def parse_lookml_file(lookml_file_name: str) -> dict:
"""Parse a LookML file into a dictionary with keys for each of its primary properties and a list of values."""
logger.info("Parsing data from LookML file {}".format(lookml_file_name))
with open(lookml_file_name, "r") as lookml_file_stream:
lookml_data = lkml.load(lookml_file_stream)
return lookml_data
def run_dbt_debug(self, *args, **kwargs) -> None:
"""Run `dbt debug` command to check if your dbt_project.yml and profiles.yml files are properly configured."""
logger.info(
"Confirming proper dbt project setup, profile and warehouse access..."
)
result = self._dbt_cli_runner(DBT_DEBUG, *args, **kwargs)
logger.info(result)
def run_cli_command(
command: Union[List[str], str],
working_directory: str,
use_shell: bool = False,
output_as_text: bool = True,
capture_output: bool = True,
**kwargs,
):
"""Execute command line subprocess"""
result = subprocess.run(
command,
shell=use_shell,
cwd=working_directory,
text=output_as_text,
capture_output=capture_output,
**kwargs,
)
if result.stderr:
raise subprocess.CalledProcessError(returncode=result.returncode,
cmd=result.args,
stderr=result.stderr)
if result.stdout:
if "Encountered an error" in result.stdout: # Handle for dbt stdout errors
logger.error("dbt Error: {}".format(result.stdout))
raise subprocess.CalledProcessError(returncode=result.returncode,
cmd=result.args,
stderr=result.stdout)
logger.debug("Command Result:\n{}".format(result.stdout))
return result.stdout
def assemble_view(
cls,
view_name: str,
sql_table_name: str = None,
derived_table: str = None,
dimensions: List[dict] = None,
dimension_groups: List[dict] = None,
measures: List[dict] = None,
sets: List[dict] = None,
parameters: List[dict] = None,
label: str = None,
required_access_grants: list = None,
extends: str = None,
extension_is_required: bool = False,
include_suggestions: bool = True,
):
assembled_view_dict = {"view": {"name": view_name}}
logger.info("Creating LookML View: {}".format(view_name))
# Validate inputs
if not sql_table_name and not derived_table and not extends:
raise DbteaException(
name="missing-lookml-view-properties",
title="Missing Necessary LookML View Properties",
detail="Created LookML Views must specify either a `sql_table_name`, `derived_table` or `extends` in order "
"to properly specify the view source",
)
# Add optional view options as needed
if label:
assembled_view_dict["view"]["label"] = label
if extends:
assembled_view_dict["view"]["extends"] = extends
if extension_is_required:
assembled_view_dict["view"]["extension"] = "required"
if sql_table_name:
assembled_view_dict["view"]["sql_table_name"] = sql_table_name
if derived_table:
assembled_view_dict["view"]["derived_table"] = derived_table
if required_access_grants:
assembled_view_dict["view"][
"required_access_grants"
] = required_access_grants
if not include_suggestions:
assembled_view_dict["view"]["suggestions"] = "no"
# Add body of View
if parameters:
assembled_view_dict["view"]["parameters"] = parameters
if dimensions:
assembled_view_dict["view"]["dimensions"] = dimensions
if dimension_groups:
assembled_view_dict["view"]["dimension_groups"] = dimension_groups
if measures:
assembled_view_dict["view"]["measures"] = measures
if sets:
assembled_view_dict["view"]["sets"] = sets
return lkml.dump(assembled_view_dict)
def dbt_model_schemas_to_lookml_views(dbt_models_data: List[dict]) -> dict:
""""""
lookml_views_dbt_data = dbt_models_data
for model_data in lookml_views_dbt_data:
if "columns" in model_data:
model_data["dimensions"] = list()
model_data["dimension_groups"] = list()
invalid_column_properties = list()
for index, column_data in enumerate(model_data["columns"]):
for column_property_key in column_data.keys():
if column_property_key not in VALID_LOOKML_DIMENSION_PROPERTIES:
invalid_column_properties.append(column_property_key)
for invalid_column_property in invalid_column_properties:
logger.debug(
"Removing property invalid for LookML for dimension {}: {}".format(
model_data.get("name"), invalid_column_property
)
)
if model_data["columns"][index].get(invalid_column_property):
del model_data["columns"][index][invalid_column_property]
if column_data.get("type") in LOOKML_DIMENSION_GROUP_TYPES:
column_data.update({"sql": "${TABLE}." + column_data.get("name")})
model_data["dimension_groups"].append(column_data)
else:
column_data.update({"sql": "${TABLE}." + column_data.get("name")})
model_data["dimensions"].append(column_data)
# model_data["dimensions"][index]["sql"] = "${TABLE}." + column_data.get("name")
del model_data["columns"]
if "alias" in model_data:
model_data["sql_table_name"] = model_data["alias"]
else:
model_data["sql_table_name"] = model_data.get(
"name"
) # TODO: Needs to be fully qualified name]
# Process model-level meta fields
if "meta" in model_data:
for meta_key, meta_value in model_data["meta"].items():
if meta_key[:7] == "looker_" or meta_key[:7] == "lookml_":
model_data[meta_key[7:]] = model_data["meta"][meta_key]
invalid_properties = list()
for model_property in model_data.keys():
if model_property not in VALID_LOOKML_VIEW_PROPERTIES:
invalid_properties.append(model_property)
for invalid_property in invalid_properties:
logger.debug(
"Removing property invalid for LookML for view {}: {}".format(
model_data.get("name"), invalid_property
)
)
del model_data[invalid_property]
return {"views": lookml_views_dbt_data}
def write_data_to_file(self, replace_if_exists: bool = True) -> None:
""""""
if self._config_file_exists() and self._config_data_exists(
) and not replace_if_exists:
logger.warning(
"Dbtea config file already exists, ignoring write config data to file step"
)
else:
logger.info("Creating {} file at path: {}".format(
self.file_name, self.config_dir))
utils.write_to_yaml_file(self.config_data,
self.config_name_and_path)
def run_dbt_run_operation(
self, macro_name: str, macro_args: dict = None, *args, **kwargs
) -> None:
"""Run `dbt run-operation` command to run dbt macros."""
logger.info("Executing dbt macro operation {}...".format(macro_name))
operation_with_macro = DBT_RUN_OPERATION.copy()
operation_with_macro.append(macro_name)
if macro_args:
operation_with_macro.append(f"--args '{macro_args}'")
result = self._dbt_cli_runner(operation_with_macro, *args, **kwargs)
logger.info(result)
def write_looker_config(self):
""""""
logger.info("Writing Looker config file at path: {}".format(
self.looker_config_path))
self.looker_config.add_section(self.looker_config_section)
for key, value in self.config_data.get(self.dbt_project, {}).items():
if key.startswith("looker_sdk_"):
self.looker_config.set(self.looker_config_section,
str(key.replace("looker_sdk_", "")),
str(value))
with open(self.looker_config_path, "w") as config_stream:
self.looker_config.write(config_stream)
def timed_function(*args, **kwargs):
start_time = timeit.default_timer()
try:
result = fn(*args, **kwargs)
finally:
elapsed_time = timeit.default_timer() - start_time
elapsed_formatted = human_readable(elapsed_time)
message_detail = get_detail(fn.__name__)
logger.info(
f"Completed {message_detail}operation in {elapsed_formatted}.\n"
)
return result
def read_data_from_file(self, local_lookml_project_path: str) -> dict:
"""Parse a LookML file into a dictionary with keys for each of its primary properties and a list of values."""
logger.info(
"Parsing data from local LookML file {}".format(
self.lookml_file_name_and_path
)
)
with open(
utils.assemble_path(
local_lookml_project_path, self.lookml_file_name_and_path
),
"r",
) as lookml_file:
return lkml.load(lookml_file)
def _parse_artifact(self, artifact_file: str):
""""""
if artifact_file not in ARTIFACT_DATA_FILES.values():
logger.warning(
"You have specified an artifact file which is not in the list of known dbt artifacts"
)
artifact_path = utils.assemble_path(
self.project_root, self.target_path, artifact_file
)
if not utils.file_exists(artifact_path):
raise DbteaException(
name="artifact-file-missing",
title="Artifact file {} is missing".format(artifact_file),
detail="There is no artifact {} at path {}. You may not have yet generated this artifact and "
"need to run models, source freshness or docs generation".format(
artifact_file, artifact_path
),
)
return utils.parse_json_file(artifact_path)
def fetch_dbt_project_directory(custom_project_directory: str = None) -> str:
"""Return path to the base of the closest dbt project by traversing from current working directory backwards in
order to find a dbt_project.yml file.
If an optional custom project path is specified (which should be a full path to the base project path of a dbt
project), return that directory instead.
"""
project_directory = os.getcwd()
root_path = os.path.abspath(os.sep)
if custom_project_directory:
custom_directory_project_file = assemble_path(custom_project_directory,
DBT_PROJECT_FILE)
if os.path.exists(custom_directory_project_file):
return custom_project_directory
else:
raise DbteaException(
name="invalid-custom-dbt-project-directory",
title="No dbt project found at supplied custom directory",
detail=
"No dbt_project.yml file found at supplied custom project directory {}, confirm your "
"custom project directory is valid".format(
custom_project_directory),
)
while project_directory != root_path:
dbt_project_file = assemble_path(project_directory, DBT_PROJECT_FILE)
if os.path.exists(dbt_project_file):
logger.info("Running dbtea against dbt project at path: {}".format(
project_directory))
return project_directory
project_directory = os.path.dirname(project_directory)
raise DbteaException(
name="missing-dbt-project",
title="No dbt project found",
detail=
"No dbt_project.yml file found in current or any direct parent paths. You need to run dbtea "
"from within dbt project in order to use its tooling, or supply a custom project directory",
)
def create_pull_request(
organization_name: str,
repository_name: str,
git_token: str,
head_branch: str,
base_branch: str = "main",
title: str = "dbtea updates",
description: str = "dbtea metadata refresh",
):
"""Creates the pull request for the head_branch against the base_branch"""
github_pulls_url = utils.assemble_path(GITHUB_API_URL, "repos",
organization_name, repository_name,
"pulls")
headers = {
"Authorization": "token {}".format(git_token),
"Content-Type": "application/json",
}
payload = {
"title": title,
"body": description,
"head": head_branch,
"base": base_branch,
}
response = requests.post(github_pulls_url,
headers=headers,
data=json.dumps(payload))
if response.status_code >= 400:
raise GitException(
name="pull-request-create-fail",
provider="github",
title="Error Creating GitHub Pull Request via API",
status=response.status_code,
detail=response.json().get("errors"),
response=response,
)
logger.info("Created pull request for branch {} at URL: {}".format(
head_branch,
response.json().get("html_url")))
def wrapper(*args, **kwargs):
try:
return function(*args, **kwargs)
except DbteaException as error:
logger.error(
f"\n{error}\n\n" +
"For support, please create an issue at https://github.com/spectacles-ci/spectacles/issues"
+ "\n")
sys.exit(error.exit_code)
except KeyboardInterrupt as error:
logger.debug(error, exc_info=True)
logger.info("Spectacles was manually interrupted.")
sys.exit(1)
except Exception as error:
logger.debug(error, exc_info=True)
logger.error(
f'\nEncountered unexpected {error.__class__.__name__}: "{error}"\n'
f"Full error traceback logged to file.\n\n" +
"For support, please create an issue at https://github.com/spectacles-ci/spectacles/issues"
+ "\n")
sys.exit(1)
def convert_to_lookml_data_type(field_name: str,
field_type: str,
include_postal_code: bool = False) -> str:
""""""
if include_postal_code and field_name in LOOKML_ZIPCODE_FIELD_NAMES:
lookml_data_type = LOOKML_TYPE_ZIP
elif field_type in LOOKML_YESNO_DATA_TYPES:
lookml_data_type = LOOKML_TYPE_BOOL
elif field_type in LOOKML_TIME_DATA_TYPES:
lookml_data_type = LOOKML_TYPE_DATETIME
elif field_type in LOOKML_DATE_DATA_TYPES:
lookml_data_type = LOOKML_TYPE_DATE
elif field_type in LOOKML_NUMBER_DATA_TYPES:
lookml_data_type = LOOKML_TYPE_NUMBER
else:
lookml_data_type = LOOKML_TYPE_STRING
logger.debug(
f"Field: {field_name} with data type: {field_type} was mapped to LookML type: {lookml_data_type}"
)
return lookml_data_type
def _dbt_cli_runner(self, input_command_as_list: list, *args, **kwargs):
"""Run dbt CLI command based on input options."""
arg_flags = list()
kwarg_flags = list()
if args:
for flag in args:
arg_flags.append(f"--{flag}")
input_command_as_list.extend(arg_flags)
if kwargs:
for flag, flag_value in kwargs.items():
kwarg_flags.append(f"--{flag} '{flag_value}'")
input_command_as_list.extend(kwarg_flags)
input_command = " ".join(input_command_as_list)
logger.info("Running dbt command: ".format(" ".join(input_command_as_list)))
return utils.run_cli_command(
input_command,
working_directory=self.project_root,
use_shell=True,
output_as_text=True,
capture_output=True,
)
def run_dbt_deps(self, require_codegen: bool = False, *args, **kwargs) -> None:
"""Run `dbt deps` command to install dbt project dependencies; the `codegen` package must be included."""
project_packages_file = utils.assemble_path(self.project_root, "packages.yml")
if require_codegen:
if not utils.file_exists(project_packages_file):
raise FileExistsError(
"You must have a packages.yml file specified in your project"
)
package_data = utils.parse_yaml_file(project_packages_file)
package_list = [
entry.get("package") for entry in package_data.get("packages", {})
]
if "fishtown-analytics/codegen" not in package_list:
raise ValueError(
"You have not brought the codegen dbt package into your project! You must include the "
"package 'fishtown-analytics/codegen' in your `packages.yml` file to codegen in bulk."
)
logger.info("Fetching dbt project package dependencies...")
result = self._dbt_cli_runner(DBT_DEPS, *args, **kwargs)
logger.info(result)
def run_dbt_parse(self, *args, **kwargs) -> None:
"""Run `dbt parse` command to provide information on performance."""
logger.info("Parsing dbt project for performance details...")
result = self._dbt_cli_runner(DBT_PARSE, *args, **kwargs)
logger.info(result)
def run_dbt_list(self, *args, **kwargs) -> None:
"""Run `dbt list` (ls) command to list all resources within the dbt project."""
logger.info("Listing dbt project resources...")
result = self._dbt_cli_runner(DBT_LIST, *args, **kwargs)
logger.info(result)
def run_dbt_run(self, *args, **kwargs) -> None:
"""Run `dbt run` command to run dbt models."""
logger.info("Running dbt models...")
result = self._dbt_cli_runner(DBT_RUN, *args, **kwargs)
logger.info(result)
def run_dbt_rpc(self, *args, **kwargs) -> None:
"""Run `dbt rpc` command to spin up an RPC server."""
logger.info("Starting dbt RPC server...")
result = self._dbt_cli_runner(DBT_RPC, *args, **kwargs)
logger.info(result)
def run_dbt_snapshot(self, *args, **kwargs) -> None:
"""Run `dbt snapshot` command to execute dbt snapshots."""
logger.info("Running dbt snapshots...")
result = self._dbt_cli_runner(DBT_SNAPSHOT, *args, **kwargs)
logger.info(result)
def run_dbt_seed(self, *args, **kwargs) -> None:
"""Run `dbt seed` command to upload seed data."""
logger.info("Uploading dbt seed data files to data warehouse...")
result = self._dbt_cli_runner(DBT_SEED, *args, **kwargs)
logger.info(result)
def run_dbt_source_snapshot_freshness(self, *args, **kwargs) -> None:
"""Run `dbt source snapshot-freshness` command to get freshness of data sources."""
logger.info("Checking freshness of data source tables...")
result = self._dbt_cli_runner(DBT_SOURCE_SNAPSHOT_FRESHNESS, *args, **kwargs)
logger.info(result)
def create_lookml_model(
model_name: str,
output_to: str = "stdout",
connection: str = None,
label: str = None,
includes: list = None,
explores: List[dict] = None,
access_grants: List[dict] = None,
tests: List[dict] = None,
datagroups: List[dict] = None,
map_layers: List[dict] = None,
named_value_formats: List[dict] = None,
fiscal_month_offset: int = None,
persist_for: str = None,
persist_with: str = None,
week_start_day: str = None,
case_sensitive: bool = True,
output_directory: str = None,
) -> Optional[str]:
""""""
assembled_model_dict = dict()
logger.info("Creating LookML Model: {}".format(model_name))
# Validate inputs
if output_to not in OUTPUT_TO_OPTIONS:
raise DbteaException(
name="invalid-lookml-model-properties",
title="Invalid LookML Model Properties",
detail="You must choose a valid output_to option from the following: {}".format(
OUTPUT_TO_OPTIONS
),
)
if output_to == "file" and not output_directory:
raise DbteaException(
name="missing-output-directory",
title="No Model Output Directory Specified",
detail="You must include an output_directory param if outputting model to a file",
)
# Add optional model options
if connection:
assembled_model_dict["connection"] = connection
if label:
assembled_model_dict["label"] = label
if includes:
assembled_model_dict["includes"] = includes
if persist_for:
assembled_model_dict["persist_for"] = persist_for
if persist_with:
assembled_model_dict["persist_with"] = persist_with
if fiscal_month_offset:
assembled_model_dict["fiscal_month_offset"] = fiscal_month_offset
if week_start_day:
assembled_model_dict["week_start_day"] = week_start_day
if not case_sensitive:
assembled_model_dict["case_sensitive"] = "no"
# Add body of Model
if datagroups:
assembled_model_dict["datagroups"] = datagroups
if access_grants:
assembled_model_dict["access_grants"] = access_grants
if explores:
assembled_model_dict["explores"] = explores
if named_value_formats:
assembled_model_dict["named_value_formats"] = named_value_formats
if map_layers:
assembled_model_dict["map_layers"] = map_layers
if tests:
assembled_model_dict["tests"] = tests
if output_to == "stdout":
return lkml.dump(assembled_model_dict)
else:
model_file_name = utils.assemble_path(
output_directory, model_name + ".model.lkml"
)
with open(model_file_name, "w") as output_stream:
output_stream.write(lkml.dump(assembled_model_dict))
def assemble_model(
cls,
model_name: str,
directory_path: str,
connection: str = None,
label: str = None,
includes: list = None,
explores: List[dict] = None,
access_grants: List[dict] = None,
tests: List[dict] = None,
datagroups: List[dict] = None,
map_layers: List[dict] = None,
named_value_formats: List[dict] = None,
fiscal_month_offset: int = None,
persist_for: str = None,
persist_with: str = None,
week_start_day: str = None,
case_sensitive: bool = True,
):
""""""
assembled_model_dict = dict()
logger.info("Creating LookML Model: {}".format(model_name))
# Add optional model options
if connection:
assembled_model_dict["connection"] = connection
if label:
assembled_model_dict["label"] = label
if includes:
assembled_model_dict["includes"] = includes
if persist_for:
assembled_model_dict["persist_for"] = persist_for
if persist_with:
assembled_model_dict["persist_with"] = persist_with
if fiscal_month_offset:
assembled_model_dict["fiscal_month_offset"] = fiscal_month_offset
if week_start_day:
assembled_model_dict["week_start_day"] = week_start_day
if not case_sensitive:
assembled_model_dict["case_sensitive"] = "no"
# Add body of Model
if datagroups:
assembled_model_dict["datagroups"] = datagroups
if access_grants:
assembled_model_dict["access_grants"] = access_grants
if explores:
assembled_model_dict["explores"] = explores
if named_value_formats:
assembled_model_dict["named_value_formats"] = named_value_formats
if map_layers:
assembled_model_dict["map_layers"] = map_layers
if tests:
assembled_model_dict["tests"] = tests
return super().__init__(
model_name,
"model",
directory_path=directory_path,
lookml_data=assembled_model_dict,
)
def run_dbt_test(self, *args, **kwargs) -> None:
"""Run `dbt test` command to run dbt models."""
logger.info("Running dbt tests to validate models...")
result = self._dbt_cli_runner(DBT_TEST, *args, **kwargs)
logger.info(result)
def run_dbt_clean(self, *args, **kwargs) -> None:
"""Run `dbt clean` command to remove clean target folders (usually dbt_modules, target) from dbt project."""
logger.info("Removing dbt clean target folders from dbt project...")
result = self._dbt_cli_runner(DBT_CLEAN, *args, **kwargs)
logger.info(result)
def run_dbt_compile(self, *args, **kwargs) -> None:
"""Run `dbt compile` command to compile dbt models."""
logger.info("Compiling dbt models...")
result = self._dbt_cli_runner(DBT_COMPILE, *args, **kwargs)
logger.info(result)
def run_dbt_init(self, *args, **kwargs) -> None:
"""Run `dbt init` command to create a base dbt project."""
logger.info("Initialize a base dbt project...")
result = self._dbt_cli_runner(DBT_INIT, *args, **kwargs)
logger.info(result)
