This reference describes the Protocol Buffer messages in Prism API 2.4 and how Polypheny-DB uses them. The server checks the major API version during ConnectionRequest and reports its major and minor versions in ConnectionResponse. Check the server version and feature support after connecting. Prism API 2.4 · Server compatibility
Request and response messages
Prism exchanges serialized Protobuf messages over a framed socket connection. Every Request has a positive id and a payload chosen from its type oneof. The server copies the id to each Response for that request. A response with last = true ends that request’s response sequence. A result Frame has its own is_last flag, which indicates whether the result set has more frames to fetch. The first request opens the session with ConnectionRequest. See the development guide for authentication and socket framing. Prism development guide
Error responses
A failed request can end with Response.error_response and last = true. The response carries the original request ID. During direct execution, the server may already have returned a statement ID in an intermediate response. Keep that ID so you can close the statement if the session is still usable.
Prism 2.4 adds ErrorResponse.details to the existing error message. Use ErrorResponse.message as the human-readable diagnostic. Use structured fields for programmatic handling when the server provides them. Response schema · Error details schema
| Message | Field | Number | Meaning |
|---|---|---|---|
Response |
error_response |
256 | Error payload in the response type oneof. |
ErrorResponse |
message |
1 | Human-readable error text. |
ErrorResponse |
details |
2 | Optional ErrorDetails message, available in the 2.4 schema. |
ErrorDetails |
error_code |
1 | Optional signed 32-bit error code from the error source. |
ErrorDetails |
state |
2 | Optional five-character SQLSTATE, including implementation-defined states. |
ErrorDetails |
message |
3 | Optional copy of the outer ErrorResponse.message. |
Clients generated from earlier Prism 2.x schemas can still read the original message field. Clients using 2.4 or later should check whether details is present, then check each optional field individually. An explicit error code of 0 differs from an absent code. Handle absent or unrecognized states as unclassified errors and retain the original diagnostic. Client decoding example
Statements and result lifetime
Direct queries and prepare requests take a language_name, statement text and an optional namespace_name. The language names are sql, mongo and cypher. Each Cypher request must contain exactly one statement. The parser accepts a trailing semicolon and treats semicolons inside strings and comments as part of the text. A multi-statement Cypher request fails before applying any changes.
Named Cypher parameters
Send PrepareStatementRequest in prepare_named_statement_request with language_name = "cypher" and value expressions such as $name. The returned PreparedStatementSignature lists each distinct parameter once, in order of first appearance, with type name ANY. Execute via ExecuteNamedStatementRequest with the returned statement ID. Supply exactly the expected names as keys in NamedParameters.parameters, with the leading dollar sign omitted. Repeated references use the same value. Missing or unexpected keys fail the execution. A value argument can be an explicit ProtoNull.
Parameters stand for values. Write labels, relationship types and identifiers in the query text. Use direct execution for Cypher DDL. Indexed preparation and indexed batches are available for SQL. Cypher uses named preparation for parameterized execution and direct batches for unparameterized writes. Prepared Cypher example
Cypher bindings accept booleans, integers, decimals, finite floating-point numbers, strings, bytes, dates, times, timestamps and nulls. Encode a string-keyed map as ProtoValue.document. Its values may include nested maps, lists and supported scalars. Lists are supported within a map parameter. A Cypher map result carries the same document encoding and can be passed back as a named value parameter. Cypher graph property rules still apply when writing values to a node or relationship.
Batches
A direct batch returns StatementBatchResponse.scalars in input order. Use separate execute requests for queries that return rows, documents or graph elements.
When a direct batch contains Cypher, Polypheny-DB validates every member before running the batch. Each member must contain one statement that produces a scalar update count or command acknowledgment. Run DDL separately from a multi-statement Cypher batch. Transactional writes in such a batch run together, including supported SQL or MongoQL writes. In auto-commit, the server commits on success. With manual commit, successful writes stay in the session transaction. A planning or execution error after the batch starts rolls back that transaction, including earlier pending writes in manual mode.
Close a successful direct batch with CloseStatementRequest using its batch_id. An indexed prepared batch runs one statement with multiple ordered parameter sets and uses the prepared statement ID.
Fetch and close
Check for the presence of StatementResult.frame. If it is absent, read the integer scalar, where zero is a valid result. A direct MongoQL write can return its count in a DocumentFrame, typically in a field named updateCount.
For a paged result, use the statement ID with FetchRequest until Frame.is_last is true. A positive fetch_size limits source rows per page. A flattened graph page can contain several graph elements per source row. A zero fetch size advances no rows and can be used to inspect column metadata before fetching rows. A negative size fetches all remaining rows. The default page size when omitted is 100.
Use CloseResultRequest with a valid statement ID to release an open result and keep its handle. Use CloseStatementRequest to release both. In auto-commit, reading the final page or closing the result ends its transaction once there are no other open results in that transaction.
An open auto-commit result blocks execution of a new Cypher statement. An open Cypher result, including a tabular one, blocks other statement executions. The previous result remains fetchable if the new execution is rejected. A named Cypher prepare only parses the query, so it can run while a result is open. The resulting handle can execute after the open result finishes.
Read-only SQL and MongoQL statements may overlap in auto-commit, using separate transactions for overlapping reads. Finish open results before running a write or DDL. Finish or close open Cypher results and overlapping reads before changing auto-commit. With manual commit, end the transaction explicitly with CommitRequest or RollbackRequest.
Result frames
Dispatch on the Frame.result oneof for each execution. The server selects the encoding from the query’s declared result and the session options. SQL and tabular Cypher results use RelationalFrame. MongoQL document results use DocumentFrame. An execution keeps its frame type and any column metadata across pages, including empty pages. A new execution can choose its own format.
Graph values in result rows
By default, a Cypher projection that returns only graph-typed columns (NODE, EDGE, PATH or GRAPH) uses GraphFrame. This sends a flat stream of nodes and edges in projection and traversal order. It omits null graph values and may send the same element more than once. Read GraphFrame.element for each page. A mixed projection such as RETURN n, n.name uses RelationalFrame and requires the client to advertise Prism 2.3 or later in the 2.x series. Scalar projections also use RelationalFrame.
To keep the row and column structure of graph-only projections, set the client-info property CypherResultFormat to relational. Send ClientInfoProperties in set_client_info_properties_request with properties = {"CypherResultFormat": "relational"}. Use auto for the default selection. Clients need Prism 2.3 graph-value support to select relational. A change affects subsequent executions and leaves open cursors in their original format.
Decode each tabular cell through the ProtoValue.value oneof. Typed node, edge, path and graph cells require Prism 2.3 support, including when the values appear inside lists or maps. Use ProtoValue.null to detect a null cell, such as an unmatched value from OPTIONAL MATCH. ColumnMeta.type_meta.proto_value_type describes the logical type. A Cypher map has logical type MAP and the cell uses document for its string-keyed entries. Prepared Cypher example
Document collection metadata
Send EntitiesRequest with a document namespace in namespace_name. Set entity_pattern to filter collection names or leave it unset to list the collections in the namespace. PolyDBMS uses % for any number of characters and _ for one character in catalog name patterns. Matching collections appear in EntitiesResponse.entities as Entity.document values. Inspect the Entity oneof to distinguish document collections from relational tables. Request schema · Response schema
The server infers field metadata from up to 100 documents per collection and follows nested objects up to 32 path segments. It reports arrays as fields. Treat the paths, types and optionality as observations from that sample. The sample_document_json field, when present, contains one document as JSON text and can include application data. An empty collection still has namespace and collection names, with no observed fields or sample document. Use Protobuf field presence to test whether the sample was supplied. PolyDBMS document metadata implementation
Nested paths use . as a separator. address.city represents two keys, while address\.city represents one key containing a dot. A literal backslash in a key is doubled. PolyDBMS represents a field observed only with null values as null, an integer and floating-point mix as double, and other heterogeneous types as mixed. A field is optional if it was null or missing in at least one sampled document. Field definitions
document_indexes contains the indexed paths reported by the storage adapter. An empty list indicates that the server provided no index metadata. unique = true describes a single-field unique index and applies only to documents covered by the index. A compound index reports its component paths with unique = false. Index definitions
The following Python example constructs a request and decodes an illustrative response locally. Install the released Protobuf bindings first.
python -m pip install polypheny-prism-api==2.4
import json
from org.polypheny.prism import protointerface_pb2 as prism
request = prism.Request(id=1)
request.entities_request.namespace_name = "clinic"
request.entities_request.entity_pattern = "symptoms%"
payload = request.SerializeToString()
assert prism.Request.FromString(payload).WhichOneof("type") == "entities_request"
example = prism.Response(id=request.id, last=True)
document = example.entities_response.entities.add().document
document.namespace_name = "clinic"
document.collection_name = "symptoms"
document.fields.add(field_path="patient_id", type_name="int")
document.sample_document_json = '{"patient_id": 7}'
document.document_indexes.add(field_path="patient_id", index_type="btree", unique=True)
response = prism.Response.FromString(example.SerializeToString())
assert response.id == request.id and response.last
assert response.WhichOneof("type") == "entities_response"
for entity in response.entities_response.entities:
if entity.WhichOneof("Entity") == "document":
collection = entity.document
print(f"{collection.namespace_name}.{collection.collection_name}")
for field in collection.fields:
print(field.field_path, field.type_name, field.is_optional)
if collection.HasField("sample_document_json"):
print(json.loads(collection.sample_document_json))
for index in collection.document_indexes:
print(index.field_path, index.index_type, index.unique)
The Python and JDBC guides show how to request collection metadata from a running server.
Remote Procedure Calls
The table maps each request payload to the response payloads used by PolyDBMS. A failed request returns error_response. Request and response wrappers
| Request | Response |
|---|---|
dbms_version_request · DbmsVersionRequest |
dbms_version_response · DbmsVersionResponse |
default_namespace_request · DefaultNamespaceRequest |
default_namespace_response · DefaultNamespaceResponse |
table_types_request · TableTypesRequest |
table_types_response · TableTypesResponse |
types_request · TypesRequest |
types_response · TypesResponse |
procedures_request · ProceduresRequest |
procedures_response · ProceduresResponse |
functions_request · FunctionsRequest |
functions_response · FunctionsResponse |
namespaces_request · NamespacesRequest |
namespaces_response · NamespacesResponse |
entities_request · EntitiesRequest |
entities_response · EntitiesResponse |
sql_string_functions_request · SqlStringFunctionsRequest |
sql_string_functions_response · MetaStringResponse |
sql_system_functions_request · SqlSystemFunctionsRequest |
sql_system_functions_response · MetaStringResponse |
sql_time_date_functions_request · SqlTimeDateFunctionsRequest |
sql_time_date_functions_response · MetaStringResponse |
sql_numeric_functions_request · SqlNumericFunctionsRequest |
sql_numeric_functions_response · MetaStringResponse |
sql_keywords_request · SqlKeywordsRequest |
sql_keywords_response · MetaStringResponse |
connection_request · ConnectionRequest |
connection_response · ConnectionResponse |
connection_check_request · ConnectionCheckRequest |
connection_check_response · ConnectionCheckResponse |
disconnect_request · DisconnectRequest |
disconnect_response · DisconnectResponse |
client_info_properties_request · ClientInfoPropertiesRequest |
client_info_properties_response · ClientInfoProperties |
set_client_info_properties_request · ClientInfoProperties |
set_client_info_properties_response · ClientInfoPropertiesResponse |
execute_unparameterized_statement_request · ExecuteUnparameterizedStatementRequest |
statement_response · StatementResponse |
execute_unparameterized_statement_batch_request · ExecuteUnparameterizedStatementBatchRequest |
statement_batch_response · StatementBatchResponse |
prepare_indexed_statement_request · PrepareStatementRequest |
prepared_statement_signature · PreparedStatementSignature |
execute_indexed_statement_request · ExecuteIndexedStatementRequest |
statement_result · StatementResult |
execute_indexed_statement_batch_request · ExecuteIndexedStatementBatchRequest |
statement_batch_response · StatementBatchResponse |
prepare_named_statement_request · PrepareStatementRequest |
prepared_statement_signature · PreparedStatementSignature |
execute_named_statement_request · ExecuteNamedStatementRequest |
statement_result · StatementResult |
fetch_request · FetchRequest |
frame · Frame |
close_statement_request · CloseStatementRequest |
close_statement_response · CloseStatementResponse |
commit_request · CommitRequest |
commit_response · CommitResponse |
rollback_request · RollbackRequest |
rollback_response · RollbackResponse |
connection_properties_update_request · ConnectionPropertiesUpdateRequest |
connection_properties_update_response · ConnectionPropertiesUpdateResponse |
close_result_request · CloseResultRequest |
close_result_response · CloseResultResponse |
For direct, unparameterized execution, the server first sends a statement or batch identifier with last = false. The final response carries the result or update counts. Prepared statements return their identifiers from the prepare call. Fetch additional result frames using FetchRequest and the statement identifier.
Message Categories
Files ending in _requests.proto and _responses.proto define the corresponding payloads. The other files define wrappers, result frames, errors and data values.
namespace_meta_requests
Messages related to describing the metadata of namespaces and their entities.
- Proto File:
namespace_meta_requests.proto
transaction_requests
Messages concerning the handling of transactions.
- Proto File:
transaction_requests.proto
namespace_meta_responses
Messages related to describing the metadata of namespaces and their entities.
- Proto File:
namespace_meta_responses.proto
meta_responses
Messages that are used for querying general metadata.
- Proto File:
meta_responses.proto
protointerface
This file contains basic wrapper messages used for the communication between client and server. Examples for such messages are a response and a request wrapper that are used to wrap a specific response or request message. Further messages for error responses are included.
- Proto File:
protointerface.proto
meta_requests
Messages that are used for querying general metadata.
- Proto File:
meta_requests.proto
connection_requests
Messages related to establishing and maintaining a connection.
- Proto File:
connection_requests.proto
statement_responses
Messages related to statements in a query language and their results.
- Proto File:
statement_responses.proto
error
Messages used to send errors to the client.
- Proto File:
error.proto
relational_frame
Relational result frames, rows and column metadata.
- Proto File:
relational_frame.proto
document_frame
Document query results.
- Proto File:
document_frame.proto
graph_frame
Graph query results containing nodes and edges.
- Proto File:
graph_frame.proto
statement_requests
Messages related to statements in a query language and their results.
- Proto File:
statement_requests.proto
transaction_responses
Messages concerning the handling of transactions.
- Proto File:
transaction_responses.proto
value
ProtoValue represents a single value supported by the DBMS. It contains a wide range of fields each corresponding to one supported data type. Only one field can be set at a time.
- Proto File:
value.proto
connection_responses
Messages related to establishing and maintaining a connection.
- Proto File:
connection_responses.proto
Protocol Messages
Message types are grouped by their Protocol Buffer source file.
org/polypheny/prism/namespace_meta_requests.proto
The NamespacesRequest message allows querying for namespaces based on specific patterns in their names as well as their types.
optionalstringnamespace_pattern: Optional name pattern. In PolyDBMS,%matches any number of characters and_matches one.optionalstringnamespace_type: Optional PolyDBMS namespace type,RELATIONAL,DOCUMENTorGRAPH.
The EntitiesRequest message is meant for querying details about entities within a specific namespace, possibly using a pattern to narrow down the results.
stringnamespace_name: The name of the namespace within which entities are to be searched.optionalstringentity_pattern: Optional pattern for collection or table names. In PolyDBMS,%matches any number of characters and_matches one.
org/polypheny/prism/transaction_requests.proto
Represents a request to commit a transaction. This message is used to signal the system to finalize all changes made during the transaction, making them permanent and visible to other transactions.
Represents a request to rollback a transaction. This message is used to signal the system to undo all changes made during the transaction, returning the state of the database to what it was before the transaction began.
org/polypheny/prism/namespace_meta_responses.proto
The NamespacesResponse message delivers a list of namespaces based on the provided criteria in the NamespacesRequest.
repeatedNamespacenamespaces: A list of Namespace messages containing information about each namespace that matches the search criteria.
The EntitiesResponse message returns a list of entities that match the provided criteria from the EntitiesRequest.
repeatedEntityentities: A list of Entity messages containing information about each entity within the requested namespace.
The Namespace message provides detailed information about a specific namespace.
stringnamespace_name: The name of the namespace.boolis_case_sensitive: Indicates if the namespace name is case sensitive.optionalstringnamespace_type: An optional field providing the type of the namespace, if applicable.
The Entity message acts as a generic wrapper for different of entities metadata messages. Only one of the entities metadata fields can be set for a given message instance. Different entity metadata messages are used to represent the metadata of different entity types. The available entity metadata messages are, Table for relational entities, Graph for labeled property graph entities and Document for entities in the document model.
Tabletable: An entity of type Table.Graphgraph: An entity of type Graph.Documentdocument: An entity of type Document.
The Table message provides comprehensive information about a specific table within a database.
stringnamespace_name: The namespace or schema where the table is located.stringtable_name: The name of the table.stringtable_type: The type of the table (e.g., BASE TABLE, VIEW).repeatedColumncolumns: A list of Column messages providing information about each column in the table.optionalPrimaryKeyprimary_key: An optional PrimaryKey message, providing details about the primary key of the table.repeatedForeignKeyforeign_keys: A list of ForeignKey messages detailing the foreign keys associated with the table.repeatedForeignKeyexported_keys: A list of ForeignKey messages detailing the keys in other tables that reference this table.repeatedIndexindexes: A list of Index messages detailing the indexes defined on the table.
The Column message describes the attributes of a specific column within a table.
stringnamespace_name: The namespace or schema of the column’s table.stringtable_name: The table to which this column belongs.stringcolumn_name: The name of the column.stringtype_name: The data type of the column.optionalint32type_length: The length or size of the column type, if applicable.optionalint32type_scale: The scale of the column type, if applicable.boolis_nullable: Indicates if the column can contain null values.optionalstringdefault_value_as_string: The default value of the column, represented as a string.int32column_index: The positional index of the column in the table.optionalstringcollation: The collation name for the column, if applicable.boolis_hidden: Indicates if the column is hidden.ColumnTypecolumn_type: An enumeration indicating the type of column (e.g., REGULAR, CALCULATED, VERSION).
The PrimaryKey message provides details about a table’s primary key.
stringdatabase_name: The database where the primary key’s table is located.stringnamespace_name: The namespace or schema of the primary key’s table.stringtable_name: The table with this primary key.repeatedColumncolumns: The columns that make up the primary key.
The ForeignKey message details a foreign key relationship between tables, providing information about the linkage between a table’s column(s) and the column(s) of another table it references. This relationship is crucial for maintaining referential integrity within the database.
optionalstringreferenced_namespace_name: The namespace or schema of the referenced table. This indicates the specific namespace or schema within the database where the referenced table resides.stringreferenced_table_name: The table being referenced by the foreign key. This field identifies the target table that contains the referenced columns.repeatedColumnreferenced_columns: The columns in the referenced table. These are the columns in the referenced table that the foreign key points to.repeatedColumnforeign_columns: The columns in the current table that make up the foreign key. These columns in the table contain the foreign key constraint reference the corresponding columns in the referenced table.int32update_rule: A numeric representation of the action taken when a referenced row is updated. This rule determines how changes to the referenced columns affect the rows that hold the foreign key.int32delete_rule: A numeric representation of the action taken when a referenced row is deleted. Similar to the update rule, this defines the behavior of the foreign key relationship when rows in the referenced table are deleted.optionalstringkey_name: An optional name for the foreign key constraint. This allows for the naming of the foreign key constraint for easier reference and management within the database.
The Index message provides details about an index defined on a table. Indexes are crucial for improving the performance of data retrieval operations by effectively organizing the data stored in a database table. This message includes information about the index’s database, namespace (or schema), table, uniqueness, name, columns involved, physical location, and type.
stringnamespace_name: The namespace or schema of the indexed table. This indicates the specific namespace or schema within the database where the indexed table resides.stringtable_name: The table on which the index is defined. This field identifies the specific table within the namespace that the index belongs to.boolunique: Indicates if the index enforces unique values. A boolean value where true means the index ensures that all values in the index are unique.stringindex_name: The name of the index. This allows for the naming of the index for easier reference and management within the database.repeatedColumncolumns: The columns that are part of the index. This repeated field lists all the columns included in the index, defining the order and structure of the index.int64location: A numeric identifier representing the location or storage of the index. This field is used to identify where the index is physically stored within the database system, if applicable.int32index_type: A numeric representation indicating the type or mechanism of the index. This could represent different types of indexing strategies used by the database system, such as B-tree, hash, etc.
Reserved for graph entity metadata. Graph query results use the frame and value encodings described in Result frames.
Document
Metadata for a document collection, including field observations from the server and available index metadata.
| Field | Type | Description |
|---|---|---|
namespace_name |
string |
Name of the document namespace. |
collection_name |
string |
Name of the collection. |
fields |
repeated DocumentField |
Known keys and nested object paths. |
sample_document_json |
optional string |
A sample document encoded as JSON, when available. Check field presence before reading it. |
document_indexes |
repeated DocumentIndex |
Index metadata available to the server from the collection’s storage adapters. An empty list indicates that the server provided no index metadata. |
DocumentField
An observed or declared field in a document collection.
| Field | Type | Description |
|---|---|---|
field_path |
string |
Dot-separated path to the field. The server escapes literal dots and backslashes within a key. |
type_name |
string |
The field type, such as string, int, double, bool, array, object, null or mixed. |
is_optional |
bool |
True if the field was missing or null in the sample, or the server declares it optional. False describes the sample and places no constraint on future writes. |
DocumentIndex
Index metadata for one field path in a document collection.
| Field | Type | Description |
|---|---|---|
field_path |
string |
The indexed path uses the same escaping as document field paths. Wildcard segments such as $** retain their index meaning. |
index_type |
string |
The index method, such as btree, hash, text or wildcard. |
unique |
bool |
True for a unique single-field index within the set of documents covered by that index. False for compound index components. Partial and sparse indexes enforce uniqueness within their coverage. |
org/polypheny/prism/meta_responses.proto
The DbmsVersionResponse message relays details about the version and name of the database management system (DBMS) in response to a DbmsVersionRequest.
stringdbms_name: The name of the database management system.stringversion_name: The full version name of the DBMS.int32major_version: The major version number of the DBMS.int32minor_version: The minor version number of the DBMS.
The default namespace
stringdefault_namespace
The TableTypesResponse message conveys information about the various table types present in the system.
repeatedTableTypetable_types: A list of TableType objects detailing each table type present in the system.
The TableType message delineates the nature or category of a table.
stringtable_type: A descriptor denoting the type or category of a table, such as “BASE TABLE”, “VIEW”, or any other classification present in the system.
The TypesResponse message provides a list of native data types supported by the system.
repeatedTypetypes: A list of Type objects detailing each native data type supported by the system.
The Type message describes the characteristics and specifications of a particular native data type.
stringtype_name: The name of the data type.int32precision: The number of total digits for numeric types or the maximum length for string and binary types.optionalstringliteral_prefix: The character(s) used to prefix a literal of this type (e.g., a single quote for strings).optionalstringliteral_suffix: The character(s) used to suffix a literal of this type.boolis_case_sensitive: Indicates whether the data type is case sensitive.int32is_searchable: Indicates how the data type can be used in WHERE clauses.boolis_auto_increment: Indicates whether columns of this data type auto-increment.int32min_scale: The minimum scale for this data type.int32max_scale: The maximum scale for this data type.int32radix: The base number used in representing numeric values.
The UserDefinedType message encapsulates details about a specific user-defined data type.
stringtype_name: The name of the user-defined data type.repeatedValueMetavalue_metas: A list of ValueMeta objects detailing the metadata associated with fields of this data type.
The ValueMeta message offers insights into the metadata of a specific field related to a user-defined type.
stringvalue_name: The name or descriptor of the value’s metadata.
The MetaStringResponse message encapsulates a string that represents some specific metadata information.
stringstring: A string representing the requested metadata information.
The ProceduresResponse message provides a list of stored procedures available within the system.
repeatedProcedureprocedures: A list of Procedure objects detailing each stored procedure present in the system.
The Procedure message encapsulates details about a specific stored procedure in the system.
stringtrivial_name: The simple name of the procedure.stringinput_parameters: The input parameters required by the procedure.stringdescription: A brief description or purpose of the procedure.ReturnTypereturn_type: Enum indicating the return type of the procedure. It can be unspecified, unknown, no result, or result.stringunique_name: A unique name or identifier for the procedure.
The ClientInfoProperties message encapsulates key-value pairs representing client-specific information.
map<string, string>properties: A mapping of property names to their corresponding values, providing insights or settings related to the client.
Acknowledges a successful set_client_info_properties_request. Use client_info_properties_request to read the property values.
ClientInfoPropertyMetaResponse
The ClientInfoPropertyMetaResponse message provides metadata about various client-specific properties.
repeatedClientInfoPropertyMetaclient_info_property_metas: A list of ClientInfoPropertyMeta objects detailing metadata for each client-specific property.
The ClientInfoPropertyMeta message encapsulates metadata about a specific client property.
stringkey: The unique key identifier for the client property.stringdefault_value: The default value of the property.int32maxlength: Maximum length of the property value. Useful for validation purposes.stringdescription: A brief description explaining the purpose or usage of the property.
The FunctionsResponse message provides a list of functions available within the system.
repeatedFunctionfunctions: A list of Function objects detailing each function present in the system.
The Function message encapsulates details about a specific function in the system.
stringname: The name of the function.stringsyntax: The expected syntax for using the function.stringfunction_category: The category to which the function belongs. Helps in categorizing functions for better management and understanding.boolis_table_function: Indicates whether the function is a table function. If true, it means the function returns a table; otherwise, it returns a scalar value.
org/polypheny/prism/protointerface.proto
Carries a request ID and the selected payload in the type oneof.
uint64id: The id of the request. Must be strictly greater than 0.DbmsVersionRequestdbms_version_requestDefaultNamespaceRequestdefault_namespace_requestTableTypesRequesttable_types_requestTypesRequesttypes_requestProceduresRequestprocedures_requestFunctionsRequestfunctions_requestNamespacesRequestnamespaces_requestEntitiesRequestentities_requestSqlStringFunctionsRequestsql_string_functions_requestSqlSystemFunctionsRequestsql_system_functions_requestSqlTimeDateFunctionsRequestsql_time_date_functions_requestSqlNumericFunctionsRequestsql_numeric_functions_requestSqlKeywordsRequestsql_keywords_requestConnectionRequestconnection_requestConnectionCheckRequestconnection_check_requestDisconnectRequestdisconnect_requestClientInfoPropertiesRequestclient_info_properties_requestClientInfoPropertiesset_client_info_properties_requestExecuteUnparameterizedStatementRequestexecute_unparameterized_statement_requestExecuteUnparameterizedStatementBatchRequestexecute_unparameterized_statement_batch_requestPrepareStatementRequestprepare_indexed_statement_requestExecuteIndexedStatementRequestexecute_indexed_statement_requestExecuteIndexedStatementBatchRequestexecute_indexed_statement_batch_requestPrepareStatementRequestprepare_named_statement_requestExecuteNamedStatementRequestexecute_named_statement_requestFetchRequestfetch_requestCloseStatementRequestclose_statement_requestCommitRequestcommit_requestRollbackRequestrollback_requestConnectionPropertiesUpdateRequestconnection_properties_update_requestCloseResultRequestclose_result_request
Payload of an error response.
stringmessage: Human-readable error message, field 1.ErrorDetailsdetails: Optional structured error details, field 2, since Prism 2.4.
Carries a request ID, the end-of-response flag and the selected payload in the type oneof.
uint64id: The id of the request this response is related to.boollast: If a request triggers a stream of responses from the server, this flag represents a response to be the end of the stream.ErrorResponseerror_responseDbmsVersionResponsedbms_version_responseDefaultNamespaceResponsedefault_namespace_responseTableTypesResponsetable_types_responseTypesResponsetypes_responseProceduresResponseprocedures_responseFunctionsResponsefunctions_responseNamespacesResponsenamespaces_responseEntitiesResponseentities_responseMetaStringResponsesql_string_functions_responseMetaStringResponsesql_system_functions_responseMetaStringResponsesql_time_date_functions_responseMetaStringResponsesql_numeric_functions_responseMetaStringResponsesql_keywords_responseConnectionResponseconnection_responseConnectionCheckResponseconnection_check_responseDisconnectResponsedisconnect_responseClientInfoPropertiesclient_info_properties_responseClientInfoPropertiesResponseset_client_info_properties_responseStatementResponsestatement_responseStatementBatchResponsestatement_batch_responseStatementResultstatement_resultPreparedStatementSignatureprepared_statement_signatureFrameframeCloseStatementResponseclose_statement_responseCommitResponsecommit_responseRollbackResponserollback_responseConnectionPropertiesUpdateResponseconnection_properties_update_responseCloseResultResponseclose_result_response
org/polypheny/prism/meta_requests.proto
The DbmsVersionRequest message is utilized to request the version information of the database management system (DBMS) in use. It acts as a trigger for the server to respond with the specific version details of the DBMS. This message does not contain any fields. It simply acts as an indicator to prompt the server for DbmsVersionResponse.
Request the default namespace
The TableTypesRequest message is designed to request information about the different types of tables that are supported or recognized by the DBMS. This message does not contain any fields. It’s a prompt for the server to respond with details about the supported table types in a TableTypesResponse.
The TypesRequest message is deployed to obtain a list of data types supported by the database management system. This helps clients understand the range of data types they can utilize when defining or querying tables. This message does not contain any fields. It simply prompts the server to return a list of supported data types in a TypesResponse.
The SqlStringFunctionsRequest message is used to solicit information about the string functions supported by the SQL implementation of the DBMS. This message is field less and prompts the server to return details of the available string functions in an SqlStringFunctionsResponse.
The SqlSystemFunctionsRequest message aims to retrieve the list of system functions provided by the SQL implementation of the DBMS. Without any fields, this message indicates the server to respond with details about the system functions in an SqlSystemFunctionsResponse.
The SqlTimeDateFunctionsRequest message is dispatched to fetch a list of time and date functions supported by the SQL implementation of the DBMS. This message, being field less, acts as a request for the server to list time and date functions in a SqlTimeDateFunctionsResponse.
The SqlNumericFunctionsRequest message endeavors to obtain details about the numeric functions provided by the SQL implementation of the DBMS. This message, devoid of fields, prompts the server to respond with information about numeric functions in an SqlNumericFunctionsResponse.
The SqlKeywordsRequest message is designed to request a list of reserved keywords utilized by the SQL implementation of the DBMS. With no fields, this message acts as an indicator for the server to provide the list of SQL keywords in an SqlKeywordsResponse.
The ProceduresRequest message is employed to retrieve a list of stored procedures in the database for a specified query language. The client can also narrow down the results by specifying a procedure name pattern.
stringlanguage: The query language for which the procedures are defined. This determines the set of procedures that will be retrieved.optionalstringprocedure_name_pattern: A pattern to filter the names of procedures. For example, providing “get%” might retrieve procedures named getUser, getDetails, etc. If not specified, all procedures for the provided language will be returned. Like in sql, the symbol _ can be used to match a single character.
The ClientInfoPropertiesRequest message facilitates the acquisition of client information properties stored in the database. These properties can offer additional context about the connected client. This message doesn’t possess any fields. It acts as a directive for the server to provide the associated client information properties in a ClientInfoPropertiesResponse.
The FunctionsRequest message is wielded to obtain a list of functions from the database based on the specified query language and function category.
stringquery_language: Defines the query language for which the functions are sought. This field determines the range of functions that will be retrieved.stringfunction_category: Categorizes the function, allowing clients to filter results based on specific categories, such as “NUMERIC”, “STRING”, or “SYSTEM”. This helps in refining the search for specific types of functions.
org/polypheny/prism/connection_requests.proto
Opens a session and supplies the client API version, authentication fields and initial connection properties.
int32major_api_version: Specifies the major version of the API that the client is using.int32minor_api_version: Represents the minor version of the API in use.optionalstringusername: (Optional) The username for authentication when establishing the connection.optionalstringpassword: (Optional) The password associated with the specified username for authentication purposes.optionalConnectionPropertiesconnection_properties: Initial transaction and default namespace settings.
ConnectionPropertiesUpdateRequest
Updates the properties of an open connection.
ConnectionPropertiesconnection_properties: Specifies the new settings for the connection. Contains properties such as transaction auto-commit status and target namespace preference.
The ConnectionProperties message defines specific properties related to the client-server connection. It allows clients to specify certain behaviors and settings for the connection, such as transaction auto-commit status and target namespace preference.
optionalboolis_auto_commit: (Optional) Indicates whether transactions are automatically committed upon execution or require manual commitment.optionalstringnamespace_name: (Optional) Specifies the preferred namespace within the database or system that the client wants to interact with.
Closes the Prism session.
The ConnectionCheckRequest message in combination with the corresponding remote procedure call is utilized to verify the current state of an established connection. It acts as a simple “ping” request, enabling clients to ascertain if the server is responsive and if the connection is still valid.
org/polypheny/prism/statement_responses.proto
Response of the server on the request of a client to close the result set of a specific statement. This message has no fields. It merely serves as an acknowledgement.
Represents the response after executing a statement. This message contains the identifier of the executed statement and optionally, the result of the execution. The result can vary based on the type of statement executed, ranging from scalar values to complex data frames.
int32statement_id: The unique identifier of the executed statement.optionalStatementResultresult: (Optional) The result of the executed statement. This could be a scalar value or a data frame, depending on the type of statement executed.
Encapsulates the result of an executed statement. This can include scalar values for simple queries or data frames for queries that return multiple rows or complex data.
int64scalar: A scalar result value, often used to represent the number of rows affected by an operation.optionalFrameframe: (Optional) A frame of data, which could be a relational table, a graph, or a set of documents.
Represents the response after executing a batch of statements. This includes the batch identifier and a list of scalar results for each statement in the batch, typically representing the number of rows affected by each operation.
int32batch_id: The unique identifier of the executed batch.repeatedint64scalars: List of scalar results for each statement in the batch.
Provides metadata about a prepared statement, including its parameters. This information is crucial for correctly binding parameters to the statement before execution.
int32statement_id: The unique identifier of the prepared statement.repeatedParameterMetaparameter_metas: Metadata about each parameter required by the statement, detailing type, name, and other characteristics.
Provides metadata about a single parameter of a prepared statement. This includes details about the parameter’s precision, scale, data type, and names.
int32precision: Precision of the parameter, typically used for numeric types.int32scale: Scale of the parameter, typically used for decimal types.stringtype_name: The data type of the parameter.optionalstringparameter_name: (Optional) The name of the placeholder specified for this value.stringname: The name of the parameter, if specified.
Represents the response after closing a prepared statement. The message is currently empty, serving as a simple acknowledgment of the action. This message does not contain any fields. It simply serves as an acknowledgment.
Carries relational, graph or document results in the result oneof.
boolis_last: Indicates if this frame is the last in a series of frames, useful for paginated results.RelationalFramerelational_frameGraphFramegraph_frameDocumentFramedocument_frame
org/polypheny/prism/error.proto
Optional structured fields carried by ErrorResponse.details since Prism 2.4.
optional int32error_code: Source-specific numeric error code, field 1.optional stringstate: Five-character SQLSTATE, field 2.optional stringmessage: Copy of the outerErrorResponse.message, field 3.
org/polypheny/prism/relational_frame.proto
The RelationalFrame message encapsulates the structure of a relational frame, which includes metadata about columns and rows of data within a relational table. This structure is essential for representing the schema and data of a table in a structured format, facilitating data manipulation and query operations.
repeatedColumnMetacolumn_meta: A collection of metadata about each column in the frame. This provides details such as index, nullability, length, precision, and type information, defining the schema of the table represented by the relational frame.repeatedRowrows: The rows of data within the frame. Each row consists of a series of ProtoValue objects representing the values in the row, mirroring the actual data stored in the table’s rows according to the schema defined by the column metadata.
The ColumnMeta message holds metadata related to a specific column within a relational table.
int32column_index: Index of the column within the table.boolis_nullable: Indicates whether the column can contain null values.int32length: The length of the column values.stringcolumn_label: Label associated with the column, if any.stringcolumn_name: Name of the column.int32precision: Precision of the column, usually for numerical values.stringentity_name: Name of the entity that the column represents.stringschema_name: Name of the schema that the column belongs to.TypeMetatype_meta: Metadata describing the type of the column.int32scale: Scale of the column, typically for numerical values.stringnamespace: Namespace the column resides in.
The FieldMeta message provides metadata for individual fields, which can be part of complex structures like arrays or structs.
int32field_index: Index of the field within its containing structure.boolis_nullable: Indicates whether the field can contain null values.int32length: Length of the field’s value.stringfield_name: Name of the field.int32precision: Precision of the field, usually for numerical values.TypeMetatype_meta: Metadata describing the type of the field.int32scale: Scale of the field, typically for numerical values.
Describes a result type through ProtoPolyType and optional struct or array metadata.
ProtoPolyTypeproto_value_type: The base type fromProtoPolyType.optionalStructMetastruct_meta: If the proto_value_type indicates a structure, this field provides the metadata for that structure. Contains details about the fields within the structure. This field will be absent for non-structured types.optionalArrayMetaarray_meta: If the proto_value_type indicates an array, this field provides metadata about the type of elements within the array. This field will be absent for non-array types.
The StructMeta message provides metadata specific to structured types, giving details about the fields that make up the structure.
repeatedFieldMetafield_metas: A collection of metadata entries, one for each field within the structure. Each FieldMeta provides details like the field’s name, type, nullability, etc.
The ArrayMeta message provides metadata about array types, detailing the kind of elements they can hold.
TypeMetaelement_type: TypeMeta This field describes the type of elements contained within the array. It uses the TypeMeta message, meaning an array can contain any data type, including nested arrays or structures.
The Row message represents a single row within a relational frame.
repeatedProtoValuevalues: The values within the row, represented as a series of ProtoValue objects.
org/polypheny/prism/document_frame.proto
The DocumentFrame message represents a frame containing one or more documents and is used to relay results or data entries structured as documents from the server to the client.
repeatedProtoDocumentdocuments: A list of documents contained within this frame. Each entry in this list is an instance of ProtoDocument, representing an individual document.
org/polypheny/prism/graph_frame.proto
Carries nodes and edges returned by a graph query. A graph query with a tabular result can return a relational frame. All frames in a result set use the same model.
repeatedGraphElementelement: The nodes and edges in this frame.
Wraps a node or edge in the element oneof.
ProtoNodenode: This field stores the node if the graph element is a node.ProtoEdgeedge: This field stores the edge if the graph element is an edge.
org/polypheny/prism/statement_requests.proto
Closes the result set of an existing statement and keeps its handle. If the result is already closed, the server acknowledges the request. An unknown statement ID returns an error.
int32statement_id: The id of the statement of which the result set should be closed.
ExecuteUnparameterizedStatementRequest
Represents a request to execute a statement without any parameters. This message includes details about the language in which the statement is written, the statement itself, and optional attributes such as fetch size and namespace name to control how the statement is executed.
stringlanguage_name: Specifies the query language of the statement being executed.stringstatement: The statement to be executed.optionalint32fetch_size: (Optional) Number of rows to include into the response message of this call. More rows can be fetched afterwards. If not specified, a default value is used.optionalstringnamespace_name: (Optional) The namespace under which the statement should be executed. If not specified, the namespace specified for the connection is used.
ExecuteUnparameterizedStatementBatchRequest
Encapsulates a batch of unparameterized statements to be executed. This is useful for executing multiple statements in a single call, improving efficiency and performance by reducing the number of round trips to the server.
repeatedExecuteUnparameterizedStatementRequeststatements: List of statements to be executed. Each statement is an instance of ExecuteUnparameterizedStatementRequest, allowing for individual control over the execution parameters of each statement in the batch.
Represents a request to prepare a statement for execution. This message includes the language of the statement, the statement text itself, and optionally, the namespace where the statement should be prepared. Preparing a statement can improve performance for repeated execution.
stringlanguage_name: Specifies the query language of the statement to be prepared.stringstatement: The statement to be prepared.optionalstringnamespace_name: (Optional) The namespace under which the statement should be prepared.
ExecuteIndexedStatementRequest
Represents a request to execute a previously prepared statement by its ID, using indexed parameters. This allows for efficient execution of prepared statements with the provided parameters.
int32statement_id: The unique identifier of the prepared statement to be executed.IndexedParametersparameters: The indexed parameters to be used with the statement.optionalint32fetch_size: (Optional) Number of rows to include into the response message of this call. More rows can be fetched afterwards.
ExecuteIndexedStatementBatchRequest
Encapsulates a batch request to execute a prepared statement multiple times with different sets of indexed parameters. This is useful for executing the same statement with various parameter sets in a single operation, enhancing performance and efficiency.
int32statement_id: The unique identifier of the prepared statement to be executed.repeatedIndexedParametersparameters: List of sets of indexed parameters for multiple executions of the statement.
Represents a request to execute a previously prepared statement by its ID, using named parameters. This approach allows for parameter values to be associated with specific parameter names in the statement, providing clarity and reducing errors in parameter placement.
int32statement_id: The unique identifier of the prepared statement to be executed.NamedParametersparameters: The named parameters to be used with the statement.optionalint32fetch_size: (Optional) Number of rows to include into the response message of this call. More rows can be fetched afterwards.
Encapsulates indexed parameters for a prepared statement. This represents a list of parameter values to be used in the statement, where each parameter’s position in the list corresponds to its position in the statement.
repeatedProtoValueparameters: List of values to be used as parameters, indexed by their position in the list.
Encapsulates named parameters for a prepared statement. This allows for parameters to be specified by name, providing a flexible and clear way to supply parameters to a prepared statement.
map<string, ProtoValue>parameters: Map of parameter names to their corresponding values.
Represents a request to close a statement. This operation is used to release resources allocated to the statement on the server. Closing a statement when it is no longer needed is a good practice to manage resources efficiently.
int32statement_id: The unique identifier of the statement to be closed.
Represents a request to fetch a batch of results for a previously executed statement. This is typically used in conjunction with large result sets, where the results are fetched in multiple batches to manage memory and network resources effectively.
int32statement_id: The unique identifier of the statement whose results are to be fetched.optionalint32fetch_size: (Optional) Number of rows to retrieve in this batch. If no fetch size is specified, a default value is used.
org/polypheny/prism/transaction_responses.proto
Represents the response after attempting to commit a transaction. This message serves as an acknowledgment that the commit request was received and processed, finalizing all changes made during the transaction. This message does not contain any fields. It simply serves as an acknowledgment of the commit.
Represents the response after attempting to rollback a transaction. This message serves as an acknowledgment that the rollback request was received and processed, undoing all changes made during the transaction. This message does not contain any fields. It simply serves as an acknowledgment of the rollback.
org/polypheny/prism/value.proto
ProtoValue represents a single value supported by the DBMS. It contains a wide range of fields each corresponding to one supported data type. Only one field can be set at a time.
ProtoBooleanboolean: Field used to represent a boolean. If set, all other fields must be empty.ProtoIntegerinteger: A 32-bit integer.ProtoLonglong: Field used to represent a long integer of 64 bits. If set, all other fields must be empty.ProtoBigDecimalbig_decimal: A decimal represented by an unscaled integer and a scale.ProtoFloatfloat: Field used to represent a floating-point number of 32 bits. If set, all other fields must be empty.ProtoDoubledouble: Field used to represent a double-precision (64 bit) floating-point number. If set, all other fields must be empty.ProtoDatedate: Field used to represent a date. Dates are stored in days since epoch till the start of the specified day. If set, all other fields must be empty.ProtoTimetime: Field used to represent a time. The value is specified as milliseconds since midnight of a day. If set, all other fields must be empty.ProtoTimestamptimestamp: Field used to represent a timestamp. The value is specified as milliseconds since epoch. If set, all other fields must be empty.ProtoIntervalinterval: Field used to represent a time interval. Intervals are either specified in milliseconds or months. The unit is specified along the value. If set, all other fields must be empty.ProtoStringstring: A character string.ProtoBinarybinary: Field used to represent binary data of arbitrary length. If set, all other fields must be empty.ProtoNullnull: Field used to represent a null value. If set, all other fields must be empty.ProtoListlist: Field used to represent a list of ProtoValues. If set, all other fields must be empty.ProtoDocumentdocument: A document with string keys and Protobuf values.ProtoFilefile: Field used to represent a FILE type such as FILE, IMAGE, VIDEO or AUDIO.ProtoNodenode (20): A node in a labeled property graph.ProtoEdgeedge (21): A relationship with endpoints and a direction.ProtoPathpath (22): An ordered path through a graph.ProtoGraphgraph (23): Sets of nodes and relationships.
Represents a boolean value.
boolboolean: A boolean value.
Represents a 32-bit integer value.
int32integer: A 32-bit integer value.
Represents a 64-bit long integer value.
int64long: A 64-bit long integer value.
An arbitrary-precision decimal represented as an unscaled integer and a scale. Its value is the unscaled integer multiplied by 10 to the power of the negated scale.
int32scale: The scale of the BigDecimal.bytesunscaled_value: The unscaled value of the BigDecimal.
Represents a single precision floating point number.
floatfloat: A single precision floating point number.
Represents a double precision floating point number.
doubledouble: A double precision floating point number.
Represents a date as a 64-bit integer.
int64date: Days since epoch till start of the specified day.
Represents a time value.
int32time: Milliseconds since midnight of an arbitrary day.
Represents a timestamp value.
int64timestamp: The timestamp value in milliseconds since epoch.
Represents an interval value.
int64millisecondsint64months
Represents a string value.
stringstring: A string value of arbitrary length.
Represents binary data.
bytesbinary: Binary data of arbitrary length.
Represents a null value. This message does not contain any fields.
Represents a list of ProtoValue items.
repeatedProtoValuevalues: The list of ProtoValue items.
A document containing string keys and ProtoValue values.
map<string, ProtoValue>entries: The document’s fields and values.
Represents a file containing large binary data.
bytesbinary: The content of the file in binary.
A node in a labeled property graph.
stringid: The system internal unique identifier for that node.stringname: The variable name when available. A data property callednameis stored inproperties.map<string, ProtoValue>properties: A map of key-value pairs of proto values. These are used as a map to store properties with their corresponding names.repeatedstringlabels: This list contains all labels associated with this node. Normal strings are used instead of ProtoStrings.
An edge in a labeled property graph.
stringid: The server identifier for the edge.stringname: The variable name when available.map<string, ProtoValue>properties: A map of key-value pairs of proto values. These are used as a map to store properties with their corresponding names.repeatedstringlabels: The labels associated with the edge.stringleft: The identifier of the left endpoint.stringright: The identifier of the right endpoint.Directiondirection: The edge direction relative to its endpoints.
repeated ProtoNodenodes: Nodes in traversal order. The path can visit a node more than once.repeated ProtoEdgeedges: Edges in traversal order. Edgeiconnects nodesiandi + 1. There is one more node than edge, so a zero-hop path contains just one node. An edge retains its endpoints and direction when traversed in reverse.
repeated ProtoNodenodes: The graph’s nodes, including isolated nodes.repeated ProtoEdgeedges: The graph’s edges. Neither sequence has a guaranteed order.
org/polypheny/prism/connection_responses.proto
The ConnectionResponse message is sent by the server in response to a ConnectionRequest from the client. It provides feedback regarding the compatibility of the client’s API version with the server, as well as other important information about the connection.
boolis_compatible: Indicates whether the client’s API version is compatible with the server. If true, the connection can proceed; if false, the client should consider updating or downgrading its API version.int32major_api_version: The major version number of the server’s API. Helps the client ascertain the API level of the server.int32minor_api_version: The minor version number of the server’s API. Provides finer granularity about the server’s capabilities.
The DisconnectionResponse message is sent by the server in response to a DisconnectRequest from the client. It provides an acknowledgment that the disconnection process has been acknowledged and processed by the server. This message does not contain any fields. It simply serves as an acknowledgment of the disconnection process.
The ConnectionCheckResponse message is sent by the server in response to a ConnectionCheckRequest from the client. It acts as an affirmation that the server is responsive, and the connection is still valid. This message does not contain any fields. Its receipt by the client is an indicator of the server’s responsiveness and the validity of the connection.
ConnectionPropertiesUpdateResponse
The ConnectionPropertiesUpdateResponse message is sent by the server in response to a ConnectionPropertiesUpdateRequest from the client. It acknowledges the success of the requested connection properties update. This message does not contain any fields. It serves as an acknowledgment of the update process for the connection properties.
Enums
Enums and their numeric values, grouped by source file.
org/polypheny/prism/namespace_meta_responses.proto
The kind of column described by Column.column_type. The values identify an unspecified, regular, calculated or version column.
UNSPECIFIED= 0REGULAR= 1CALCULATED= 2VERSION= 3
org/polypheny/prism/meta_responses.proto
The return type enum is part of the message Procedure and indicates the possible types of return values for a stored procedure.
UNSPECIFIED= 0 — Indicates that the return type is unspecified.UNKNOWN= 1 — Indicates that the return type is unknown.NO_RESULT= 2 — The procedure does not return a result.RESULT= 3 — The procedure returns a result.
org/polypheny/prism/value.proto
Specifies the type of a proto value. This enum contains a value for each datatype supported by the DBMS.
UNSPECIFIED= 0 — Default unspecified type.BOOLEAN= 1 — Boolean type, true or false.TINYINT= 2 — 8-bit integer.SMALLINT= 3 — 16-bit integer.INTEGER= 4 — 32-bit integer.BIGINT= 5 — 64-bit integer.DECIMAL= 6 — Arbitrary precision number.REAL= 7 — Single precision floating point.FLOAT= 8 — Alias for REAL.DOUBLE= 9 — Double precision floating point.DATE= 10 — Date value representing a specific day independent of time. Specified in days since epoch till the start of the specified day.TIME= 11 — Time value without timezone in milliseconds since midnight.TIMESTAMP= 13 — Timestamp without timezone. Specified as milliseconds since epoch.INTERVAL= 15 — IntervalCHAR= 28 — Fixed length character string.VARCHAR= 29 — Variable length character string.TEXT= 58 — Text data type.BINARY= 30 — Binary data as byte string.VARBINARY= 31 — Variable length binary data as byte string.NULL= 32 — Null type.ARRAY= 33 — Array type containing elements of one of the ProtoPolyTypes.MAP= 34 — Map type storing key value pairs. Key and value can be arbitrary values of one of the ProtoPolyTypes.DOCUMENT= 35 — Document type similar to a map but used in the document model to represent documents returned as a result.GRAPH= 36 — Graph type. Returned by a query in the graph data model.NODE= 37 — Graph node type. Used inside the graph type to represent a graphs nodes.EDGE= 38 — Graph edge type. Used inside the graph type to represent a graphs edges.PATH= 39 — Graph path type. Used in results of the graph data model to return a the result of graph queries returning a path.IMAGE= 40 — Image data type. Subtype of the file type. Same as binary but containing image data.VIDEO= 41 — Video data type. Subtype of the file type. Same as binary but containing video data.AUDIO= 42 — Audio data type. Subtype of the file type. Same as binary but containing audio data.FILE= 43 — File data type. Same as binary but containing the data of an arbitrary file.DISTINCT= 44 — Distinct type (a named scalar type). Not supported yet!STRUCTURED= 45 — Structured type. Contains an arbitrary set of other ProtoValues of any of the ProtoValueTypes. Not supported yet!ROW= 46 — Row type.OTHER= 47 — Other types not specifically listed.CURSOR= 48 — Cursor type.COLUMN_LIST= 49 — Column list type.DYNAMIC_STAR= 50 — Dynamic star type (used for dynamic column projection).GEOMETRY= 51 — Geometric data type.SYMBOL= 52 — Symbol type.JSON= 53 — JSON data type.MULTISET= 54 — Multiset type.ANY= 55 — Any type (a type that can hold any value).USER_DEFINED_TYPE= 56 — User-defined type.ROW_ID= 57 — Row identifier type.
Direction of a graph edge, relative to its left and right endpoints.
UNSPECIFIED= 0LEFT_TO_RIGHT= 1RIGHT_TO_LEFT= 2NONE= 3