )
+    writer_params = self.writer_parameters or {}
+
+    clients = []
+    for writer_type in writers_to_use:
+      writer_client = writer.create_writer(writer_type, **writer_params)
+      if writer_type == 'bq':
+        _ = writer_client.create_or_get_dataset()
+      if writer_type == 'sheet':
+        writer_client.init_client()
+      clients.append(writer_client)
+    return clients

garf_executors/executor.py

@@ -14,14 +14,29 @@
 
 """Defines common functionality between executors."""
 
-from concurrent import futures
+import asyncio
+import inspect
+from typing import Optional
+
+from garf_core import report_fetcher
+from opentelemetry import trace
 
 from garf_executors import execution_context
+from garf_executors.telemetry import tracer
 
 
 class Executor:
   """Defines common functionality between executors."""
 
+  def __init__(
+    self,
+    preprocessors: Optional[dict[str, report_fetcher.Processor]] = None,
+    postprocessors: Optional[dict[str, report_fetcher.Processor]] = None,
+  ) -> None:
+    self.preprocessors = preprocessors or {}
+    self.postprocessors = postprocessors or {}
+
+  @tracer.start_as_current_span('api.execute_batch')
   def execute_batch(
     self,
     batch: dict[str, str],
@@ -30,6 +45,9 @@ class Executor:
   ) -> list[str]:
     """Executes batch of queries for a common context.
 
+    If an executor has any pre/post processors, executes them first while
+    modifying the context.
+
     Args:
       batch: Mapping between query_title and its text.
       context: Execution context.
@@ -38,17 +56,69 @@ class Executor:
     Returns:
       Results of execution.
     """
-    results = []
-    with futures.ThreadPoolExecutor(max_workers=parallel_threshold) as executor:
-      future_to_query = {
-        executor.submit(
-          self.execute,
-          query,
-          title,
-          context,
-        ): query
-        for title, query in batch.items()
-      }
-      for future in futures.as_completed(future_to_query):
-        results.append(future.result())
+    span = trace.get_current_span()
+    span.set_attribute('api.parallel_threshold', parallel_threshold)
+    _handle_processors(processors=self.preprocessors, context=context)
+    results = asyncio.run(
+      self._run(
+        batch=batch, context=context, parallel_threshold=parallel_threshold
+      )
+    )
+    _handle_processors(processors=self.postprocessors, context=context)
     return results
+
+  def add_preprocessor(
+    self, preprocessors: dict[str, report_fetcher.Processor]
+  ) -> None:
+    self.preprocessors.update(preprocessors)
+
+  async def aexecute(
+    self,
+    query: str,
+    title: str,
+    context: execution_context.ExecutionContext,
+  ) -> str:
+    """Performs query execution asynchronously.
+
+    Args:
+      query: Location of the query.
+      title: Name of the query.
+      context: Query execution context.
+
+    Returns:
+      Result of writing the report.
+    """
+    return await asyncio.to_thread(self.execute, query, title, context)
+
+  async def _run(
+    self,
+    batch: dict[str, str],
+    context: execution_context.ExecutionContext,
+    parallel_threshold: int,
+  ):
+    semaphore = asyncio.Semaphore(value=parallel_threshold)
+
+    async def run_with_semaphore(fn):
+      async with semaphore:
+        return await fn
+
+    tasks = [
+      self.aexecute(query=query, title=title, context=context)
+      for title, query in batch.items()
+    ]
+    return await asyncio.gather(*(run_with_semaphore(task) for task in tasks))
+
+
+def _handle_processors(
+  processors: dict[str, report_fetcher.Processor],
+  context: execution_context.ExecutionContext,
+) -> None:
+  for k, processor in processors.items():
+    processor_signature = list(inspect.signature(processor).parameters.keys())
+    if k in context.fetcher_parameters:
+      processor_parameters = {
+        k: v
+        for k, v in context.fetcher_parameters.items()
+        if k in processor_signature
+      }
+      context.fetcher_parameters[k] = processor(**processor_parameters)

garf_executors/fetchers.py

@@ -13,12 +13,18 @@
 # limitations under the License.
 
 import inspect
+import logging
 import sys
 from importlib.metadata import entry_points
 
-from garf_core import exceptions, report_fetcher
+from garf_core import report_fetcher
 
+from garf_executors.telemetry import tracer
 
+logger = logging.getLogger(name='garf_executors.fetchers')
+
+
+@tracer.start_as_current_span('find_fetchers')
 def find_fetchers() -> set[str]:
   """Identifiers all available report fetchers."""
   if entrypoints := _get_entrypoints('garf'):
@@ -26,6 +32,7 @@ def find_fetchers() -> set[str]:
   return set()
 
 
+@tracer.start_as_current_span('get_report_fetcher')
 def get_report_fetcher(source: str) -> type[report_fetcher.ApiReportFetcher]:
   """Loads report fetcher for a given source.
 
@@ -44,15 +51,19 @@ def get_report_fetcher(source: str) -> type[report_fetcher.ApiReportFetcher]:
   for fetcher in _get_entrypoints('garf'):
     if fetcher.name == source:
       try:
-        fetcher_module = fetcher.load()
+        with tracer.start_as_current_span('load_fetcher_module') as span:
+          fetcher_module = fetcher.load()
+          span.set_attribute('loaded_module', fetcher_module.__name__)
         for name, obj in inspect.getmembers(fetcher_module):
           if inspect.isclass(obj) and issubclass(
             obj, report_fetcher.ApiReportFetcher
           ):
             return getattr(fetcher_module, name)
-      except ModuleNotFoundError:
-        continue
-  raise exceptions.ApiReportFetcherError(
+      except ModuleNotFoundError as e:
+        raise report_fetcher.ApiReportFetcherError(
+          f'Failed to load fetcher for source {source}, reason: {e}'
+        )
+  raise report_fetcher.ApiReportFetcherError(
     f'No fetcher available for the source "{source}"'
   )
 

garf_executors/garf_pb2.py

@@ -0,0 +1,45 @@
+# -*- coding: utf-8 -*-
+# Generated by the protocol buffer compiler.  DO NOT EDIT!
+# NO CHECKED-IN PROTOBUF GENCODE
+# source: garf.proto
+# Protobuf Python Version: 6.31.1
+"""Generated protocol buffer code."""
+from google.protobuf import descriptor as _descriptor
+from google.protobuf import descriptor_pool as _descriptor_pool
+from google.protobuf import runtime_version as _runtime_version
+from google.protobuf import symbol_database as _symbol_database
+from google.protobuf.internal import builder as _builder
+_runtime_version.ValidateProtobufRuntimeVersion(
+    _runtime_version.Domain.PUBLIC,
+    6,
+    31,
+    1,
+    '',
+    'garf.proto'
+)
+# @@protoc_insertion_point(imports)
+
+_sym_db = _symbol_database.Default()
+
+
+from google.protobuf import struct_pb2 as google_dot_protobuf_dot_struct__pb2
+
+
+DESCRIPTOR = _descriptor_pool.Default().AddSerializedFile(b'\n\ngarf.proto\x12\x04garf\x1a\x1cgoogle/protobuf/struct.proto\"g\n\x0e\x45xecuteRequest\x12\x0e\n\x06source\x18\x01 \x01(\t\x12\r\n\x05title\x18\x02 \x01(\t\x12\r\n\x05query\x18\x03 \x01(\t\x12\'\n\x07\x63ontext\x18\x04 \x01(\x0b\x32\x16.garf.ExecutionContext\"\xbc\x01\n\x10\x45xecutionContext\x12/\n\x10query_parameters\x18\x01 \x01(\x0b\x32\x15.garf.QueryParameters\x12\x33\n\x12\x66\x65tcher_parameters\x18\x02 \x01(\x0b\x32\x17.google.protobuf.Struct\x12\x0e\n\x06writer\x18\x03 \x01(\t\x12\x32\n\x11writer_parameters\x18\x04 \x01(\x0b\x32\x17.google.protobuf.Struct\"d\n\x0fQueryParameters\x12&\n\x05macro\x18\x01 \x01(\x0b\x32\x17.google.protobuf.Struct\x12)\n\x08template\x18\x02 \x01(\x0b\x32\x17.google.protobuf.Struct\"\"\n\x0f\x45xecuteResponse\x12\x0f\n\x07results\x18\x01 \x03(\t2G\n\x0bGarfService\x12\x38\n\x07\x45xecute\x12\x14.garf.ExecuteRequest\x1a\x15.garf.ExecuteResponse\"\x00\x62\x06proto3')
+
+_globals = globals()
+_builder.BuildMessageAndEnumDescriptors(DESCRIPTOR, _globals)
+_builder.BuildTopDescriptorsAndMessages(DESCRIPTOR, 'garf_pb2', _globals)
+if not _descriptor._USE_C_DESCRIPTORS:
+  DESCRIPTOR._loaded_options = None
+  _globals['_EXECUTEREQUEST']._serialized_start=50
+  _globals['_EXECUTEREQUEST']._serialized_end=153
+  _globals['_EXECUTIONCONTEXT']._serialized_start=156
+  _globals['_EXECUTIONCONTEXT']._serialized_end=344
+  _globals['_QUERYPARAMETERS']._serialized_start=346
+  _globals['_QUERYPARAMETERS']._serialized_end=446
+  _globals['_EXECUTERESPONSE']._serialized_start=448
+  _globals['_EXECUTERESPONSE']._serialized_end=482
+  _globals['_GARFSERVICE']._serialized_start=484
+  _globals['_GARFSERVICE']._serialized_end=555
+# @@protoc_insertion_point(module_scope)

garf_executors/garf_pb2_grpc.py

@@ -0,0 +1,97 @@
+# Generated by the gRPC Python protocol compiler plugin. DO NOT EDIT!
+"""Client and server classes corresponding to protobuf-defined services."""
+import grpc
+import warnings
+
+from . import garf_pb2 as garf__pb2
+
+GRPC_GENERATED_VERSION = '1.75.0'
+GRPC_VERSION = grpc.__version__
+_version_not_supported = False
+
+try:
+    from grpc._utilities import first_version_is_lower
+    _version_not_supported = first_version_is_lower(GRPC_VERSION, GRPC_GENERATED_VERSION)
+except ImportError:
+    _version_not_supported = True
+
+if _version_not_supported:
+    raise RuntimeError(
+        f'The grpc package installed is at version {GRPC_VERSION},'
+        + f' but the generated code in garf_pb2_grpc.py depends on'
+        + f' grpcio>={GRPC_GENERATED_VERSION}.'
+        + f' Please upgrade your grpc module to grpcio>={GRPC_GENERATED_VERSION}'
+        + f' or downgrade your generated code using grpcio-tools<={GRPC_VERSION}.'
+    )
+
+
+class GarfServiceStub(object):
+    """Missing associated documentation comment in .proto file."""
+
+    def __init__(self, channel):
+        """Constructor.
+
+        Args:
+            channel: A grpc.Channel.
+        """
+        self.Execute = channel.unary_unary(
+                '/garf.GarfService/Execute',
+                request_serializer=garf__pb2.ExecuteRequest.SerializeToString,
+                response_deserializer=garf__pb2.ExecuteResponse.FromString,
+                _registered_method=True)
+
+
+class GarfServiceServicer(object):
+    """Missing associated documentation comment in .proto file."""
+
+    def Execute(self, request, context):
+        """Missing associated documentation comment in .proto file."""
+        context.set_code(grpc.StatusCode.UNIMPLEMENTED)
+        context.set_details('Method not implemented!')
+        raise NotImplementedError('Method not implemented!')
+
+
+def add_GarfServiceServicer_to_server(servicer, server):
+    rpc_method_handlers = {
+            'Execute': grpc.unary_unary_rpc_method_handler(
+                    servicer.Execute,
+                    request_deserializer=garf__pb2.ExecuteRequest.FromString,
+                    response_serializer=garf__pb2.ExecuteResponse.SerializeToString,
+            ),
+    }
+    generic_handler = grpc.method_handlers_generic_handler(
+            'garf.GarfService', rpc_method_handlers)
+    server.add_generic_rpc_handlers((generic_handler,))
+    server.add_registered_method_handlers('garf.GarfService', rpc_method_handlers)
+
+
+ # This class is part of an EXPERIMENTAL API.
+class GarfService(object):
+    """Missing associated documentation comment in .proto file."""
+
+    @staticmethod
+    def Execute(request,
+            target,
+            options=(),
+            channel_credentials=None,
+            call_credentials=None,
+            insecure=False,
+            compression=None,
+            wait_for_ready=None,
+            timeout=None,
+            metadata=None):
+        return grpc.experimental.unary_unary(
+            request,
+            target,
+            '/garf.GarfService/Execute',
+            garf__pb2.ExecuteRequest.SerializeToString,
+            garf__pb2.ExecuteResponse.FromString,
+            options,
+            channel_credentials,
+            insecure,
+            call_credentials,
+            compression,
+            wait_for_ready,
+            timeout,
+            metadata,
+            _registered_method=True)

garf_executors/sql_executor.py

@@ -25,11 +25,14 @@ except ImportError as e:
 
 import logging
 import re
+import uuid
 
 import pandas as pd
 from garf_core import query_editor, report
+from opentelemetry import trace
 
 from garf_executors import exceptions, execution_context, executor
+from garf_executors.telemetry import tracer
 
 logger = logging.getLogger(__name__)
 
@@ -54,6 +57,7 @@ class SqlAlchemyQueryExecutor(
         engine: Initialized Engine object to operated on a given database.
     """
     self.engine = engine
+    super().__init__()
 
   @classmethod
   def from_connection_string(
@@ -66,6 +70,7 @@ class SqlAlchemyQueryExecutor(
     engine = sqlalchemy.create_engine(connection_string)
     return cls(engine)
 
+  @tracer.start_as_current_span('sql.execute')
   def execute(
     self,
     query: str,
@@ -84,35 +89,53 @@ class SqlAlchemyQueryExecutor(
     Returns:
       Report with data if query returns some data otherwise empty Report.
     """
-    logging.info('Executing script: %s', title)
+    span = trace.get_current_span()
+    logger.info('Executing script: %s', title)
     query_text = self.replace_params_template(query, context.query_parameters)
     with self.engine.begin() as conn:
       if re.findall(r'(create|update) ', query_text.lower()):
-        conn.connection.executescript(query_text)
-        results = report.GarfReport()
+        try:
+          conn.connection.executescript(query_text)
+          results = report.GarfReport()
+        except Exception as e:
+          raise SqlAlchemyQueryExecutorError(
+            f'Failed to execute query {title}: Reason: {e}'
+          ) from e
       else:
-        temp_table_name = f'temp_{title}'.replace('.', '_')
+        temp_table_name = f'temp_{uuid.uuid4().hex}'
         query_text = f'CREATE TABLE {temp_table_name} AS {query_text}'
         conn.connection.executescript(query_text)
         try:
           results = report.GarfReport.from_pandas(
             pd.read_sql(f'SELECT * FROM {temp_table_name}', conn)
           )
+        except Exception as e:
+          raise SqlAlchemyQueryExecutorError(
+            f'Failed to execute query {title}: Reason: {e}'
+          ) from e
         finally:
           conn.connection.execute(f'DROP TABLE {temp_table_name}')
       if context.writer and results:
-        writer_client = context.writer_client
-        logger.debug(
-          'Start writing data for query %s via %s writer',
-          title,
-          type(writer_client),
-        )
-        writing_result = writer_client.write(results, title)
-        logger.debug(
-          'Finish writing data for query %s via %s writer',
-          title,
-          type(writer_client),
-        )
-        logger.info('%s executed successfully', title)
-        return writing_result
+        writer_clients = context.writer_clients
+        if not writer_clients:
+          logger.warning('No writers configured, skipping write operation')
+        else:
+          writing_results = []
+          for writer_client in writer_clients:
+            logger.debug(
+              'Start writing data for query %s via %s writer',
+              title,
+              type(writer_client),
+            )
+            writing_result = writer_client.write(results, title)
+            logger.debug(
+              'Finish writing data for query %s via %s writer',
+              title,
+              type(writer_client),
+            )
+            writing_results.append(writing_result)
+          logger.info('%s executed successfully', title)
+          # Return the last writer's result for backward compatibility
+          return writing_results[-1] if writing_results else None
+      span.set_attribute('execute.num_results', len(results))
       return results

garf_executors/telemetry.py

@@ -0,0 +1,20 @@
+# Copyright 2025 Google LLC
+#
+# Licensed 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
+#
+#     https://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.
+
+# pylint: disable=C0330, g-bad-import-order, g-multiple-import
+from opentelemetry import trace
+
+tracer = trace.get_tracer(
+  instrumenting_module_name='garf_executors',
+)

garf_executors/workflow.py

@@ -0,0 +1,96 @@
+# Copyright 2026 Google LLC
+#
+# Licensed 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
+#
+#     https://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.
+from __future__ import annotations
+
+import os
+import pathlib
+
+import pydantic
+import smart_open
+import yaml
+
+from garf_executors.execution_context import ExecutionContext
+
+
+class QueryPath(pydantic.BaseModel):
+  """Path file with query."""
+
+  path: str
+
+
+class QueryDefinition(pydantic.BaseModel):
+  """Definition of a query."""
+
+  query: Query
+
+
+class Query(pydantic.BaseModel):
+  """Query elements.
+
+  Attributes:
+    text: Query text.
+    title: Name of the query.
+  """
+
+  text: str
+  title: str
+
+
+class ExecutionStep(ExecutionContext):
+  """Common context for executing one or more queries.
+
+  Attributes:
+    fetcher: Name of a fetcher to get data from API.
+    alias: Optional alias to identify execution step.
+    queries: Queries to run for a particular fetcher.
+    context: Execution context for queries and fetcher.
+  """
+
+  fetcher: str | None = None
+  alias: str | None = None
+  queries: list[QueryPath | QueryDefinition] | None = None
+
+  @property
+  def context(self) -> ExecutionContext:
+    return ExecutionContext(
+      writer=self.writer,
+      writer_parameters=self.writer_parameters,
+      query_parameters=self.query_parameters,
+      fetcher_parameters=self.fetcher_parameters,
+    )
+
+
+class Workflow(pydantic.BaseModel):
+  """Orchestrates execution of queries for multiple fetchers.
+
+  Attributes:
+    steps: Contains one or several fetcher executions.
+  """
+
+  steps: list[ExecutionStep]
+
+  @classmethod
+  def from_file(cls, path: str | pathlib.Path | os.PathLike[str]) -> Workflow:
+    """Builds workflow from local or remote yaml file."""
+    with smart_open.open(path, 'r', encoding='utf-8') as f:
+      data = yaml.safe_load(f)
+    return Workflow(steps=data.get('steps'))
+
+  def save(self, path: str | pathlib.Path | os.PathLike[str]) -> str:
+    """Saves workflow to local or remote yaml file."""
+    with smart_open.open(path, 'w', encoding='utf-8') as f:
+      yaml.dump(
+        self.model_dump(exclude_none=True).get('steps'), f, encoding='utf-8'
+      )
+    return f'Workflow is saved to {str(path)}'