Managing Data Stored in BigQuery - Google BigQuery Analytics - page 454

$ bq query \
'SELECT __key__.name, __has_error__, __error __
FROM ch11.devices'
+--------------+---------------+-----------+
| __key___name | __has_error__ | __error__ |
+--------------+---------------+-----------+
| H1JInRKWgXoU | false | NULL |
| WOBWlk3th8HW | false | NULL |
| cVDuditLBGN_ | false | NULL |
+--------------+---------------+-----------+
In this case there were no errors encountered when the data was loaded.
The reason errors occur during transformation is that the Datastore type
system and BigQuery type system differ, and this results in some values
being unsupported. For example, BigQuery has a 64 K limit of the string
type, but Datastore has no hard limit on the size of TextProperty fields,
so values are truncated. An error will be added to the record to indicate that
a loss-y transformation was applied to a field.
Finally take a quick look at the actual data that appears in the table:
$ bq query \
'SELECT __key__.name, owner.email, make, model
FROM ch11.devices'
+--------------+-----------------------+-------+-------+
| __key___name | owner_email | make | model
|
+--------------+-----------------------+-------+-------+
| H1JInRKWgXoU | [e-mail] | nexus | s
|
| WOBWlk3th8HW | [e-mail] | nexus | 7
|
| cVDuditLBGN_ | [e-mail] | nexus | 10
|
+--------------+-----------------------+-------+-------+
This is not a particularly interesting query, but it gives you an idea of how
you might use this data inside BigQuery. The sample application has a