Class BatchSnapshot (2.1.0)

BatchSnapshot(database, read_timestamp=None, exact_staleness=None)

Wrapper for generating and processing read / query batches.

Parameters

Name	Description
database	`Database` database to use
read_timestamp	`datetime.datetime` Execute all reads at the given timestamp.
exact_staleness	`datetime.timedelta` Execute all reads at a timestamp that is `exact_staleness` old.

Inheritance

builtins.object > BatchSnapshot

Methods

close

close()

Clean up underlying session.

.. note::

If the transaction has been shared across multiple machines, calling this on any machine would invalidate the transaction everywhere. Ideally this would be called when data has been read from all the partitions.

execute_sql

execute_sql(*args, **kw)

Convenience method: perform query operation via snapshot.

See execute_sql.

from_dict

from_dict(database, mapping)

Reconstruct an instance from a mapping.

Parameters

Name	Description
database	`Database` database to use
mapping	`mapping` serialized state of the instance

generate_query_batches

generate_query_batches(
    sql,
    params=None,
    param_types=None,
    partition_size_bytes=None,
    max_partitions=None,
    query_options=None,
)

Start a partitioned query operation.

Uses the PartitionQuery API request to start a partitioned query operation. Returns a list of batch information needed to peform the actual queries.

Parameters

Name	Description
sql	`str` SQL query statement
params	`dict, {str -> column value}` values for parameter replacement. Keys must match the names used in `sql`.
param_types	`dict[str -> Union[dict, .types.Type]]` (Optional) maps explicit types for one or more param values; required if parameters are passed.
partition_size_bytes	`int` (Optional) desired size for each partition generated. The service uses this as a hint, the actual partition size may differ.
max_partitions	`int` (Optional) desired maximum number of partitions generated. The service uses this as a hint, the actual number of partitions may differ.
query_options	QueryOptions or `dict` (Optional) Query optimizer configuration to use for the given query. If a dict is provided, it must be of the same form as the protobuf message QueryOptions

Returns

Type	Description
iterable of dict	mappings of information used peform actual partitioned reads via `process_read_batch`.

generate_read_batches

generate_read_batches(
    table, columns, keyset, index="", partition_size_bytes=None, max_partitions=None
)

Start a partitioned batch read operation.

Uses the PartitionRead API request to initiate the partitioned read. Returns a list of batch information needed to perform the actual reads.

Parameters

Name	Description
table	`str` name of the table from which to fetch data
columns	`list of str` names of columns to be retrieved
keyset	`KeySet` keys / ranges identifying rows to be retrieved
index	`str` (Optional) name of index to use, rather than the table's primary key
partition_size_bytes	`int` (Optional) desired size for each partition generated. The service uses this as a hint, the actual partition size may differ.
max_partitions	`int` (Optional) desired maximum number of partitions generated. The service uses this as a hint, the actual number of partitions may differ.

Returns

Type	Description
iterable of dict	mappings of information used peform actual partitioned reads via `process_read_batch`.

process

process(batch)

Process a single, partitioned query or read.

Parameter

Name	Description
batch	`mapping` one of the mappings returned from an earlier call to `generate_query_batches`.

Exceptions

Type	Description
ValueError	if batch does not contain either 'read' or 'query'

Returns

Type	Description
StreamedResultSet	a result set instance which can be used to consume rows.

process_query_batch

process_query_batch(batch)

Process a single, partitioned query.

Parameter

Name	Description
batch	`mapping` one of the mappings returned from an earlier call to `generate_query_batches`.

Returns

Type	Description
StreamedResultSet	a result set instance which can be used to consume rows.

process_read_batch

process_read_batch(batch)

Process a single, partitioned read.

Parameter

Name	Description
batch	`mapping` one of the mappings returned from an earlier call to `generate_read_batches`.

Returns

Type	Description
StreamedResultSet	a result set instance which can be used to consume rows.

read

read(*args, **kw)

Convenience method: perform read operation via snapshot.

See read.

to_dict

to_dict()

Return state as a dictionary.

Result can be used to serialize the instance and reconstitute it later using from_dict.