#
# 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.
from __future__ import annotations
import enum
from collections import namedtuple
from collections.abc import Iterable, Mapping, Sequence
from tempfile import NamedTemporaryFile
from typing import TYPE_CHECKING, Any, cast
from typing_extensions import Literal
from airflow.exceptions import AirflowException
from airflow.hooks.base import BaseHook
from airflow.models import BaseOperator
from airflow.providers.amazon.aws.hooks.s3 import S3Hook
if TYPE_CHECKING:
import pandas as pd
from airflow.providers.common.sql.hooks.sql import DbApiHook
from airflow.utils.context import Context
[docs]FileOptions = namedtuple("FileOptions", ["mode", "suffix", "function"])
[docs]FILE_OPTIONS_MAP = {
FILE_FORMAT.CSV: FileOptions("r+", ".csv", "to_csv"),
FILE_FORMAT.JSON: FileOptions("r+", ".json", "to_json"),
FILE_FORMAT.PARQUET: FileOptions("rb+", ".parquet", "to_parquet"),
}
[docs]class SqlToS3Operator(BaseOperator):
"""
Saves data from a specific SQL query into a file in S3.
.. seealso::
For more information on how to use this operator, take a look at the guide:
:ref:`howto/operator:SqlToS3Operator`
:param query: the sql query to be executed. If you want to execute a file, place the absolute path of it,
ending with .sql extension. (templated)
:param s3_bucket: bucket where the data will be stored. (templated)
:param s3_key: desired key for the file. It includes the name of the file. (templated)
:param replace: whether or not to replace the file in S3 if it previously existed
:param sql_conn_id: reference to a specific database.
:param sql_hook_params: Extra config params to be passed to the underlying hook.
Should match the desired hook constructor params.
:param parameters: (optional) the parameters to render the SQL query with.
:param aws_conn_id: reference to a specific S3 connection
:param verify: Whether or not to verify SSL certificates for S3 connection.
By default SSL certificates are verified.
You can provide the following values:
- ``False``: do not validate SSL certificates. SSL will still be used
(unless use_ssl is False), but SSL certificates will not be verified.
- ``path/to/cert/bundle.pem``: A filename of the CA cert bundle to uses.
You can specify this argument if you want to use a different
CA cert bundle than the one used by botocore.
:param file_format: the destination file format, only string 'csv', 'json' or 'parquet' is accepted.
:param max_rows_per_file: (optional) argument to set destination file number of rows limit, if source data
is larger than that, it will be dispatched into multiple files.
Will be ignored if ``groupby_kwargs`` argument is specified.
:param pd_kwargs: arguments to include in DataFrame ``.to_parquet()``, ``.to_json()`` or ``.to_csv()``.
:param groupby_kwargs: argument to include in DataFrame ``groupby()``.
"""
[docs] template_fields: Sequence[str] = (
"s3_bucket",
"s3_key",
"query",
"sql_conn_id",
)
[docs] template_ext: Sequence[str] = (".sql",)
[docs] template_fields_renderers = {
"query": "sql",
"pd_kwargs": "json",
}
def __init__(
self,
*,
query: str,
s3_bucket: str,
s3_key: str,
sql_conn_id: str,
sql_hook_params: dict | None = None,
parameters: None | Mapping[str, Any] | list | tuple = None,
replace: bool = False,
aws_conn_id: str | None = "aws_default",
verify: bool | str | None = None,
file_format: Literal["csv", "json", "parquet"] = "csv",
max_rows_per_file: int = 0,
pd_kwargs: dict | None = None,
groupby_kwargs: dict | None = None,
**kwargs,
) -> None:
super().__init__(**kwargs)
self.query = query
self.s3_bucket = s3_bucket
self.s3_key = s3_key
self.sql_conn_id = sql_conn_id
self.aws_conn_id = aws_conn_id
self.verify = verify
self.replace = replace
self.pd_kwargs = pd_kwargs or {}
self.parameters = parameters
self.max_rows_per_file = max_rows_per_file
self.groupby_kwargs = groupby_kwargs or {}
self.sql_hook_params = sql_hook_params
if "path_or_buf" in self.pd_kwargs:
raise AirflowException("The argument path_or_buf is not allowed, please remove it")
if self.max_rows_per_file and self.groupby_kwargs:
raise AirflowException(
"SqlToS3Operator arguments max_rows_per_file and groupby_kwargs "
"can not be both specified. Please choose one."
)
try:
self.file_format = FILE_FORMAT[file_format.upper()]
except KeyError:
raise AirflowException(f"The argument file_format doesn't support {file_format} value.")
@staticmethod
def _fix_dtypes(df: pd.DataFrame, file_format: FILE_FORMAT) -> None:
"""
Mutate DataFrame to set dtypes for float columns containing NaN values.
Set dtype of object to str to allow for downstream transformations.
"""
try:
import numpy as np
import pandas as pd
except ImportError as e:
from airflow.exceptions import AirflowOptionalProviderFeatureException
raise AirflowOptionalProviderFeatureException(e)
for col in df:
if df[col].dtype.name == "object" and file_format == "parquet":
# if the type wasn't identified or converted, change it to a string so if can still be
# processed.
df[col] = df[col].astype(str)
if "float" in df[col].dtype.name and df[col].hasnans:
# inspect values to determine if dtype of non-null values is int or float
notna_series: Any = df[col].dropna().values
if np.equal(notna_series, notna_series.astype(int)).all():
# set to dtype that retains integers and supports NaNs
# The type ignore can be removed here if https://github.com/numpy/numpy/pull/23690
# is merged and released as currently NumPy does not consider None as valid for x/y.
df[col] = np.where(df[col].isnull(), None, df[col]) # type: ignore[call-overload]
df[col] = df[col].astype(pd.Int64Dtype())
elif np.isclose(notna_series, notna_series.astype(int)).all():
# set to float dtype that retains floats and supports NaNs
# The type ignore can be removed here if https://github.com/numpy/numpy/pull/23690
# is merged and released
df[col] = np.where(df[col].isnull(), None, df[col]) # type: ignore[call-overload]
df[col] = df[col].astype(pd.Float64Dtype())
[docs] def execute(self, context: Context) -> None:
sql_hook = self._get_hook()
s3_conn = S3Hook(aws_conn_id=self.aws_conn_id, verify=self.verify)
data_df = sql_hook.get_pandas_df(sql=self.query, parameters=self.parameters)
self.log.info("Data from SQL obtained")
self._fix_dtypes(data_df, self.file_format)
file_options = FILE_OPTIONS_MAP[self.file_format]
for group_name, df in self._partition_dataframe(df=data_df):
with NamedTemporaryFile(mode=file_options.mode, suffix=file_options.suffix) as tmp_file:
self.log.info("Writing data to temp file")
getattr(df, file_options.function)(tmp_file.name, **self.pd_kwargs)
self.log.info("Uploading data to S3")
object_key = f"{self.s3_key}_{group_name}" if group_name else self.s3_key
s3_conn.load_file(
filename=tmp_file.name, key=object_key, bucket_name=self.s3_bucket, replace=self.replace
)
def _partition_dataframe(self, df: pd.DataFrame) -> Iterable[tuple[str, pd.DataFrame]]:
"""Partition dataframe using pandas groupby() method."""
try:
import secrets
import string
import numpy as np
except ImportError:
pass
# if max_rows_per_file argument is specified, a temporary column with a random unusual name will be
# added to the dataframe. This column is used to dispatch the dataframe into smaller ones using groupby()
random_column_name = ""
if self.max_rows_per_file and not self.groupby_kwargs:
random_column_name = "".join(secrets.choice(string.ascii_letters) for _ in range(20))
df[random_column_name] = np.arange(len(df)) // self.max_rows_per_file
self.groupby_kwargs = {"by": random_column_name}
if not self.groupby_kwargs:
yield "", df
return
for group_label in (grouped_df := df.groupby(**self.groupby_kwargs)).groups:
yield (
cast(str, group_label),
grouped_df.get_group(group_label)
.drop(random_column_name, axis=1, errors="ignore")
.reset_index(drop=True),
)
def _get_hook(self) -> DbApiHook:
self.log.debug("Get connection for %s", self.sql_conn_id)
conn = BaseHook.get_connection(self.sql_conn_id)
hook = conn.get_hook(hook_params=self.sql_hook_params)
if not callable(getattr(hook, "get_pandas_df", None)):
raise AirflowException(
"This hook is not supported. The hook class must have get_pandas_df method."
)
return hook