def get_query_results(self, query, use_legacy_sql=False, max_wait_secs=None):
# type: (str, Optional[bool], Optional[int]) -> List[Tuple[Any]]
"""Returns a list or rows, each of which is a tuple of values.
Args:
query: A string with a complete SQL query.
use_legacy_sql: Whether to use legacy SQL
max_wait_secs: The maximum number of seconds to wait for the query to complete. If not
set, the class default will be used.
Returns:
A list of tuples of values.
"""
config = QueryJobConfig()
if self.maximum_billing_tier:
config.maximum_billing_tier = self.maximum_billing_tier
config.use_legacy_sql = use_legacy_sql
query_job = self.gclient.query(query, job_config=config,
retry=self.default_retry_for_api_calls)
# The above retry is for errors encountered in executing the jobs. The below retry is
# for errors encountered in polling to see whether the job is done.
query_job._retry = self.default_retry_for_async_jobs
rows = self._wait_for_job(query_job, query,
max_wait_secs=max_wait_secs or self.max_wait_secs)
if query_job.errors:
logging.warning('Errors in get_query_results: {}'.format(query_job.errors))
return [x.values() for x in list(rows)]
def select_insert(self,
source_table_id,
destination_table_id,
query_field,
prefix=' ',
fg='yellow'):
query = 'SELECT {query_field} FROM {dataset_id}.{source_table_id}'.format(
query_field=query_field,
dataset_id=self._dataset_ref.dataset_id,
source_table_id=source_table_id)
destination_table = self.dataset.table(destination_table_id)
job_config = QueryJobConfig()
job_config.use_legacy_sql = False
job_config.use_query_cache = False
job_config.write_disposition = WriteDisposition.WRITE_TRUNCATE
job_config.destination = destination_table
job = self._client.query(query, job_config)
echo('Inserting... {0}'.format(job.job_id),
prefix=prefix,
fg=fg,
no_color=self.no_color)
echo(' {0}'.format(job.query),
prefix=prefix,
fg=fg,
no_color=self.no_color)
job.result()
assert job.state == 'DONE'
error_result = job.error_result
if error_result:
raise RuntimeError(job.errors)
def execute(self,
query,
destination_table,
write_disposition="WRITE_TRUNCATE",
allow_large_results=True):
"""
:param query_file: query file path
:param destination_table: target table
:param write_disposition: default is to replace existing table. To append: WRITE_APPEND
:param allow_large_results: default to True
:return:
"""
query_configuration = QueryJobConfig()
query_configuration.use_legacy_sql = False
if destination_table:
ref = TableReferenceBuilder(destination_table, self._dataset,
self._project)
query_configuration.write_disposition = write_disposition
query_configuration.default_dataset = ref.dataset_reference
query_configuration.destination = ref.table_reference
query_configuration.allow_large_results = allow_large_results
sql_query = self.__get_query(query)
if not self._quiet:
print("-- #### {}\n{}\n".format(destination_table or "",
sql_query))
self._query_job = bigquery.Client(project=self._project).query(
sql_query, job_config=query_configuration)
if self._query_job.errors:
raise Exception(self._query_job.errors)
def get_query_results(self,
query,
use_legacy_sql=False,
max_wait_secs=None):
# type: (str, Optional[bool], Optional[int]) -> List[Tuple[Any]]
"""Returns a list or rows, each of which is a tuple of values.
Args:
query: A string with a complete SQL query.
use_legacy_sql: Whether to use legacy SQL
max_wait_secs: The maximum number of seconds to wait for the query to complete. If not
set, the class default will be used.
Returns:
A list of tuples of values.
"""
config = QueryJobConfig()
if self.maximum_billing_tier:
config.maximum_billing_tier = self.maximum_billing_tier
config.use_legacy_sql = use_legacy_sql
query_job = self._run_async_query(query, job_config=config)
rows = self._wait_for_job(query_job,
query,
max_wait_secs=max_wait_secs
or self.max_wait_secs)
if query_job.errors:
logging.warning('Errors in get_query_results: {}'.format(
query_job.errors))
return [x.values() for x in list(rows)]
def execute_sync_query(project_id, query_str, bq_client=None):
if bq_client is None:
bq_client = bigquery.Client(project_id)
config = QueryJobConfig()
config.use_legacy_sql = False
config.use_query_cache = False
query_job = bq_client.query(query_str, job_config=config, location="EU")
result = []
for row in query_job:
result.append(row)
return result
def create_table_from_query(
self,
query, # type: str
table_path, # type: str
write_disposition='WRITE_EMPTY', # type: Optional[str]
use_legacy_sql=False, # type: Optional[bool]
max_wait_secs=None, # type: Optional[int]
expected_schema=None # type: Optional[List[SchemaField]]
):
# type: (...) -> None
"""Creates a table in BigQuery from a specified query.
Args:
query: The query to run.
table_path: The path to the table (in the client's project) to write
the results to.
write_disposition: Specifies behavior if table already exists. See options here:
https://cloud.google.com/bigquery/docs/reference/rest/v2/jobs under
configuration.query.writeDisposition
use_legacy_sql: Whether the query is written in standard or legacy sql.
max_wait_secs: Seconds to wait for the query before timing out. If not
set, the class default will be used.
expected_schema: The expected schema of the resulting table; unused in this implementation
"""
if write_disposition not in [
'WRITE_TRUNCATE', 'WRITE_APPEND', 'WRITE_EMPTY'
]:
raise ValueError(
'write_disposition must be one of WRITE_TRUNCATE, '
'WRITE_APPEND, or WRITE_EMPTY')
config = QueryJobConfig()
if self.maximum_billing_tier:
config.maximum_billing_tier = self.maximum_billing_tier
config.use_legacy_sql = use_legacy_sql
config.write_disposition = write_disposition
config.allow_large_results = True
config.destination = self.get_table_reference_from_path(table_path)
query_job = self._run_async_query(query, job_config=config)
return self._wait_for_job(query_job,
query,
max_wait_secs=max_wait_secs
or self.max_wait_secs)
def execute(self, query, tbl_ref=None, append=False, preview=True):
sql_query = self.__get_query(query)
if tbl_ref:
print("-- ## " + str(tbl_ref))
print("{}{}".format("-- preview: \n" if preview else "", sql_query))
if preview:
return
job_conf = QueryJobConfig()
job_conf.use_legacy_sql = False
if tbl_ref:
job_conf.write_disposition = "WRITE_APPEND" if append else "WRITE_TRUNCATE"
job_conf.default_dataset = tbl_ref.dataset_ref
job_conf.destination = tbl_ref.table_ref
job_conf.allow_large_results = True
query_job = self.connect(tbl_ref.project if tbl_ref else None).query(
sql_query, job_config=job_conf)
if query_job.errors:
raise Exception(query_job.errors)
def create_table_from_query(self,
query, # type: str
table_path, # type: str
write_disposition='WRITE_EMPTY', # type: Optional[str]
use_legacy_sql=False, # type: Optional[bool]
max_wait_secs=None, # type: Optional[int]
expected_schema=None # type: Optional[List[SchemaField]]
):
# type: (...) -> None
"""Creates a table in BigQuery from a specified query.
Args:
query: The query to run.
table_path: The path to the table (in the client's project) to write
the results to.
write_disposition: One of 'WRITE_TRUNCATE', 'WRITE_APPEND',
'WRITE_EMPTY'. Default is WRITE_EMPTY.
use_legacy_sql: Whether the query is written in standard or legacy sql.
max_wait_secs: Seconds to wait for the query before timing out. If not
set, the class default will be used.
expected_schema: The expected schema of the resulting table; unused in this implementation
"""
if write_disposition not in ['WRITE_TRUNCATE', 'WRITE_APPEND', 'WRITE_EMPTY']:
raise ValueError('write_disposition must be one of WRITE_TRUNCATE, '
'WRITE_APPEND, or WRITE_EMPTY')
config = QueryJobConfig()
if self.maximum_billing_tier:
config.maximum_billing_tier = self.maximum_billing_tier
config.use_legacy_sql = use_legacy_sql
config.write_disposition = write_disposition
config.allow_large_results = True
config.destination = self.get_table_reference_from_path(table_path)
query_job = self.gclient.query(query, job_config=config, retry=self.default_retry)
return query_job.result(timeout=max_wait_secs or self.max_wait_secs)
def get_query_results(self, query, use_legacy_sql=False, max_wait_secs=None):
# type: (str, Optional[Bool], Optional[int]) -> List[Tuple[Any]]
"""Returns a list or rows, each of which is a tuple of values.
Args:
query: A string with a complete SQL query.
use_legacy_sql: Whether to use legacy SQL
max_wait_secs: The maximum number of seconds to wait for the query to complete. If not
set, the class default will be used.
Returns:
A list of tuples of values.
"""
config = QueryJobConfig()
if self.maximum_billing_tier:
config.maximum_billing_tier = self.maximum_billing_tier
config.use_legacy_sql = use_legacy_sql
query_job = self.gclient.query(query, job_config=config, retry=self.default_retry)
rows = query_job.result(retry=self.default_retry,
timeout=max_wait_secs or self.max_wait_secs)
return [x.values() for x in list(rows)]
def create_config():
config = QueryJobConfig()
config.use_legacy_sql = False
return config