gbq.py

# Copyright (c) 2017 pandas-gbq Authors All rights reserved.
# Use of this source code is governed by a BSD-style
# license that can be found in the LICENSE file.

import copy
import concurrent.futures
from datetime import datetime
import logging
import re
import time
import typing
from typing import Any, Dict, Optional, Sequence, Union
import warnings

import numpy as np

# Only import at module-level at type checking time to avoid circular
# dependencies in the pandas package, which has an optional dependency on
# pandas-gbq.
if typing.TYPE_CHECKING:  # pragma: NO COVER
    import pandas

from pandas_gbq.exceptions import AccessDenied, GenericGBQException
from pandas_gbq.features import FEATURES
import pandas_gbq.schema
import pandas_gbq.timestamp

try:
    import tqdm  # noqa
except ImportError:
    tqdm = None

logger = logging.getLogger(__name__)


def _test_google_api_imports():
    try:
        import pkg_resources  # noqa
    except ImportError as ex:  # pragma: NO COVER
        raise ImportError("pandas-gbq requires setuptools") from ex

    try:
        import db_dtypes  # noqa
    except ImportError as ex:  # pragma: NO COVER
        raise ImportError("pandas-gbq requires db-dtypes") from ex

    try:
        import pydata_google_auth  # noqa
    except ImportError as ex:  # pragma: NO COVER
        raise ImportError("pandas-gbq requires pydata-google-auth") from ex

    try:
        from google_auth_oauthlib.flow import InstalledAppFlow  # noqa
    except ImportError as ex:  # pragma: NO COVER
        raise ImportError("pandas-gbq requires google-auth-oauthlib") from ex

    try:
        import google.auth  # noqa
    except ImportError as ex:  # pragma: NO COVER
        raise ImportError("pandas-gbq requires google-auth") from ex

    try:
        from google.cloud import bigquery  # noqa
    except ImportError as ex:  # pragma: NO COVER
        raise ImportError("pandas-gbq requires google-cloud-bigquery") from ex


def _is_query(query_or_table: str) -> bool:
    return re.search(r"\s", query_or_table.strip(), re.MULTILINE) is not None


class DatasetCreationError(ValueError):
    """
    Raised when the create dataset method fails
    """

    pass


class InvalidColumnOrder(ValueError):
    """
    Raised when the provided column order for output
    results DataFrame does not match the schema
    returned by BigQuery.
    """

    pass


class InvalidIndexColumn(ValueError):
    """
    Raised when the provided index column for output
    results DataFrame does not match the schema
    returned by BigQuery.
    """

    pass


class InvalidPageToken(ValueError):
    """
    Raised when Google BigQuery fails to return,
    or returns a duplicate page token.
    """

    pass


class InvalidSchema(ValueError):
    """
    Raised when the provided DataFrame does
    not match the schema of the destination
    table in BigQuery.
    """

    def __init__(self, message: str):
        self._message = message

    @property
    def message(self) -> str:
        return self._message


class NotFoundException(ValueError):
    """
    Raised when the project_id, table or dataset provided in the query could
    not be found.
    """

    pass


class QueryTimeout(ValueError):
    """
    Raised when the query request exceeds the timeoutMs value specified in the
    BigQuery configuration.
    """

    pass


class TableCreationError(ValueError):
    """
    Raised when the create table method fails
    """

    def __init__(self, message: str):
        self._message = message

    @property
    def message(self) -> str:
        return self._message


class Context(object):
    """Storage for objects to be used throughout a session.

    A Context object is initialized when the ``pandas_gbq`` module is
    imported, and can be found at :attr:`pandas_gbq.context`.
    """

    def __init__(self):
        self._credentials = None
        self._project = None
        # dialect defaults to None so that read_gbq can stop warning if set.
        self._dialect = None

    @property
    def credentials(self):
        """
        Credentials to use for Google APIs.

        These credentials are automatically cached in memory by calls to
        :func:`pandas_gbq.read_gbq` and :func:`pandas_gbq.to_gbq`. To
        manually set the credentials, construct an
        :class:`google.auth.credentials.Credentials` object and set it as
        the context credentials as demonstrated in the example below. See
        `auth docs`_ for more information on obtaining credentials.

        .. _auth docs: http://google-auth.readthedocs.io
            /en/latest/user-guide.html#obtaining-credentials

        Returns
        -------
        google.auth.credentials.Credentials

        Examples
        --------

        Manually setting the context credentials:

        >>> import pandas_gbq
        >>> from google.oauth2 import service_account
        >>> credentials = service_account.Credentials.from_service_account_file(
        ...     '/path/to/key.json',
        ... )
        >>> pandas_gbq.context.credentials = credentials
        """
        return self._credentials

    @credentials.setter
    def credentials(self, value):
        self._credentials = value

    @property
    def project(self):
        """Default project to use for calls to Google APIs.

        Returns
        -------
        str

        Examples
        --------

        Manually setting the context project:

        >>> import pandas_gbq
        >>> pandas_gbq.context.project = 'my-project'
        """
        return self._project

    @project.setter
    def project(self, value):
        self._project = value

    @property
    def dialect(self):
        """
        Default dialect to use in :func:`pandas_gbq.read_gbq`.

        Allowed values for the BigQuery SQL syntax dialect:

        ``'legacy'``
            Use BigQuery's legacy SQL dialect. For more information see
            `BigQuery Legacy SQL Reference
            <https://cloud.google.com/bigquery/docs/reference/legacy-sql>`__.
        ``'standard'``
            Use BigQuery's standard SQL, which is
            compliant with the SQL 2011 standard. For more information
            see `BigQuery Standard SQL Reference
            <https://cloud.google.com/bigquery/docs/reference/standard-sql/>`__.

        Returns
        -------
        str

        Examples
        --------

        Setting the default syntax to standard:

        >>> import pandas_gbq
        >>> pandas_gbq.context.dialect = 'standard'
        """
        return self._dialect

    @dialect.setter
    def dialect(self, value):
        self._dialect = value


# Create an empty context, used to cache credentials.
context = Context()
"""A :class:`pandas_gbq.Context` object used to cache credentials.

Credentials automatically are cached in-memory by :func:`pandas_gbq.read_gbq`
and :func:`pandas_gbq.to_gbq`.
"""


class GbqConnector(object):
    def __init__(
        self,
        project_id,
        reauth=False,
        private_key=None,
        auth_local_webserver=True,
        dialect="standard",
        location=None,
        credentials=None,
        use_bqstorage_api=False,
    ):
        global context
        from google.api_core.exceptions import GoogleAPIError
        from google.api_core.exceptions import ClientError
        from pandas_gbq import auth

        self.http_error = (ClientError, GoogleAPIError)
        self.project_id = project_id
        self.location = location
        self.reauth = reauth
        self.private_key = private_key
        self.auth_local_webserver = auth_local_webserver
        self.dialect = dialect
        self.credentials = credentials
        default_project = None

        # Service account credentials have a project associated with them.
        # Prefer that project if none was supplied.
        if self.project_id is None and hasattr(self.credentials, "project_id"):
            self.project_id = credentials.project_id

        # Load credentials from cache.
        if not self.credentials:
            self.credentials = context.credentials
            default_project = context.project

        # Credentials were explicitly asked for, so don't use the cache.
        if private_key or reauth or not self.credentials:
            self.credentials, default_project = auth.get_credentials(
                private_key=private_key,
                project_id=project_id,
                reauth=reauth,
                auth_local_webserver=auth_local_webserver,
            )

        if self.project_id is None:
            self.project_id = default_project

        if self.project_id is None:
            raise ValueError("Could not determine project ID and one was not supplied.")

        # Cache the credentials if they haven't been set yet.
        if context.credentials is None:
            context.credentials = self.credentials
        if context.project is None:
            context.project = self.project_id

        self.client = self.get_client()
        self.use_bqstorage_api = use_bqstorage_api

        # BQ Queries costs $5 per TB. First 1 TB per month is free
        # see here for more: https://cloud.google.com/bigquery/pricing
        self.query_price_for_TB = 5.0 / 2**40  # USD/TB

    def _start_timer(self):
        self.start = time.time()

    def get_elapsed_seconds(self):
        return round(time.time() - self.start, 2)

    def log_elapsed_seconds(self, prefix="Elapsed", postfix="s.", overlong=6):
        sec = self.get_elapsed_seconds()
        if sec > overlong:
            logger.info("{} {} {}".format(prefix, sec, postfix))

    # http://stackoverflow.com/questions/1094841/reusable-library-to-get-human-readable-version-of-file-size
    @staticmethod
    def sizeof_fmt(num, suffix="B"):
        fmt = "%3.1f %s%s"
        for unit in ["", "K", "M", "G", "T", "P", "E", "Z"]:
            if abs(num) < 1024.0:
                return fmt % (num, unit, suffix)
            num /= 1024.0
        return fmt % (num, "Y", suffix)

    def get_client(self):
        import google.api_core.client_info
        from google.cloud import bigquery
        import pandas

        client_info = google.api_core.client_info.ClientInfo(
            user_agent="pandas-{}".format(pandas.__version__)
        )
        return bigquery.Client(
            project=self.project_id,
            credentials=self.credentials,
            client_info=client_info,
        )

    @staticmethod
    def process_http_error(ex):
        # See `BigQuery Troubleshooting Errors
        # <https://cloud.google.com/bigquery/troubleshooting-errors>`__
        if "cancelled" in ex.message:
            raise QueryTimeout("Reason: {0}".format(ex))
        elif "Provided Schema does not match" in ex.message:
            error_message = ex.errors[0]["message"]
            raise InvalidSchema(f"Reason: {error_message}")
        elif "Already Exists: Table" in ex.message:
            error_message = ex.errors[0]["message"]
            raise TableCreationError(f"Reason: {error_message}")
        else:
            raise GenericGBQException("Reason: {0}".format(ex))

    def download_table(
        self,
        table_id: str,
        max_results: Optional[int] = None,
        progress_bar_type: Optional[str] = None,
        dtypes: Optional[Dict[str, Union[str, Any]]] = None,
    ) -> "pandas.DataFrame":
        from google.cloud import bigquery

        self._start_timer()

        try:
            table_ref = bigquery.TableReference.from_string(
                table_id, default_project=self.project_id
            )
            rows_iter = self.client.list_rows(table_ref, max_results=max_results)
        except self.http_error as ex:
            self.process_http_error(ex)

        return self._download_results(
            rows_iter,
            max_results=max_results,
            progress_bar_type=progress_bar_type,
            user_dtypes=dtypes,
        )

    def _wait_for_query_job(self, query_reply, timeout_ms):
        """Wait for query to complete, pausing occasionally to update progress.

        Args:
            query_reply (QueryJob):
                A query job which has started.

            timeout_ms (Optional[int]):
                How long to wait before cancelling the query.
        """
        # Wait at most 10 seconds so we can show progress.
        # TODO(https://github.com/googleapis/python-bigquery-pandas/issues/327):
        # Include a tqdm progress bar here instead of a stream of log messages.
        timeout_sec = 10.0
        if timeout_ms:
            timeout_sec = min(timeout_sec, timeout_ms / 1000.0)

        while query_reply.state != "DONE":
            self.log_elapsed_seconds("  Elapsed", "s. Waiting...")

            if timeout_ms and timeout_ms < self.get_elapsed_seconds() * 1000:
                self.client.cancel_job(
                    query_reply.job_id, location=query_reply.location
                )
                raise QueryTimeout("Query timeout: {} ms".format(timeout_ms))

            try:
                query_reply.result(timeout=timeout_sec)
            except concurrent.futures.TimeoutError:
                # Use our own timeout logic
                pass
            except self.http_error as ex:
                self.process_http_error(ex)

    def run_query(self, query, max_results=None, progress_bar_type=None, **kwargs):
        from google.auth.exceptions import RefreshError
        from google.cloud import bigquery
        import pandas

        job_config = {
            "query": {
                "useLegacySql": self.dialect
                == "legacy"
                # 'allowLargeResults', 'createDisposition',
                # 'preserveNulls', destinationTable, useQueryCache
            }
        }
        config = kwargs.get("configuration")
        if config is not None:
            job_config.update(config)

        self._start_timer()

        try:
            logger.debug("Requesting query... ")
            query_reply = self.client.query(
                query,
                job_config=bigquery.QueryJobConfig.from_api_repr(job_config),
                location=self.location,
                project=self.project_id,
            )
            logger.debug("Query running...")
        except (RefreshError, ValueError):
            if self.private_key:
                raise AccessDenied("The service account credentials are not valid")
            else:
                raise AccessDenied(
                    "The credentials have been revoked or expired, "
                    "please re-run the application to re-authorize"
                )
        except self.http_error as ex:
            self.process_http_error(ex)

        job_id = query_reply.job_id
        logger.debug("Job ID: %s" % job_id)

        timeout_ms = job_config.get("jobTimeoutMs") or job_config["query"].get(
            "timeoutMs"
        )
        timeout_ms = int(timeout_ms) if timeout_ms else None
        self._wait_for_query_job(query_reply, timeout_ms)

        if query_reply.cache_hit:
            logger.debug("Query done.\nCache hit.\n")
        else:
            bytes_processed = query_reply.total_bytes_processed or 0
            bytes_billed = query_reply.total_bytes_billed or 0
            logger.debug(
                "Query done.\nProcessed: {} Billed: {}".format(
                    self.sizeof_fmt(bytes_processed),
                    self.sizeof_fmt(bytes_billed),
                )
            )
            logger.debug(
                "Standard price: ${:,.2f} USD\n".format(
                    bytes_billed * self.query_price_for_TB
                )
            )

        dtypes = kwargs.get("dtypes")

        # Ensure destination is populated.
        try:
            query_reply.result()
        except self.http_error as ex:
            self.process_http_error(ex)

        # Avoid attempting to download results from DML queries, which have no
        # destination.
        if query_reply.destination is None:
            return pandas.DataFrame()

        rows_iter = self.client.list_rows(
            query_reply.destination, max_results=max_results
        )
        return self._download_results(
            rows_iter,
            max_results=max_results,
            progress_bar_type=progress_bar_type,
            user_dtypes=dtypes,
        )

    def _download_results(
        self,
        rows_iter,
        max_results=None,
        progress_bar_type=None,
        user_dtypes=None,
    ):
        # No results are desired, so don't bother downloading anything.
        if max_results == 0:
            return None

        if user_dtypes is None:
            user_dtypes = {}

        create_bqstorage_client = self.use_bqstorage_api
        if max_results is not None:
            create_bqstorage_client = False

        to_dataframe_kwargs = {}
        if FEATURES.bigquery_needs_date_as_object:
            to_dataframe_kwargs["date_as_object"] = True

        try:
            schema_fields = [field.to_api_repr() for field in rows_iter.schema]
            conversion_dtypes = _bqschema_to_nullsafe_dtypes(schema_fields)
            conversion_dtypes.update(user_dtypes)
            df = rows_iter.to_dataframe(
                dtypes=conversion_dtypes,
                progress_bar_type=progress_bar_type,
                create_bqstorage_client=create_bqstorage_client,
                **to_dataframe_kwargs,
            )
        except self.http_error as ex:
            self.process_http_error(ex)

        df = _finalize_dtypes(df, schema_fields)

        logger.debug("Got {} rows.\n".format(rows_iter.total_rows))
        return df

    def load_data(
        self,
        dataframe,
        destination_table_ref,
        write_disposition,
        chunksize=None,
        schema=None,
        progress_bar=True,
        api_method: str = "load_parquet",
        billing_project: Optional[str] = None,
    ):
        from pandas_gbq import load

        total_rows = len(dataframe)
        try:
            chunks = load.load_chunks(
                self.client,
                dataframe,
                destination_table_ref,
                chunksize=chunksize,
                schema=schema,
                location=self.location,
                api_method=api_method,
                write_disposition=write_disposition,
                billing_project=billing_project,
            )
            if progress_bar and tqdm:
                chunks = tqdm.tqdm(chunks)
            for remaining_rows in chunks:
                logger.info(
                    "\r{} out of {} rows loaded.".format(
                        total_rows - remaining_rows, total_rows
                    )
                )
        except self.http_error as ex:
            self.process_http_error(ex)


def _bqschema_to_nullsafe_dtypes(schema_fields):
    """Specify explicit dtypes based on BigQuery schema.

    This function only specifies a dtype when the dtype allows nulls.
    Otherwise, use pandas's default dtype choice.

    See: http://pandas.pydata.org/pandas-docs/dev/missing_data.html
    #missing-data-casting-rules-and-indexing
    """
    import db_dtypes

    # If you update this mapping, also update the table at
    # `docs/reading.rst`.
    dtype_map = {
        "FLOAT": np.dtype(float),
        "INTEGER": "Int64",
        "TIME": db_dtypes.TimeDtype(),
        # Note: Other types such as 'datetime64[ns]' and db_types.DateDtype()
        # are not included because the pandas range does not align with the
        # BigQuery range. We need to attempt a conversion to those types and
        # fall back to 'object' when there are out-of-range values.
    }

    # Amend dtype_map with newer extension types if pandas version allows.
    if FEATURES.pandas_has_boolean_dtype:
        dtype_map["BOOLEAN"] = "boolean"

    dtypes = {}
    for field in schema_fields:
        name = str(field["name"])
        # Array BigQuery type is represented as an object column containing
        # list objects.
        if field["mode"].upper() == "REPEATED":
            dtypes[name] = "object"
            continue

        dtype = dtype_map.get(field["type"].upper())
        if dtype:
            dtypes[name] = dtype

    return dtypes


def _finalize_dtypes(
    df: "pandas.DataFrame", schema_fields: Sequence[Dict[str, Any]]
) -> "pandas.DataFrame":
    """
    Attempt to change the dtypes of those columns that don't map exactly.

    For example db_dtypes.DateDtype() and datetime64[ns] cannot represent
    0001-01-01, but they can represent dates within a couple hundred years of
    1970. See:
    https://github.com/googleapis/python-bigquery-pandas/issues/365
    """
    import db_dtypes
    import pandas.api.types

    # If you update this mapping, also update the table at
    # `docs/reading.rst`.
    dtype_map = {
        "DATE": db_dtypes.DateDtype(),
        "DATETIME": "datetime64[ns]",
        "TIMESTAMP": "datetime64[ns]",
    }

    for field in schema_fields:
        # This method doesn't modify ARRAY/REPEATED columns.
        if field["mode"].upper() == "REPEATED":
            continue

        name = str(field["name"])
        dtype = dtype_map.get(field["type"].upper())

        # Avoid deprecated conversion to timezone-naive dtype by only casting
        # object dtypes.
        if dtype and pandas.api.types.is_object_dtype(df[name]):
            df[name] = df[name].astype(dtype, errors="ignore")

    # Ensure any TIMESTAMP columns are tz-aware.
    df = pandas_gbq.timestamp.localize_df(df, schema_fields)

    return df


def _transform_read_gbq_configuration(configuration):
    """
    For backwards-compatibility, convert any previously client-side only
    parameters such as timeoutMs to the property name expected by the REST API.

    Makes a copy of configuration if changes are needed.
    """

    if configuration is None:
        return None

    timeout_ms = configuration.get("query", {}).get("timeoutMs")
    if timeout_ms is not None:
        # Transform timeoutMs to an actual server-side configuration.
        # https://github.com/googleapis/python-bigquery-pandas/issues/479
        configuration = copy.deepcopy(configuration)
        del configuration["query"]["timeoutMs"]
        configuration["jobTimeoutMs"] = timeout_ms

    return configuration


def read_gbq(
    query_or_table,
    project_id=None,
    index_col=None,
    col_order=None,
    reauth=False,
    auth_local_webserver=True,
    dialect=None,
    location=None,
    configuration=None,
    credentials=None,
    use_bqstorage_api=False,
    max_results=None,
    verbose=None,
    private_key=None,
    progress_bar_type="tqdm",
    dtypes=None,
):
    r"""Load data from Google BigQuery using google-cloud-python

    The main method a user calls to execute a Query in Google BigQuery
    and read results into a pandas DataFrame.

    This method uses the Google Cloud client library to make requests to
    Google BigQuery, documented `here
    <https://googleapis.dev/python/bigquery/latest/index.html>`__.

    See the :ref:`How to authenticate with Google BigQuery <authentication>`
    guide for authentication instructions.

    Parameters
    ----------
    query_or_table : str
        SQL query to return data values. If the string is a table ID, fetch the
        rows directly from the table without running a query.
    project_id : str, optional
        Google Cloud Platform project ID. Optional when available from
        the environment.
    index_col : str, optional
        Name of result column to use for index in results DataFrame.
    col_order : list(str), optional
        List of BigQuery column names in the desired order for results
        DataFrame.
    reauth : boolean, default False
        Force Google BigQuery to re-authenticate the user. This is useful
        if multiple accounts are used.
    auth_local_webserver : bool, default True
        Use the `local webserver flow
        <https://googleapis.dev/python/google-auth-oauthlib/latest/reference/google_auth_oauthlib.flow.html#google_auth_oauthlib.flow.InstalledAppFlow.run_local_server>`_
        instead of the `console flow
        <https://googleapis.dev/python/google-auth-oauthlib/latest/reference/google_auth_oauthlib.flow.html#google_auth_oauthlib.flow.InstalledAppFlow.run_console>`_
        when getting user credentials. Your code must run on the same machine
        as your web browser and your web browser can access your application
        via ``localhost:808X``.

        .. versionadded:: 0.2.0
    dialect : str, default 'standard'
        Note: The default value changed to 'standard' in version 0.10.0.

        SQL syntax dialect to use. Value can be one of:

        ``'legacy'``
            Use BigQuery's legacy SQL dialect. For more information see
            `BigQuery Legacy SQL Reference
            <https://cloud.google.com/bigquery/docs/reference/legacy-sql>`__.
        ``'standard'``
            Use BigQuery's standard SQL, which is
            compliant with the SQL 2011 standard. For more information
            see `BigQuery Standard SQL Reference
            <https://cloud.google.com/bigquery/docs/reference/standard-sql/>`__.
    location : str, optional
        Location where the query job should run. See the `BigQuery locations
        documentation
        <https://cloud.google.com/bigquery/docs/dataset-locations>`__ for a
        list of available locations. The location must match that of any
        datasets used in the query.

        .. versionadded:: 0.5.0
    configuration : dict, optional
        Query config parameters for job processing.
        For example:

            configuration = {'query': {'useQueryCache': False}}

        For more information see `BigQuery REST API Reference
        <https://cloud.google.com/bigquery/docs/reference/rest/v2/jobs#configuration.query>`__.
    credentials : google.auth.credentials.Credentials, optional
        Credentials for accessing Google APIs. Use this parameter to override
        default credentials, such as to use Compute Engine
        :class:`google.auth.compute_engine.Credentials` or Service Account
        :class:`google.oauth2.service_account.Credentials` directly.

        .. versionadded:: 0.8.0
    use_bqstorage_api : bool, default False
        Use the `BigQuery Storage API
        <https://cloud.google.com/bigquery/docs/reference/storage/>`__ to
        download query results quickly, but at an increased cost. To use this
        API, first `enable it in the Cloud Console
        <https://console.cloud.google.com/apis/library/bigquerystorage.googleapis.com>`__.
        You must also have the `bigquery.readsessions.create
        <https://cloud.google.com/bigquery/docs/access-control#roles>`__
        permission on the project you are billing queries to.

        This feature requires the ``google-cloud-bigquery-storage`` and
        ``pyarrow`` packages.

        This value is ignored if ``max_results`` is set.

        .. versionadded:: 0.10.0
    max_results : int, optional
        If set, limit the maximum number of rows to fetch from the query
        results.

        .. versionadded:: 0.12.0
    progress_bar_type (Optional[str]):
        If set, use the `tqdm <https://tqdm.github.io/>`__ library to
        display a progress bar while the data downloads. Install the
        ``tqdm`` package to use this feature.
        Possible values of ``progress_bar_type`` include:

        ``None``
            No progress bar.
        ``'tqdm'``
            Use the :func:`tqdm.tqdm` function to print a progress bar
            to :data:`sys.stderr`.
        ``'tqdm_notebook'``
            Use the :func:`tqdm.tqdm_notebook` function to display a
            progress bar as a Jupyter notebook widget.
        ``'tqdm_gui'``
            Use the :func:`tqdm.tqdm_gui` function to display a
            progress bar as a graphical dialog box.
    dtypes : dict, optional
        A dictionary of column names to pandas ``dtype``. The provided
        ``dtype`` is used when constructing the series for the column
        specified. Otherwise, a default ``dtype`` is used.
    verbose : None, deprecated
        Deprecated in Pandas-GBQ 0.4.0. Use the `logging module
        to adjust verbosity instead
        <https://pandas-gbq.readthedocs.io/en/latest/intro.html#logging>`__.
    private_key : str, deprecated
        Deprecated in pandas-gbq version 0.8.0. Use the ``credentials``
        parameter and
        :func:`google.oauth2.service_account.Credentials.from_service_account_info`
        or
        :func:`google.oauth2.service_account.Credentials.from_service_account_file`
        instead.

    Returns
    -------
    df: DataFrame
        DataFrame representing results of query.
    """
    global context

    if dialect is None:
        dialect = context.dialect

    if dialect is None:
        dialect = "standard"

    _test_google_api_imports()

    if verbose is not None and FEATURES.pandas_has_deprecated_verbose:
        warnings.warn(
            "verbose is deprecated and will be removed in "
            "a future version. Set logging level in order to vary "
            "verbosity",
            FutureWarning,
            stacklevel=2,
        )

    if dialect not in ("legacy", "standard"):
        raise ValueError("'{0}' is not valid for dialect".format(dialect))

    configuration = _transform_read_gbq_configuration(configuration)

    if configuration and "query" in configuration and "query" in configuration["query"]:
        if query_or_table is not None:
            raise ValueError(
                "Query statement can't be specified "
                "inside config while it is specified "
                "as parameter"
            )
        query_or_table = configuration["query"].pop("query")

    connector = GbqConnector(
        project_id,
        reauth=reauth,
        dialect=dialect,
        auth_local_webserver=auth_local_webserver,
        location=location,
        credentials=credentials,
        private_key=private_key,
        use_bqstorage_api=use_bqstorage_api,
    )

    if _is_query(query_or_table):
        final_df = connector.run_query(
            query_or_table,
            configuration=configuration,
            max_results=max_results,
            progress_bar_type=progress_bar_type,
            dtypes=dtypes,
        )
    else:
        final_df = connector.download_table(
            query_or_table,
            max_results=max_results,
            progress_bar_type=progress_bar_type,
            dtypes=dtypes,
        )

    # Reindex the DataFrame on the provided column
    if index_col is not None:
        if index_col in final_df.columns:
            final_df.set_index(index_col, inplace=True)
        else:
            raise InvalidIndexColumn(
                'Index column "{0}" does not exist in DataFrame.'.format(index_col)
            )

    # Change the order of columns in the DataFrame based on provided list
    if col_order is not None:
        if sorted(col_order) == sorted(final_df.columns):
            final_df = final_df[col_order]
        else:
            raise InvalidColumnOrder("Column order does not match this DataFrame.")

    connector.log_elapsed_seconds(
        "Total time taken",
        datetime.now().strftime("s.\nFinished at %Y-%m-%d %H:%M:%S."),
    )

    return final_df


def to_gbq(
    dataframe,
    destination_table,
    project_id=None,
    chunksize=None,
    reauth=False,
    if_exists="fail",
    auth_local_webserver=True,
    table_schema=None,
    location=None,
    progress_bar=True,
    credentials=None,
    api_method: str = "default",
    verbose=None,
    private_key=None,
):
    """Write a DataFrame to a Google BigQuery table.

    The main method a user calls to export pandas DataFrame contents to Google BigQuery table.

    This method uses the Google Cloud client library to make requests to Google BigQuery, documented `here
    <https://googleapis.dev/python/bigquery/latest/index.html>`__.

    See the :ref:`How to authenticate with Google BigQuery <authentication>`
    guide for authentication instructions.

    Parameters
    ----------
    dataframe : pandas.DataFrame
        DataFrame to be written to a Google BigQuery table.
    destination_table : str
        Name of table to be written, in the form ``dataset.tablename`` or
        ``project.dataset.tablename``.
    project_id : str, optional
        Google Cloud Platform project ID. Optional when available from
        the environment.
    chunksize : int, optional
        Number of rows to be inserted in each chunk from the dataframe.
        Set to ``None`` to load the whole dataframe at once.
    reauth : bool, default False
        Force Google BigQuery to re-authenticate the user. This is useful
        if multiple accounts are used.
    if_exists : str, default 'fail'
        Behavior when the destination table exists. Value can be one of:

        ``'fail'``
            If table exists, do nothing.
        ``'replace'``
            If table exists, drop it, recreate it, and insert data.
        ``'append'``
            If table exists, insert data. Create if does not exist.
    auth_local_webserver : bool, default True
        Use the `local webserver flow
        <https://googleapis.dev/python/google-auth-oauthlib/latest/reference/google_auth_oauthlib.flow.html#google_auth_oauthlib.flow.InstalledAppFlow.run_local_server>`_
        instead of the `console flow
        <https://googleapis.dev/python/google-auth-oauthlib/latest/reference/google_auth_oauthlib.flow.html#google_auth_oauthlib.flow.InstalledAppFlow.run_console>`_
        when getting user credentials. Your code must run on the same machine
        as your web browser and your web browser can access your application
        via ``localhost:808X``.

        .. versionadded:: 0.2.0
    table_schema : list of dicts, optional
        List of BigQuery table fields to which according DataFrame
        columns conform to, e.g. ``[{'name': 'col1', 'type':
        'STRING'},...]``.  The ``type`` values must be BigQuery type names.

        - If ``table_schema`` is provided, it may contain all or a subset of
          DataFrame columns. If a subset is provided, the rest will be
          inferred from the DataFrame dtypes.  If ``table_schema`` contains
          columns not in the DataFrame, they'll be ignored.
        - If ``table_schema`` is **not** provided, it will be
          generated according to dtypes of DataFrame columns. See
          `Inferring the Table Schema
          <https://pandas-gbq.readthedocs.io/en/latest/writing.html#writing-schema>`__.
          for a description of the schema inference.

        See `BigQuery API documentation on valid column names
        <https://cloud.google.com/bigquery/docs/schemas#column_names`>__.

        .. versionadded:: 0.3.1
    location : str, optional
        Location where the load job should run. See the `BigQuery locations
        documentation
        <https://cloud.google.com/bigquery/docs/dataset-locations>`__ for a
        list of available locations. The location must match that of the
        target dataset.

        .. versionadded:: 0.5.0
    progress_bar : bool, default True
        Use the library `tqdm` to show the progress bar for the upload,
        chunk by chunk.

        .. versionadded:: 0.5.0
    credentials : google.auth.credentials.Credentials, optional
        Credentials for accessing Google APIs. Use this parameter to override
        default credentials, such as to use Compute Engine
        :class:`google.auth.compute_engine.Credentials` or Service Account
        :class:`google.oauth2.service_account.Credentials` directly.

        .. versionadded:: 0.8.0
    api_method : str, optional
        API method used to upload DataFrame to BigQuery. One of "load_parquet",
        "load_csv". Default "load_parquet" if pandas is version 1.1.0+,
        otherwise "load_csv".

        .. versionadded:: 0.16.0
    verbose : bool, deprecated
        Deprecated in Pandas-GBQ 0.4.0. Use the `logging module
        to adjust verbosity instead
        <https://pandas-gbq.readthedocs.io/en/latest/intro.html#logging>`__.
    private_key : str, deprecated
        Deprecated in pandas-gbq version 0.8.0. Use the ``credentials``
        parameter and
        :func:`google.oauth2.service_account.Credentials.from_service_account_info`
        or
        :func:`google.oauth2.service_account.Credentials.from_service_account_file`
        instead.
    """

    _test_google_api_imports()

    from google.api_core import exceptions as google_exceptions
    from google.cloud import bigquery

    if verbose is not None and FEATURES.pandas_has_deprecated_verbose:
        warnings.warn(
            "verbose is deprecated and will be removed in "
            "a future version. Set logging level in order to vary "
            "verbosity",
            FutureWarning,
            stacklevel=1,
        )

    if api_method == "default":
        # Avoid using parquet if pandas doesn't support lossless conversions to
        # parquet timestamp. See: https://stackoverflow.com/a/69758676/101923
        if FEATURES.pandas_has_parquet_with_lossless_timestamp:
            api_method = "load_parquet"
        else:
            api_method = "load_csv"

    if chunksize is not None:
        if api_method == "load_parquet":
            warnings.warn(
                "chunksize is ignored when using api_method='load_parquet'",
                DeprecationWarning,
                stacklevel=2,
            )
        else:
            warnings.warn(
                "chunksize will be ignored when using api_method='load_csv' in a future version of pandas-gbq",
                PendingDeprecationWarning,
                stacklevel=2,
            )

    if "." not in destination_table:
        raise NotFoundException(
            "Invalid Table Name. Should be of the form 'datasetId.tableId' or "
            "'projectId.datasetId.tableId'"
        )

    if if_exists not in ("fail", "replace", "append"):
        raise ValueError("'{0}' is not valid for if_exists".format(if_exists))

    if_exists_list = ["fail", "replace", "append"]
    dispositions = ["WRITE_EMPTY", "WRITE_TRUNCATE", "WRITE_APPEND"]
    dispositions_dict = dict(zip(if_exists_list, dispositions))

    write_disposition = dispositions_dict[if_exists]

    connector = GbqConnector(
        project_id,
        reauth=reauth,
        auth_local_webserver=auth_local_webserver,
        location=location,
        credentials=credentials,
        private_key=private_key,
    )
    bqclient = connector.client

    destination_table_ref = bigquery.table.TableReference.from_string(
        destination_table, default_project=connector.project_id
    )

    project_id_table = destination_table_ref.project
    dataset_id = destination_table_ref.dataset_id
    table_id = destination_table_ref.table_id

    default_schema = _generate_bq_schema(dataframe)
    # If table_schema isn't provided, we'll create one for you
    if not table_schema:
        table_schema = default_schema
    # It table_schema is provided, we'll update the default_schema to the provided table_schema
    else:
        table_schema = pandas_gbq.schema.update_schema(
            default_schema, dict(fields=table_schema)
        )

    try:
        # Try to get the table
        table = bqclient.get_table(destination_table_ref)
    except google_exceptions.NotFound:
        # If the table doesn't already exist, create it
        table_connector = _Table(
            project_id_table,
            dataset_id,
            location=location,
            credentials=connector.credentials,
        )
        table_connector.create(table_id, table_schema)
    else:
        # Convert original schema (the schema that already exists) to pandas-gbq API format
        original_schema = pandas_gbq.schema.to_pandas_gbq(table.schema)

        # Update the local `table_schema` so mode (NULLABLE/REQUIRED)
        # matches. See: https://github.com/pydata/pandas-gbq/issues/315
        table_schema = pandas_gbq.schema.update_schema(table_schema, original_schema)

    if dataframe.empty:
        # Create the table (if needed), but don't try to run a load job with an
        # empty file. See: https://github.com/pydata/pandas-gbq/issues/237
        return

    connector.load_data(
        dataframe,
        destination_table_ref,
        write_disposition=write_disposition,
        chunksize=chunksize,
        schema=table_schema,
        progress_bar=progress_bar,
        api_method=api_method,
        billing_project=project_id,
    )


def generate_bq_schema(df, default_type="STRING"):
    """DEPRECATED: Given a passed df, generate the associated Google BigQuery
    schema.

    Parameters
    ----------
    df : DataFrame
    default_type : string
        The default big query type in case the type of the column
        does not exist in the schema.
    """
    # deprecation TimeSeries, #11121
    warnings.warn(
        "generate_bq_schema is deprecated and will be removed in " "a future version",
        FutureWarning,
        stacklevel=2,
    )

    return _generate_bq_schema(df, default_type=default_type)


def _generate_bq_schema(df, default_type="STRING"):
    """DEPRECATED: Given a dataframe, generate a Google BigQuery schema.

    This is a private method, but was used in external code to work around
    issues in the default schema generation. Now that individual columns can
    be overridden: https://github.com/pydata/pandas-gbq/issues/218, this
    method can be removed after there is time to migrate away from this
    method."""
    from pandas_gbq import schema

    return schema.generate_bq_schema(df, default_type=default_type)


class _Table(GbqConnector):
    def __init__(
        self,
        project_id,
        dataset_id,
        reauth=False,
        location=None,
        credentials=None,
        private_key=None,
    ):
        self.dataset_id = dataset_id
        super(_Table, self).__init__(
            project_id,
            reauth,
            location=location,
            credentials=credentials,
            private_key=private_key,
        )

    def _table_ref(self, table_id):
        """Return a BigQuery client library table reference"""
        from google.cloud.bigquery import DatasetReference
        from google.cloud.bigquery import TableReference

        return TableReference(
            DatasetReference(self.project_id, self.dataset_id), table_id
        )

    def exists(self, table_id):
        """Check if a table exists in Google BigQuery

        Parameters
        ----------
        table : str
            Name of table to be verified

        Returns
        -------
        boolean
            true if table exists, otherwise false
        """
        from google.api_core.exceptions import NotFound

        table_ref = self._table_ref(table_id)
        try:
            self.client.get_table(table_ref)
            return True
        except NotFound:
            return False
        except self.http_error as ex:
            self.process_http_error(ex)

    def create(self, table_id, schema):
        """Create a table in Google BigQuery given a table and schema

        Parameters
        ----------
        table : str
            Name of table to be written
        schema : str
            Use the generate_bq_schema to generate your table schema from a
            dataframe.
        """
        from google.cloud.bigquery import DatasetReference
        from google.cloud.bigquery import Table
        from google.cloud.bigquery import TableReference

        if self.exists(table_id):
            raise TableCreationError("Table {0} already exists".format(table_id))

        if not _Dataset(self.project_id, credentials=self.credentials).exists(
            self.dataset_id
        ):
            _Dataset(
                self.project_id,
                credentials=self.credentials,
                location=self.location,
            ).create(self.dataset_id)

        table_ref = TableReference(
            DatasetReference(self.project_id, self.dataset_id), table_id
        )
        table = Table(table_ref)
        table.schema = pandas_gbq.schema.to_google_cloud_bigquery(schema)

        try:
            self.client.create_table(table)
        except self.http_error as ex:
            self.process_http_error(ex)

    def delete(self, table_id):
        """Delete a table in Google BigQuery

        Parameters
        ----------
        table : str
            Name of table to be deleted
        """
        from google.api_core.exceptions import NotFound

        table_ref = self._table_ref(table_id)
        try:
            self.client.delete_table(table_ref)
        except NotFound:
            # Ignore 404 error which may occur if table already deleted
            pass
        except self.http_error as ex:
            self.process_http_error(ex)


class _Dataset(GbqConnector):
    def __init__(
        self,
        project_id,
        reauth=False,
        location=None,
        credentials=None,
        private_key=None,
    ):
        super(_Dataset, self).__init__(
            project_id,
            reauth,
            credentials=credentials,
            location=location,
            private_key=private_key,
        )

    def _dataset_ref(self, dataset_id):
        """Return a BigQuery client library dataset reference"""
        from google.cloud.bigquery import DatasetReference

        return DatasetReference(self.project_id, dataset_id)

    def exists(self, dataset_id):
        """Check if a dataset exists in Google BigQuery

        Parameters
        ----------
        dataset_id : str
            Name of dataset to be verified

        Returns
        -------
        boolean
            true if dataset exists, otherwise false
        """
        from google.api_core.exceptions import NotFound

        try:
            self.client.get_dataset(self._dataset_ref(dataset_id))
            return True
        except NotFound:
            return False
        except self.http_error as ex:
            self.process_http_error(ex)

    def create(self, dataset_id):
        """Create a dataset in Google BigQuery

        Parameters
        ----------
        dataset : str
            Name of dataset to be written
        """
        from google.cloud.bigquery import Dataset

        if self.exists(dataset_id):
            raise DatasetCreationError(
                "Dataset {0} already " "exists".format(dataset_id)
            )

        dataset = Dataset(self._dataset_ref(dataset_id))

        if self.location is not None:
            dataset.location = self.location

        try:
            self.client.create_dataset(dataset)
        except self.http_error as ex:
            self.process_http_error(ex)