:py:mod:`bigquery.utils`
========================

.. py:module:: bigquery.utils


Module Contents
---------------


Functions
~~~~~~~~~

.. autoapisummary::

   bigquery.utils.flatten
   bigquery.utils.parse
   bigquery.utils.query_response_to_dict



Attributes
~~~~~~~~~~

.. autoapisummary::

   bigquery.utils.log
   bigquery.utils.utc


.. py:data:: log

   

.. py:data:: utc

   

.. py:function:: flatten(x)

   Flatten response objects into something we can actually work with.

   The API returns data of the form:

       {'f': [{'v': ...}]}

   to indicate groupings of fields and their potential for having multiple
   values. We want those to just be plain old objects.


.. py:function:: parse(field, value)

   Parse a given field back to a Python object.

   This is often trivial: convert the value from a string to the type
   specified in the field's schema. There's a couple caveats we've identified
   so far, though:

   * NULLABLE fields should be handled specially, eg. so as not to
     accidentally convert them to the schema type.
   * REPEATED fields are nested a biot differently than expected, so we need
     to flatten *first*, then convert.

   `Field = Dict[str, Union[str, 'Field']]`, but wow is that difficult to
   represent in a backwards-enough compatible fashion.


.. py:function:: query_response_to_dict(response)

   Convert a query response to a dictionary.

   API responses for job queries are packed into a difficult-to-use format.
   This method deserializes a response into a List of rows, with each row
   being a dictionary of field names to the row's value.

   This method also handles converting the values according to the schema
   defined in the response (eg. into builtin python types).