Optional. Optional name of the property to store the result of the aggregation.

If not provided, Datastore will pick a default name following the format property_<incremental_id++>. For example:

AGGREGATE
  COUNT_UP_TO(1) AS count_up_to_1,
  COUNT_UP_TO(2),
  COUNT_UP_TO(3) AS count_up_to_3,
  COUNT(*)
OVER (
  ...
);

becomes:

AGGREGATE
  COUNT_UP_TO(1) AS count_up_to_1,
  COUNT_UP_TO(2) AS property_1,
  COUNT_UP_TO(3) AS count_up_to_3,
  COUNT(*) AS property_2
OVER (
  ...
);

Requires:

• Must be unique across all aggregation aliases.
• Conform to [entity property name][google.datastore.v1.Entity.properties] limitations.

Count aggregator.

Sum aggregator.

Average aggregator.

Optional. Series of aggregations to apply over the results of the nested_query.

Requires:

• A minimum of one and maximum of five aggregations per query.

Nested query for aggregation

The result of the aggregation functions, ex: COUNT(*) AS total_entities.

The key is the [alias][google.datastore.v1.AggregationQuery.Aggregation.alias] assigned to the aggregation function on input and the size of this map equals the number of aggregation functions in the query.

The aggregation results for this batch.

The state of the query after the current batch. Only COUNT(*) aggregations are supported in the initial launch. Therefore, expected result type is limited to NO_MORE_RESULTS.

Read timestamp this batch was returned from.

In a single transaction, subsequent query result batches for the same query can have a greater timestamp. Each batch's read timestamp is valid for all preceding batches.

Required. A list of keys with incomplete key paths for which to allocate IDs. No key may be reserved/read-only.

Required. The ID of the project against which to make the request.

The ID of the database against which to make the request.

'(default)' is not allowed; please use empty string '' to refer the default database.

The keys specified in the request (in the same order), each with its key path completed with a newly allocated ID.

Values in the array. The order of values in an array is preserved as long as all values have identical settings for 'exclude_from_indexes'.

The property to aggregate on.

Required. The ID of the project against which to make the request.

The ID of the database against which to make the request.

'(default)' is not allowed; please use empty string '' to refer the default database.

Options for a new transaction.

The transaction identifier (always present).

The type of commit to perform. Defaults to TRANSACTIONAL.

The mutations to perform.

When mode is TRANSACTIONAL, mutations affecting a single entity are applied in order. The following sequences of mutations affecting a single entity are not permitted in a single Commit request:

• insert followed by insert
• update followed by insert
• upsert followed by insert
• delete followed by update

When mode is NON_TRANSACTIONAL, no two mutations may affect a single entity.

Required. The ID of the project against which to make the request.

The ID of the database against which to make the request.

'(default)' is not allowed; please use empty string '' to refer the default database.

The identifier of the transaction associated with the commit. A transaction identifier is returned by a call to [Datastore.BeginTransaction][google.datastore.v1.Datastore.BeginTransaction].

Options for beginning a new transaction for this request. The transaction is committed when the request completes. If specified, [TransactionOptions.mode][google.datastore.v1.TransactionOptions] must be [TransactionOptions.ReadWrite][google.datastore.v1.TransactionOptions.ReadWrite].

The result of performing the mutations. The i-th mutation result corresponds to the i-th mutation in the request.

The number of index entries updated during the commit, or zero if none were updated.

The transaction commit timestamp. Not set for non-transactional commits.

The operator for combining multiple filters.

The list of filters to combine.

Requires:

• At least one filter is present.

Optional. Optional constraint on the maximum number of entities to count.

This provides a way to set an upper bound on the number of entities to scan, limiting latency, and cost.

Unspecified is interpreted as no bound.

If a zero value is provided, a count result of zero should always be expected.

High-Level Example:

AGGREGATE COUNT_UP_TO(1000) OVER ( SELECT * FROM k );

Requires:

• Must be non-negative when present.

The entity's key.

An entity must have a key, unless otherwise documented (for example, an entity in Value.entity_value may have no key). An entity's kind is its key path's last element's kind, or null if it has no key.

The entity's properties. The map's keys are property names. A property name matching regex __.*__ is reserved. A reserved property name is forbidden in certain documented contexts. The map keys, represented as UTF-8, must not exceed 1,500 bytes and cannot be empty.

The resulting entity.

A cursor that points to the position after the result entity. Set only when the EntityResult is part of a QueryResultBatch message.

The version of the entity, a strictly positive number that monotonically increases with changes to the entity.

This field is set for [FULL][google.datastore.v1.EntityResult.ResultType.FULL] entity results.

For [missing][google.datastore.v1.LookupResponse.missing] entities in LookupResponse, this is the version of the snapshot that was used to look up the entity, and it is always set except for eventually consistent reads.

The time at which the entity was last changed. This field is set for [FULL][google.datastore.v1.EntityResult.ResultType.FULL] entity results. If this entity is missing, this field will not be set.

The time at which the entity was created. This field is set for [FULL][google.datastore.v1.EntityResult.ResultType.FULL] entity results. If this entity is missing, this field will not be set.

A composite filter.

A filter on a property.

A string of the format described here.

When false, the query string must not contain any literals and instead must bind all values. For example, SELECT * FROM Kind WHERE a = 'string literal' is not allowed, while SELECT * FROM Kind WHERE a = @value is.

Numbered binding site @1 references the first numbered parameter, effectively using 1-based indexing, rather than the usual 0.

For each binding site numbered i in query_string, there must be an i-th numbered parameter. The inverse must also be true.

For each non-reserved named binding site in the query string, there must be a named parameter with that name, but not necessarily the inverse.

Key must match regex [A-Za-z_$][A-Za-z_$0-9]*, must not match regex __.*__, and must not be "".

A value parameter.

A query cursor. Query cursors are returned in query result batches.

Entities are partitioned into subsets, currently identified by a project ID and namespace ID. Queries are scoped to a single partition.

The entity path. An entity path consists of one or more elements composed of a kind and a string or numerical identifier, which identify entities. The first element identifies a root entity, the second element identifies a child of the root entity, the third element identifies a child of the second entity, and so forth. The entities identified by all prefixes of the path are called the element's ancestors.

An entity path is always fully complete: all of the entity's ancestors are required to be in the path along with the entity identifier itself. The only exception is that in some documented cases, the identifier in the last path element (for the entity) itself may be omitted. For example, the last path element of the key of Mutation.insert may have no identifier.

A path can never be empty, and a path can have at most 100 elements.

The name of the kind.

The options for this lookup request.

Required. Keys of entities to look up.

Required. The ID of the project against which to make the request.

The ID of the database against which to make the request.

'(default)' is not allowed; please use empty string '' to refer the default database.

Entities found as ResultType.FULL entities. The order of results in this field is undefined and has no relation to the order of the keys in the input.

Entities not found as ResultType.KEY_ONLY entities. The order of results in this field is undefined and has no relation to the order of the keys in the input.

A list of keys that were not looked up due to resource constraints. The order of results in this field is undefined and has no relation to the order of the keys in the input.

The identifier of the transaction that was started as part of this Lookup request.

Set only when [ReadOptions.new_transaction][google.datastore.v1.ReadOptions.new_transaction] was set in [LookupRequest.read_options][google.datastore.v1.LookupRequest.read_options].

The time at which these entities were read or found missing.

The entity to insert. The entity must not already exist. The entity key's final path element may be incomplete.

The entity to update. The entity must already exist. Must have a complete key path.

The entity to upsert. The entity may or may not already exist. The entity key's final path element may be incomplete.

The key of the entity to delete. The entity may or may not already exist. Must have a complete key path and must not be reserved/read-only.

The version of the entity that this mutation is being applied to. If this does not match the current version on the server, the mutation conflicts.

The update time of the entity that this mutation is being applied to. If this does not match the current update time on the server, the mutation conflicts.

The automatically allocated key. Set only when the mutation allocated a key.

The version of the entity on the server after processing the mutation. If the mutation doesn't change anything on the server, then the version will be the version of the current entity or, if no entity is present, a version that is strictly greater than the version of any previous entity and less than the version of any possible future entity.

Whether a conflict was detected for this mutation. Always false when a conflict detection strategy field is not set in the mutation.

The update time of the entity on the server after processing the mutation. If the mutation doesn't change anything on the server, then the timestamp will be the update timestamp of the current entity. This field will not be set after a 'delete'.

The create time of the entity. This field will not be set after a 'delete'.

The ID of the project to which the entities belong.

If not empty, the ID of the database to which the entities belong.

If not empty, the ID of the namespace to which the entities belong.

The kind of the entity.

A kind matching regex __.*__ is reserved/read-only. A kind must not contain more than 1500 bytes when UTF-8 encoded. Cannot be "".

Must be valid UTF-8 bytes. Legacy values that are not valid UTF-8 are encoded as __bytes<X>__ where <X> is the base-64 encoding of the bytes.

The auto-allocated ID of the entity.

Never equal to zero. Values less than zero are discouraged and may not be supported in the future.

The name of the entity.

A name matching regex __.*__ is reserved/read-only. A name must not be more than 1500 bytes when UTF-8 encoded. Cannot be "".

Must be valid UTF-8 bytes. Legacy values that are not valid UTF-8 are encoded as __bytes<X>__ where <X> is the base-64 encoding of the bytes.

The property to project.

The property to filter by.

The operator to filter by.

The value to compare the property to.

The property to order by.

The direction to order by. Defaults to ASCENDING.

A reference to a property.

Requires:

• MUST be a dot-delimited (.) string of segments, where each segment conforms to [entity property name][google.datastore.v1.Entity.properties] limitations.

The projection to return. Defaults to returning all properties.

The kinds to query (if empty, returns entities of all kinds). Currently at most 1 kind may be specified.

The filter to apply.

The order to apply to the query results (if empty, order is unspecified).

The properties to make distinct. The query results will contain the first result for each distinct combination of values for the given properties (if empty, all results are returned).

Requires:

• If order is specified, the set of distinct on properties must appear before the non-distinct on properties in order.

A starting point for the query results. Query cursors are returned in query result batches and can only be used to continue the same query.

An ending point for the query results. Query cursors are returned in query result batches and can only be used to limit the same query.

The number of results to skip. Applies before limit, but after all other constraints. Optional. Must be >= 0 if specified.

The maximum number of results to return. Applies after all other constraints. Optional. Unspecified is interpreted as no limit. Must be >= 0 if specified.

The result type for every entity in entity_results.

The results for this batch.

A cursor that points to the position after the last skipped result. Will be set when skipped_results != 0.

A cursor that points to the position after the last result in the batch.

The state of the query after the current batch.

The number of results skipped, typically because of an offset.

The version number of the snapshot this batch was returned from. This applies to the range of results from the query's start_cursor (or the beginning of the query if no cursor was given) to this batch's end_cursor (not the query's end_cursor).

In a single transaction, subsequent query result batches for the same query can have a greater snapshot version number. Each batch's snapshot version is valid for all preceding batches. The value will be zero for eventually consistent queries.

Read timestamp this batch was returned from. This applies to the range of results from the query's start_cursor (or the beginning of the query if no cursor was given) to this batch's end_cursor (not the query's end_cursor).

In a single transaction, subsequent query result batches for the same query can have a greater timestamp. Each batch's read timestamp is valid for all preceding batches. This value will not be set for eventually consistent queries in Cloud Datastore.

Reads entities at the given time.

This must be a microsecond precision timestamp within the past one hour, or if Point-in-Time Recovery is enabled, can additionally be a whole minute timestamp within the past 7 days.

The non-transactional read consistency to use.

The identifier of the transaction in which to read. A transaction identifier is returned by a call to [Datastore.BeginTransaction][google.datastore.v1.Datastore.BeginTransaction].

Options for beginning a new transaction for this request.

The new transaction identifier will be returned in the corresponding response as either [LookupResponse.transaction][google.datastore.v1.LookupResponse.transaction] or [RunQueryResponse.transaction][google.datastore.v1.RunQueryResponse.transaction].

Reads entities as they were at the given time. This value is only supported for Cloud Firestore in Datastore mode.

This must be a microsecond precision timestamp within the past one hour, or if Point-in-Time Recovery is enabled, can additionally be a whole minute timestamp within the past 7 days.

The transaction identifier of the transaction being retried.

Required. A list of keys with complete key paths whose numeric IDs should not be auto-allocated.

Required. The ID of the project against which to make the request.

The ID of the database against which to make the request.

'(default)' is not allowed; please use empty string '' to refer the default database.

Required. The transaction identifier, returned by a call to [Datastore.BeginTransaction][google.datastore.v1.Datastore.BeginTransaction].

Required. The ID of the project against which to make the request.

The ID of the database against which to make the request.

'(default)' is not allowed; please use empty string '' to refer the default database.

The options for this query.

Entities are partitioned into subsets, identified by a partition ID. Queries are scoped to a single partition. This partition ID is normalized with the standard default context partition ID.

Required. The ID of the project against which to make the request.

The ID of the database against which to make the request.

'(default)' is not allowed; please use empty string '' to refer the default database.

The query to run.

The GQL query to run. This query must be an aggregation query.

A batch of aggregation results. Always present.

The parsed form of the GqlQuery from the request, if it was set.

The identifier of the transaction that was started as part of this RunAggregationQuery request.

Set only when [ReadOptions.new_transaction][google.datastore.v1.ReadOptions.new_transaction] was set in [RunAggregationQueryRequest.read_options][google.datastore.v1.RunAggregationQueryRequest.read_options].

The options for this query.

Entities are partitioned into subsets, identified by a partition ID. Queries are scoped to a single partition. This partition ID is normalized with the standard default context partition ID.

Required. The ID of the project against which to make the request.

The ID of the database against which to make the request.

'(default)' is not allowed; please use empty string '' to refer the default database.

The query to run.

The GQL query to run. This query must be a non-aggregation query.

A batch of query results (always present).

The parsed form of the GqlQuery from the request, if it was set.

The identifier of the transaction that was started as part of this RunQuery request.

Set only when [ReadOptions.new_transaction][google.datastore.v1.ReadOptions.new_transaction] was set in [RunQueryRequest.read_options][google.datastore.v1.RunQueryRequest.read_options].

The property to aggregate on.

The transaction should allow both reads and writes.

The transaction should only allow reads.

The meaning field should only be populated for backwards compatibility.

If the value should be excluded from all indexes including those defined explicitly.

A boolean value.

An integer value.

A double value.

A key value.

An entity value.

• May have no key.
• May have a key with an incomplete key path.
• May have a reserved/read-only key.

A geo point value representing a point on the surface of Earth.

An array value. Cannot contain another array value. A Value instance that sets field array_value must not set fields meaning or exclude_from_indexes.

A timestamp value. When stored in the Datastore, precise only to microseconds; any additional precision is rounded down.

A null value.

A UTF-8 encoded string value. When exclude_from_indexes is false (it is indexed) , may have at most 1500 bytes. Otherwise, may be set to at most 1,000,000 bytes.

A blob value. May have at most 1,000,000 bytes. When exclude_from_indexes is false, may have at most 1500 bytes. In JSON requests, must be base64-encoded.

Unspecified. This value must not be used.

Ascending.

Descending.

Unspecified. This value must not be used.

Transactional: The mutations are either all applied, or none are applied. Learn about transactions here.

Non-transactional: The mutations may not apply as all or none.

Unspecified. This value is never used.

There may be additional batches to fetch from this query.

The query is finished, but there may be more results after the limit.

The query is finished, but there may be more results after the end cursor.

The query is finished, and there are no more results.

Unspecified. This value must not be used.

The results are required to satisfy each of the combined filters.

Documents are required to satisfy at least one of the combined filters.

Unspecified. This value must not be used.

The given property is less than the given value.

Requires:

• That property comes first in order_by.

The given property is less than or equal to the given value.

Requires:

• That property comes first in order_by.

The given property is greater than the given value.

Requires:

• That property comes first in order_by.

The given property is greater than or equal to the given value.

Requires:

• That property comes first in order_by.

The given property is equal to the given value.

The given property is equal to at least one value in the given array.

Requires:

• That value is a non-empty ArrayValue, subject to disjunction limits.
• No NOT_IN is in the same query.

The given property is not equal to the given value.

Requires:

• No other NOT_EQUAL or NOT_IN is in the same query.
• That property comes first in the order_by.

Limit the result set to the given entity and its descendants.

Requires:

• That value is an entity key.
• All evaluated disjunctions must have the same HAS_ANCESTOR filter.

The value of the property is not in the given array.

Requires:

• That value is a non-empty ArrayValue with at most 10 values.
• No other OR, IN, NOT_IN, NOT_EQUAL is in the same query.
• That field comes first in the order_by.

Unspecified. This value must not be used.

Strong consistency.

Eventual consistency.

Unspecified. This value is never used.

The key and properties.

A projected subset of properties. The entity may have no key.

Only the key.