garf-executors 1.0.2__py3-none-any.whl → 1.2.0__py3-none-any.whl

garf/executors/entrypoints/tracer.py

@@ -14,15 +14,20 @@
 
 import os
 
-from opentelemetry import trace
+from opentelemetry import metrics, trace
+from opentelemetry.exporter.otlp.proto.grpc.metric_exporter import (
+  OTLPMetricExporter,
+)
 from opentelemetry.exporter.otlp.proto.grpc.trace_exporter import (
   OTLPSpanExporter,
 )
+from opentelemetry.sdk.metrics import MeterProvider
+from opentelemetry.sdk.metrics.export import (
+  PeriodicExportingMetricReader,
+)
 from opentelemetry.sdk.resources import Resource
 from opentelemetry.sdk.trace import TracerProvider
-from opentelemetry.sdk.trace.export import (
-  BatchSpanProcessor,
-)
+from opentelemetry.sdk.trace.export import BatchSpanProcessor
 
 DEFAULT_SERVICE_NAME = 'garf'
 
@@ -55,3 +60,23 @@ def initialize_tracer():
       tracer_provider.add_span_processor(otlp_processor)
 
   trace.set_tracer_provider(tracer_provider)
+
+
+def initialize_meter():
+  resource = Resource.create(
+    {'service.name': os.getenv('OTLP_SERVICE_NAME', DEFAULT_SERVICE_NAME)}
+  )
+  meter_provider = MeterProvider(resource=resource)
+
+  if otel_endpoint := os.getenv('OTEL_EXPORTER_OTLP_ENDPOINT'):
+    otlp_metric_exporter = OTLPMetricExporter(
+      endpoint=f'{otel_endpoint}/v1/metrics'
+    )
+    metric_reader = PeriodicExportingMetricReader(otlp_metric_exporter)
+    meter_provider = MeterProvider(
+      resource=resource, metric_readers=[metric_reader]
+    )
+  else:
+    meter_provider = MeterProvider(resource=resource)
+  metrics.set_meter_provider(meter_provider)
+  return meter_provider

garf/executors/execution_context.py

@@ -36,8 +36,8 @@ 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. Can be a single writer string or list of writers.
-    writer_parameters: Optional parameters to setup writer.
+    writer: Type of writer(s) to use.
+    writer_parameters: Optional parameters to setup writer(s).
   """
 
   query_parameters: query_editor.GarfQueryParameters | None = pydantic.Field(
@@ -77,41 +77,25 @@ class ExecutionContext(pydantic.BaseModel):
   @property
   def writer_client(self) -> abs_writer.AbsWriter:
     """Returns single writer client."""
+    if not self.writer:
+      raise writer.GarfIoWriterError('No available writer')
     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 writer_type == 'sheet':
-      writer_client.init_client()
-    return writer_client
+    writer_clients = writer.setup_writers(
+      writers=[writer_type], writer_parameters=self.writer_parameters
+    )
+    return writer_clients[0]
 
   @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
+      raise writer.GarfIoWriterError('No available writer')
     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
+    return writer.setup_writers(
+      writers=writers_to_use, writer_parameters=self.writer_parameters
+    )

garf/executors/fetchers.py

@@ -17,7 +17,7 @@ import logging
 import sys
 from importlib.metadata import entry_points
 
-from garf.core import report_fetcher
+from garf.core import report_fetcher, simulator
 from garf.executors.telemetry import tracer
 
 logger = logging.getLogger(name='garf.executors.fetchers')
@@ -31,6 +31,14 @@ def find_fetchers() -> set[str]:
   return set()
 
 
+@tracer.start_as_current_span('find_simulators')
+def find_simulators() -> set[str]:
+  """Identifiers all available report simulators."""
+  if entrypoints := _get_entrypoints('garf_simulator'):
+    return {simulator.name for simulator in entrypoints}
+  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.
@@ -57,7 +65,10 @@ def get_report_fetcher(source: str) -> type[report_fetcher.ApiReportFetcher]:
           if inspect.isclass(obj) and issubclass(
             obj, report_fetcher.ApiReportFetcher
           ):
-            return getattr(fetcher_module, name)
+            if not hasattr(obj, 'alias'):
+              return getattr(fetcher_module, name)
+            if obj.alias == fetcher.name:
+              return getattr(fetcher_module, name)
       except ModuleNotFoundError as e:
         raise report_fetcher.ApiReportFetcherError(
           f'Failed to load fetcher for source {source}, reason: {e}'
@@ -67,6 +78,45 @@ def get_report_fetcher(source: str) -> type[report_fetcher.ApiReportFetcher]:
   )
 
 
+@tracer.start_as_current_span('get_report_simulator')
+def get_report_simulator(source: str) -> type[simulator.ApiReportSimulator]:
+  """Loads report simulator for a given source.
+
+  Args:
+    source: Alias for a source associated with a simulator.
+
+  Returns:
+    Class for a found report simulator.
+
+  Raises:
+    GarfApiReportSimulatorError: When simulator cannot be loaded.
+    MissingApiReportSimulatorError: When simulator not found.
+  """
+  if source not in find_simulators():
+    raise simulator.MissingApiReportSimulatorError(source)
+  for sim in _get_entrypoints('garf_simulator'):
+    if sim.name == source:
+      try:
+        with tracer.start_as_current_span('load_simulator_module') as span:
+          simulator_module = sim.load()
+          span.set_attribute('loaded_module', simulator_module.__name__)
+        for name, obj in inspect.getmembers(simulator_module):
+          if inspect.isclass(obj) and issubclass(
+            obj, simulator.ApiReportSimulator
+          ):
+            if not hasattr(obj, 'alias'):
+              return getattr(simulator_module, name)
+            if obj.alias == sim.name:
+              return getattr(simulator_module, name)
+      except ModuleNotFoundError as e:
+        raise simulator.GarfApiReportSimulatorError(
+          f'Failed to load simulator for source {source}, reason: {e}'
+        )
+  raise simulator.GarfApiReportSimulatorError(
+    f'No simulator available for the source "{source}"'
+  )
+
+
 def _get_entrypoints(group='garf'):
   if sys.version_info.major == 3 and sys.version_info.minor == 9:
     try:

garf/executors/garf_pb2.py

@@ -25,21 +25,27 @@ _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')
+DESCRIPTOR = _descriptor_pool.Default().AddSerializedFile(b'\n\ngarf.proto\x12\x04garf\x1a\x1cgoogle/protobuf/struct.proto\"a\n\x0c\x46\x65tchRequest\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\x12.garf.FetchContext\"G\n\rFetchResponse\x12\x0f\n\x07\x63olumns\x18\x01 \x03(\t\x12%\n\x04rows\x18\x02 \x03(\x0b\x32\x17.google.protobuf.Struct\"t\n\x0c\x46\x65tchContext\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\"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(\t2{\n\x0bGarfService\x12\x38\n\x07\x45xecute\x12\x14.garf.ExecuteRequest\x1a\x15.garf.ExecuteResponse\"\x00\x12\x32\n\x05\x46\x65tch\x12\x12.garf.FetchRequest\x1a\x13.garf.FetchResponse\"\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
+  _globals['_FETCHREQUEST']._serialized_start=50
+  _globals['_FETCHREQUEST']._serialized_end=147
+  _globals['_FETCHRESPONSE']._serialized_start=149
+  _globals['_FETCHRESPONSE']._serialized_end=220
+  _globals['_FETCHCONTEXT']._serialized_start=222
+  _globals['_FETCHCONTEXT']._serialized_end=338
+  _globals['_EXECUTEREQUEST']._serialized_start=340
+  _globals['_EXECUTEREQUEST']._serialized_end=443
+  _globals['_EXECUTIONCONTEXT']._serialized_start=446
+  _globals['_EXECUTIONCONTEXT']._serialized_end=634
+  _globals['_QUERYPARAMETERS']._serialized_start=636
+  _globals['_QUERYPARAMETERS']._serialized_end=736
+  _globals['_EXECUTERESPONSE']._serialized_start=738
+  _globals['_EXECUTERESPONSE']._serialized_end=772
+  _globals['_GARFSERVICE']._serialized_start=774
+  _globals['_GARFSERVICE']._serialized_end=897
 # @@protoc_insertion_point(module_scope)

garf/executors/garf_pb2_grpc.py

@@ -5,7 +5,7 @@ import warnings
 
 from . import garf_pb2 as garf__pb2
 
-GRPC_GENERATED_VERSION = '1.75.0'
+GRPC_GENERATED_VERSION = '1.76.0'
 GRPC_VERSION = grpc.__version__
 _version_not_supported = False
 
@@ -18,7 +18,7 @@ except ImportError:
 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'
+        + ' 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}.'
@@ -39,6 +39,11 @@ class GarfServiceStub(object):
                 request_serializer=garf__pb2.ExecuteRequest.SerializeToString,
                 response_deserializer=garf__pb2.ExecuteResponse.FromString,
                 _registered_method=True)
+        self.Fetch = channel.unary_unary(
+                '/garf.GarfService/Fetch',
+                request_serializer=garf__pb2.FetchRequest.SerializeToString,
+                response_deserializer=garf__pb2.FetchResponse.FromString,
+                _registered_method=True)
 
 
 class GarfServiceServicer(object):
@@ -50,6 +55,12 @@ class GarfServiceServicer(object):
         context.set_details('Method not implemented!')
         raise NotImplementedError('Method not implemented!')
 
+    def Fetch(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 = {
@@ -58,6 +69,11 @@ def add_GarfServiceServicer_to_server(servicer, server):
                     request_deserializer=garf__pb2.ExecuteRequest.FromString,
                     response_serializer=garf__pb2.ExecuteResponse.SerializeToString,
             ),
+            'Fetch': grpc.unary_unary_rpc_method_handler(
+                    servicer.Fetch,
+                    request_deserializer=garf__pb2.FetchRequest.FromString,
+                    response_serializer=garf__pb2.FetchResponse.SerializeToString,
+            ),
     }
     generic_handler = grpc.method_handlers_generic_handler(
             'garf.GarfService', rpc_method_handlers)
@@ -95,3 +111,30 @@ class GarfService(object):
             timeout,
             metadata,
             _registered_method=True)
+
+    @staticmethod
+    def Fetch(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/Fetch',
+            garf__pb2.FetchRequest.SerializeToString,
+            garf__pb2.FetchResponse.FromString,
+            options,
+            channel_credentials,
+            insecure,
+            call_credentials,
+            compression,
+            wait_for_ready,
+            timeout,
+            metadata,
+            _registered_method=True)

garf/executors/query_processor.py

@@ -12,29 +12,39 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+"""qQuery can be used as a parameter in garf queries."""
+
 import contextlib
 
-from garf.core import query_editor
-from garf.executors import exceptions, execution_context
+from garf.core import query_editor, query_parser
+from garf.executors import execution_context
 
 
-def process_gquery(
-  context: execution_context.ExecutionContext,
-) -> execution_context.ExecutionContext:
-  for k, v in context.fetcher_parameters.items():
+class GqueryError(query_parser.GarfQueryError):
+  """Errors on incorrect qQuery syntax."""
+
+
+def _handle_sub_context(context, sub_context):
+  for k, v in sub_context.items():
     if isinstance(v, str) and v.startswith('gquery'):
       no_writer_context = context.model_copy(update={'writer': None})
       try:
-        _, alias, query = v.split(':', maxsplit=3)
+        _, alias, *query = v.split(':', maxsplit=3)
       except ValueError:
-        raise exceptions.GarfExecutorError(
+        raise GqueryError(
           f'Incorrect gquery format, should be gquery:alias:query, got {v}'
         )
+      if not alias:
+        raise GqueryError(f'Missing alias in gquery: {v}')
+      if not query:
+        raise GqueryError(f'Missing query text in gquery: {v}')
       if alias == 'sqldb':
         from garf.executors import sql_executor
 
-        gquery_executor = sql_executor.SqlAlchemyQueryExecutor(
-          **context.fetcher_parameters
+        gquery_executor = (
+          sql_executor.SqlAlchemyQueryExecutor.from_connection_string(
+            context.fetcher_parameters.get('connection_string')
+          )
         )
       elif alias == 'bq':
         from garf.executors import bq_executor
@@ -43,19 +53,27 @@ def process_gquery(
           **context.fetcher_parameters
         )
       else:
-        raise exceptions.GarfExecutorError(
-          f'Unsupported alias for gquery: {alias}'
-        )
-      with contextlib.suppress(query_editor.GarfResourceError):
+        raise GqueryError(f'Unsupported alias {alias} for gquery: {v}')
+      with contextlib.suppress(
+        query_editor.GarfResourceError, query_parser.GarfVirtualColumnError
+      ):
+        query = ':'.join(query)
         query_spec = query_editor.QuerySpecification(
           text=query, args=context.query_parameters
         ).generate()
         if len(columns := [c for c in query_spec.column_names if c != '_']) > 1:
-          raise exceptions.GarfExecutorError(
-            f'Multiple columns in gquery: {columns}'
-          )
+          raise GqueryError(f'Multiple columns in gquery definition: {columns}')
       res = gquery_executor.execute(
         query=query, title='gquery', context=no_writer_context
       )
-      context.fetcher_parameters[k] = res.to_list(row_type='scalar')
+      if len(columns := [c for c in res.column_names if c != '_']) > 1:
+        raise GqueryError(f'Multiple columns in gquery result: {columns}')
+      sub_context[k] = res.to_list(row_type='scalar')
+
+
+def process_gquery(
+  context: execution_context.ExecutionContext,
+) -> execution_context.ExecutionContext:
+  _handle_sub_context(context, context.fetcher_parameters)
+  _handle_sub_context(context, context.query_parameters.macro)
   return context

garf/executors/setup.py

@@ -0,0 +1,76 @@
+# 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.
+"""Bootstraps executor based on provided parameters."""
+
+from __future__ import annotations
+
+import importlib
+import logging
+from typing import Any
+
+from garf.executors import executor, fetchers
+from garf.executors.api_executor import ApiQueryExecutor
+from garf.executors.telemetry import tracer
+from garf.io import writer
+
+logger = logging.getLogger('garf.executors.setup')
+
+
+@tracer.start_as_current_span('setup_executor')
+def setup_executor(
+  source: str,
+  fetcher_parameters: dict[str, str | int | bool],
+  enable_cache: bool = False,
+  cache_ttl_seconds: int = 3600,
+  simulate: bool = False,
+  writers: str | list[str] | None = None,
+  writer_parameters: dict[str, Any] | None = None,
+) -> type[executor.Executor]:
+  """Initializes executors based on a source and parameters."""
+  if simulate and enable_cache:
+    logger.warning('Simulating API responses. Disabling cache.')
+    enable_cache = False
+  if writers:
+    writer_clients = writer.setup_writers(writers, writer_parameters)
+  else:
+    writer_clients = None
+  if source == 'bq':
+    bq_executor = importlib.import_module('garf.executors.bq_executor')
+    query_executor = bq_executor.BigQueryExecutor(
+      **fetcher_parameters, writers=writer_clients
+    )
+  elif source == 'sqldb':
+    sql_executor = importlib.import_module('garf.executors.sql_executor')
+    query_executor = (
+      sql_executor.SqlAlchemyQueryExecutor.from_connection_string(
+        connection_string=fetcher_parameters.get('connection_string'),
+        writers=writer_clients,
+      )
+    )
+  else:
+    concrete_api_fetcher = fetchers.get_report_fetcher(source)
+    if simulate:
+      concrete_simulator = fetchers.get_report_simulator(source)()
+    else:
+      concrete_simulator = None
+    query_executor = ApiQueryExecutor(
+      fetcher=concrete_api_fetcher(
+        **fetcher_parameters,
+        enable_cache=enable_cache,
+        cache_ttl_seconds=cache_ttl_seconds,
+      ),
+      report_simulator=concrete_simulator,
+      writers=writer_clients,
+    )
+  return query_executor

garf/executors/sql_executor.py

@@ -31,6 +31,7 @@ import pandas as pd
 from garf.core import query_editor, report
 from garf.executors import exceptions, execution_context, executor
 from garf.executors.telemetry import tracer
+from garf.io.writers import abs_writer
 from opentelemetry import trace
 
 logger = logging.getLogger(__name__)
@@ -40,9 +41,7 @@ class SqlAlchemyQueryExecutorError(exceptions.GarfExecutorError):
   """Error when SqlAlchemyQueryExecutor fails to run query."""
 
 
-class SqlAlchemyQueryExecutor(
-  executor.Executor, query_editor.TemplateProcessorMixin
-):
+class SqlAlchemyQueryExecutor(executor.Executor):
   """Handles query execution via SqlAlchemy.
 
   Attributes:
@@ -50,7 +49,10 @@ class SqlAlchemyQueryExecutor(
   """
 
   def __init__(
-    self, engine: sqlalchemy.engine.base.Engine | None = None, **kwargs: str
+    self,
+    engine: sqlalchemy.engine.base.Engine | None = None,
+    writers: list[abs_writer.AbsWriter] | None = None,
+    **kwargs: str,
   ) -> None:
     """Initializes executor with a given engine.
 
@@ -58,18 +60,19 @@ class SqlAlchemyQueryExecutor(
         engine: Initialized Engine object to operated on a given database.
     """
     self.engine = engine or sqlalchemy.create_engine('sqlite://')
+    self.writers = writers
     super().__init__()
 
   @classmethod
   def from_connection_string(
-    cls, connection_string: str | None
+    cls, connection_string: str | None, writers: list[str] | None = None
   ) -> SqlAlchemyQueryExecutor:
     """Creates executor from SqlAlchemy connection string.
 
     https://docs.sqlalchemy.org/en/20/core/engines.html
     """
     engine = sqlalchemy.create_engine(connection_string or 'sqlite://')
-    return cls(engine)
+    return cls(engine=engine, writers=writers)
 
   @tracer.start_as_current_span('sql.execute')
   def execute(
@@ -91,8 +94,18 @@ class SqlAlchemyQueryExecutor(
       Report with data if query returns some data otherwise empty Report.
     """
     span = trace.get_current_span()
+    query_spec = (
+      query_editor.QuerySpecification(
+        text=query, title=title, args=context.query_parameters
+      )
+      .remove_comments()
+      .expand()
+    )
+    query_text = query_spec.query.text
+    title = query_spec.query.title
+    span.set_attribute('query.title', title)
+    span.set_attribute('query.text', query_text)
     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()):
         try:
@@ -116,8 +129,8 @@ class SqlAlchemyQueryExecutor(
           ) from e
         finally:
           conn.connection.execute(f'DROP TABLE {temp_table_name}')
-      if context.writer and results:
-        writer_clients = context.writer_clients
+      if results and (self.writers or context.writer):
+        writer_clients = self.writers or context.writer_clients
         if not writer_clients:
           logger.warning('No writers configured, skipping write operation')
         else:

garf/executors/workflows/gcp_workflow.yaml

@@ -0,0 +1,49 @@
+run:
+  for:
+    value: pair
+    in: ${pairs}
+    steps:
+      - log_source:
+          call: sys.log
+          args:
+            data: ${pair.alias}
+      - execute_queries:
+          parallel:
+            for:
+              value: query
+              in: ${pair.queries}
+              steps:
+                - log_query:
+                    call: sys.log
+                    args:
+                      data: ${pair}
+                - execute_single_query:
+                    try:
+                      call: http.post
+                      args:
+                        url: ${sys.get_env("GARF_ENDPOINT") + "/api/execute"}
+                        auth:
+                          type: OIDC
+                        body:
+                          source: ${pair.fetcher}
+                          # query_path: ${query.path}
+                          title: ${query.query.title}
+                          query: ${query.query.text}
+                          context:
+                            fetcher_parameters: ${pair.fetcher_parameters}
+                            writer: ${pair.writer}
+                            writer_parameters: ${pair.writer_parameters}
+                            query_parameters:
+                              macro: ${pair.query_parameters.macro}
+                              template: ${pair.query_parameters.template}
+                      result: task_resp
+                    except:
+                      as: e
+                      assign:
+                        - task_resp:
+                            status: "failed"
+                            error: ${e.message}
+                - log_result:
+                    call: sys.log
+                    args:
+                      data: ${task_resp}