Query modules Python API

This is the API documentation for mgp.py that contains definitions of the public Python API provided by Memgraph. In essence, this is a wrapper around the C API. This source file can be found in the Memgraph installation directory, under /usr/lib/memgraph/python_support.

tip

For an example of how to implement query modules in Python, take a look at the example we provided.

tip

If you install any Python modules after running Memgraph, you'll have to load them into Memgraph or restart Memgraph in order to use them.

You can also develop query modules in Python from Memgraph Lab (v2.0 and newer). Just navigate to Query Modules and click on New Module to start.

mgp.read_proc(func: Callable[[…], mgp.Record])

Register func as a read-only procedure of the current module.

read_proc is meant to be used as a decorator function to register module procedures. The registered func needs to be a callable which optionally takes ProcCtx as the first argument. Other arguments of func will be bound to values passed in the Cypher query. The full signature of func needs to be annotated with types. The return type must be Record(field_name=type, …) and the procedure must produce either a complete Record or None. To mark a field as deprecated, use Record(field_name=Deprecated(type), …). Multiple records can be produced by returning an iterable of them. Registering generator functions is currently not supported.

Example usage

    import mgp

    @mgp.read_proc
    def procedure(context: mgp.ProcCtx,
                required_arg: mgp.Nullable[mgp.Any],
                optional_arg: mgp.Nullable[mgp.Any] = None
                ) -> mgp.Record(result=str, args=list):
        args = [required_arg, optional_arg]
        # Multiple rows can be produced by returning an iterable of mgp.Record
        return mgp.Record(args=args, result='Hello World!')

The example procedure above returns 2 fields: args and result.

• args is a copy of arguments passed to the procedure.
• result is the result of this procedure, a “Hello World!” string.

Any errors can be reported by raising an Exception.

The procedure can be invoked in Cypher using the following calls:

CALL example.procedure(1, 2) YIELD args, result;
CALL example.procedure(1) YIELD args, result;

Naturally, you may pass in different arguments or yield less fields.

tip

Install the mgp Python module so your editor can use typing annotations properly and suggest methods and classes it contains. You can install the module by running pip install mgp.

mgp.write_proc(func: Callable[[…], mgp.Record])

Register func as a writeable procedure of the current module.

write_proc is meant to be used as a decorator function to register module procedures. The registered func needs to be a callable which optionally takes ProcCtx as the first argument. Other arguments of func will be bound to values passed in the Cypher query. The full signature of func needs to be annotated with types. The return type must be Record(field_name=type, …) and the procedure must produce either a complete Record or None. To mark a field as deprecated, use Record(field_name=Deprecated(type), …). Multiple records can be produced by returning an iterable of them. Registering generator functions is currently not supported.

Example usage

import mgp

@mgp.write_proc
def procedure(context: mgp.ProcCtx,
    required_arg: str,
    optional_arg: mgp.Nullable[str] = None
    ) -> mgp.Record(result=mgp.Vertex):

    vertex = context.graph.create_vertex()
    vertex_properties = vertex.properties
    vertex_properties[“required_arg”] = required_arg

    if optional_arg is not None:
        vertex_properties[“optional_arg”] = optional_arg

    return mgp.Record(result=vertex)

The example procedure above returns a newly created vertex which has at most 2 properties:

• required_arg is always present and its value is the first argument of the procedure.
• optional_arg is present if the second argument of the procedure is not null.

Any errors can be reported by raising an Exception.

The procedure can be invoked in Cypher using the following calls:

CALL example.procedure(“property value”, “another one”) YIELD result;
CALL example.procedure(“single argument”) YIELD result;

Naturally, you may pass in different arguments.

mgp.function(func: Callable[[…]])

Register func as a Memgraph function in the current module.

function is meant to be used as a decorator function to register module functions. The registered func needs to be a callable which optionally takes FuncCtx as the first argument. Other arguments of func will be bound to values passed in the Cypher query. Only the funcion arguments need to be annotated with types. The return type doesn't need to be specified, but it has to be supported by mgp.Any. Registering generator functions is currently not supported.

Example usage

import mgp

@mgp.function
def func_example(context: mgp.FuncCtx,
    required_arg: str,
    optional_arg: mgp.Nullable[str] = None
    ):

    return_args = [required_arg]

    if optional_arg is not None:
        return_args.append(optional_arg)

    # Return any kind of result supported by mgp.Any
    return return_args

The example function above returns a list of provided arguments:

• required_arg is always present and its value is the first argument of the function.
• optional_arg is present if the second argument of the function is not null.

Any errors can be reported by raising an Exception.

The function can be invoked in Cypher using the following calls:

RETURN example.func_example("first argument", "second_argument");
RETURN example.func_example("first argument");

Naturally, you may pass in different arguments.

This module provides the API for usage in custom openCypher procedures.

Label Objects

class Label()

Label of a Vertex.

name

@property
def name() -> str

Get the name of the label.

Returns:

A string that represents the name of the label.

Example:

label.name

Properties Objects

class Properties()

A collection of properties either on a Vertex or an Edge.

get()

def get(property_name: str, default=None) -> object

Get the value of a property with the given name or return default value.

Arguments:

• property_name - String that represents property name.
• default - Default value return if there is no property.

Returns:

Any object value that property under property_name has or default value otherwise.

Raises:

• InvalidContextError - If edge or vertex is out of context.
• UnableToAllocateError - If unable to allocate a mgp.Value.
• DeletedObjectError - If the object has been deleted.

Examples:

vertex.properties.get(property_name)
edge.properties.get(property_name)

set()

def set(property_name: str, value: object) -> None

Set the value of the property. When the value is None, then the property is removed.

Arguments:

• property_name - String that represents property name.
• value - Object that represents value to be set.

Raises:

• UnableToAllocateError - If unable to allocate memory for storing the property.
• ImmutableObjectError - If the object is immutable.
• DeletedObjectError - If the object has been deleted.
• SerializationError - If the object has been modified by another transaction.
• ValueConversionError - If value is vertex, edge or path.

Examples:

vertex.properties.set(property_name, value)
edge.properties.set(property_name, value)

items()

def items() -> typing.Iterable[Property]

Iterate over the properties. Doesn’t return a dynamic view of the properties but copies the current properties.

Returns:

Iterable Property of names and values.

Raises:

• InvalidContextError - If edge or vertex is out of context.
• UnableToAllocateError - If unable to allocate an iterator.
• DeletedObjectError - If the object has been deleted.

Examples:

items = vertex.properties.items()
for it in items:
  name = it.name
  value = it.value

items = edge.properties.items()
for it in items:
  name = it.name
  value = it.value

keys()

def keys() -> typing.Iterable[str]

Iterate over property names. Doesn’t return a dynamic view of the property names but copies the name of the current properties.

Returns:

Iterable list of strings that represent names/keys of properties.

Raises:

• InvalidContextError - If edge or vertex is out of context.
• UnableToAllocateError - If unable to allocate an iterator.
• DeletedObjectError - If the object has been deleted.

Examples:

graph.vertex.properties.keys()
graph.edge.properties.keys()

values()

def values() -> typing.Iterable[object]

Iterate over property values. Doesn’t return a dynamic view of the property values but copies the value of the current properties.

Returns:

Iterable list of property values.

Raises:

• InvalidContextError - If edge or vertex is out of context.
• UnableToAllocateError - If unable to allocate an iterator.
• DeletedObjectError - If the object has been deleted.

Examples:

vertex.properties.values()
edge.properties.values()

__len__

def __len__() -> int

Get the number of properties.

Returns:

A number of properties on vertex or edge.

Raises:

• InvalidContextError - If edge or vertex is out of context.
• UnableToAllocateError - If unable to allocate an iterator.
• DeletedObjectError - If the object has been deleted.

Examples:

len(vertex.properties)
len(edge.properties)

__iter__

def __iter__() -> typing.Iterable[str]

Iterate over property names.

Returns:

Iterable list of strings that represent names of properties.

Raises:

• InvalidContextError - If edge or vertex is out of context.
• UnableToAllocateError - If unable to allocate an iterator.
• DeletedObjectError - If the object has been deleted.

Examples:

iter(vertex.properties)
iter(edge.properties)

__getitem__

def __getitem__(property_name: str) -> object

Get the value of a property with the given name or raise KeyError.

Arguments:

• property_name - String that represents property name.

Returns:

Any value that property under property_name have.

Raises:

• InvalidContextError - If edge or vertex is out of context.
• UnableToAllocateError - If unable to allocate a mgp.Value.
• DeletedObjectError - If the object has been deleted.

Examples:

vertex.properties[property_name]
edge.properties[property_name]

__setitem__

def __setitem__(property_name: str, value: object) -> None

Set the value of the property. When the value is None, then the property is removed.

Arguments:

• property_name - String that represents property name.
• value - Object that represents value to be set.

Raises:

• UnableToAllocateError - If unable to allocate memory for storing the property.
• ImmutableObjectError - If the object is immutable.
• DeletedObjectError - If the object has been deleted.
• SerializationError - If the object has been modified by another transaction.
• ValueConversionError - If value is vertex, edge or path.

Examples:

vertex.properties[property_name] = value
edge.properties[property_name] = value

__contains__

def __contains__(property_name: str) -> bool

Check if there is a property with the given name.

Arguments:

• property_name - String that represents property name

Returns:

Bool value that depends if there is with a given name.

Raises:

• InvalidContextError - If edge or vertex is out of context.
• UnableToAllocateError - If unable to allocate a mgp.Value.
• DeletedObjectError - If the object has been deleted.

Examples:

if property_name in vertex.properties:

if property_name in edge.properties:

EdgeType Objects

class EdgeType()

Type of an Edge.

name

@property
def name() -> str

Get the name of EdgeType.

Returns:

A string that represents the name of EdgeType.

Example:

edge.type.name

Edge Objects

class Edge()

Edge in the graph database.

Access to an Edge is only valid during a single execution of a procedure in a query. You should not globally store an instance of an Edge. Using an invalid Edge instance will raise InvalidContextError.

is_valid()

def is_valid() -> bool

Check if edge is in a valid context and may be used.

Returns:

A bool value depends on if the edge is in a valid context.

Examples:

edge.is_valid()

underlying_graph_is_mutable()

def underlying_graph_is_mutable() -> bool

Check if the graph can be modified.

Returns:

A bool value depends on if the graph is mutable.

Examples:

edge.underlying_graph_is_mutable()

id

@property
def id() -> EdgeId

Get the ID of the edge.

Returns:

EdgeId represents ID of the edge.

Raises:

• InvalidContextError - If edge is out of context.

Examples:

edge.id

type

@property
def type() -> EdgeType

Get the type of edge.

Returns:

EdgeType describing the type of edge.

Raises:

• InvalidContextError - If edge is out of context.

Examples:

edge.type

from_vertex()

@property
def from_vertex() -> "Vertex"

Get the source vertex.

Returns:

Vertex from where the edge is directed.

Raises:

• InvalidContextError - If edge is out of context.

Examples:

edge.from_vertex

to_vertex()

@property
def to_vertex() -> "Vertex"

Get the destination vertex.

Returns:

Vertex to where the edge is directed.

Raises:

• InvalidContextError - If edge is out of context.

Examples:

edge.to_vertex

properties

@property
def properties() -> Properties

Get the properties of the edge.

Returns:

All Properties of edge.

Raises:

• InvalidContextError - If edge is out of context.

Examples:

edge.properties

__eq__

def __eq__(other) -> bool

Raise InvalidContextError.

Vertex Objects

class Vertex()

Vertex in the graph database.

Access to a Vertex is only valid during a single execution of a procedure in a query. You should not globally store an instance of a Vertex. Using an invalid Vertex instance will raise InvalidContextError.

is_valid()

def is_valid() -> bool

Checks if Vertex is in valid context and may be used.

Returns:

A bool value depends on if the Vertex is in a valid context.

Examples:

vertex.is_valid()

underlying_graph_is_mutable()

def underlying_graph_is_mutable() -> bool

Check if the graph is mutable.

Returns:

A bool value depends on if the graph is mutable.

Examples:

vertex.underlying_graph_is_mutable()

id

@property
def id() -> VertexId

Get the ID of the Vertex.

Returns:

VertexId represents ID of the vertex.

Raises:

• InvalidContextError - If vertex is out of context.

Examples:

vertex.id

labels

@property
def labels() -> typing.Tuple[Label]

Get the labels of the vertex.

Returns:

A tuple of Label representing vertex Labels

Raises:

• InvalidContextError - If vertex is out of context.
• OutOfRangeError - If some of the labels are removed while collecting the labels.
• DeletedObjectError - If Vertex has been deleted.

Examples:

vertex.labels

add_label()

def add_label(label: str) -> None

Add the label to the vertex.

Arguments:

• label - String label to be added.

Raises:

• InvalidContextError - If Vertex is out of context.
• UnableToAllocateError - If unable to allocate memory for storing the label.
• ImmutableObjectError - If Vertex is immutable.
• DeletedObjectError - If Vertex has been deleted.
• SerializationError - If Vertex has been modified by another transaction.

Examples:

vertex.add_label(label)

remove_label()

def remove_label(label: str) -> None

Remove the label from the vertex.

Arguments:

• label - String label to be deleted

Raises:

• InvalidContextError - If Vertex is out of context.
• ImmutableObjectError - If Vertex is immutable.
• DeletedObjectError - If Vertex has been deleted.
• SerializationError - If Vertex has been modified by another transaction.

Examples:

vertex.remove_label(label)

properties

@property
def properties() -> Properties

Get the properties of the vertex.

Returns:

Properties on a current vertex.

Raises:

• InvalidContextError - If Vertex is out of context.

Examples:

vertex.properties

in_edges

@property
def in_edges() -> typing.Iterable[Edge]

Iterate over inbound edges of the vertex. When the first parameter to a procedure is a projected graph, iterating will start over the inbound edges of the given vertex in the projected graph. Doesn’t return a dynamic view of the edges but copies the current inbound edges.

Returns:

Iterable list of Edge objects that are directed in towards the current vertex.

Raises:

• InvalidContextError - If Vertex is out of context.
• UnableToAllocateError - If unable to allocate an iterator.
• DeletedObjectError - If Vertex has been deleted.

Examples:

for edge in vertex.in_edges:

out_edges

@property
def out_edges() -> typing.Iterable[Edge]

Iterate over outbound edges of the vertex. When the first parameter to a procedure is a projected graph, iterating will start over the outbound edges of the given vertex in the projected graph.

Doesn’t return a dynamic view of the edges but copies the current outbound edges.

Returns:

Iterable list of Edge objects that are directed out of the current vertex.

Raises:

• InvalidContextError - If Vertex is out of context.
• UnableToAllocateError - If unable to allocate an iterator.
• DeletedObjectError - If Vertex has been deleted.

Examples:

for edge in vertex.out_edges:

__eq__

def __eq__(other) -> bool

Raise InvalidContextError

Path Objects

class Path()

Path containing Vertex and Edge instances.

__init__

def __init__(starting_vertex_or_path: typing.Union[_mgp.Path, Vertex])

Initialize with a starting Vertex.

Raises:

• InvalidContextError - If passed in Vertex is invalid.
• UnableToAllocateError - If cannot allocate a path.

is_valid()

def is_valid() -> bool

Check if Path is in valid context and may be used.

Returns:

A bool value depends on if the Path is in a valid context.

Examples:

path.is_valid()

expand()

def expand(edge: Edge)

Append an edge continuing from the last vertex on the path.

The last vertex on the path will become the other endpoint of the given edge, as continued from the current last vertex.

Arguments:

• edge - Edge that is added to the path

Raises:

• InvalidContextError - If using an invalid Path instance or if passed in Edge is invalid.
• LogicErrorError - If the current last vertex in the path is not part of the given edge.
• UnableToAllocateError - If unable to allocate memory for path extension.

Examples:

path.expand(edge)

vertices

@property
def vertices() -> typing.Tuple[Vertex, ...]

Vertices are ordered from the start to the end of the path.

Returns:

A tuple of Vertex objects order from start to end of the path.

Raises:

• InvalidContextError - If using an invalid Path instance.

Examples:

path.vertices

edges

@property
def edges() -> typing.Tuple[Edge, ...]

Edges are ordered from the start to the end of the path.

Returns:

A tuple of Edge objects order from start to end of the path

Raises:

• InvalidContextError - If using an invalid Path instance.

Examples:

path.edges

Record Objects

class Record()

Represents a record of resulting field values.

__init__

def __init__(**kwargs)

Initialize with name=value fields in kwargs.

Vertices Objects

class Vertices()

Iterable over vertices in a graph.

is_valid()

def is_valid() -> bool

Check if Vertices is in valid context and may be used.

Returns:

A bool value depends on if the Vertices is in valid context.

Examples:

vertices.is_valid()

__iter__

def __iter__() -> typing.Iterable[Vertex]

Iterate over vertices.

Returns:

Iterable list of Vertex objects.

Raises:

• InvalidContextError - If context is invalid.
• UnableToAllocateError - If unable to allocate an iterator or a vertex.

Examples:

for vertex in graph.vertices

__contains__

def __contains__(vertex)

Check if Vertices contain the given vertex.

Arguments:

• vertex - Vertex to be checked if it is a part of graph Vertices.

Returns:

Bool value depends if there is Vertex in graph Vertices.

Raises:

• UnableToAllocateError - If unable to allocate the vertex.

Examples:

if vertex in graph.vertices:

__len__

def __len__()

Get the number of vertices.

Returns:

A number of vertices in the graph.

Raises:

• InvalidContextError - If context is invalid.
• UnableToAllocateError - If unable to allocate an iterator or a vertex.

Examples:

len(graph.vertices)

Graph Objects

class Graph()

State of the graph database in current ProcCtx.

is_valid()

def is_valid() -> bool

Check if graph is in a valid context and may be used.

Returns:

A bool value depends on if the graph is in a valid context.

Examples:

graph.is_valid()

get_vertex_by_id()

def get_vertex_by_id(vertex_id: VertexId) -> Vertex

Return the Vertex corresponding to the given vertex_id from the graph. When the first parameter to a procedure is a projected graph, the vertex must also exist in the projected graph.

Access to a Vertex is only valid during a single execution of a procedure in a query. You should not globally store the returned Vertex.

Arguments:

• vertex_id - Memgraph Vertex ID

Returns:

Vertexcorresponding to vertex_id

Raises:

• IndexError - If unable to find the given vertex_id.
• InvalidContextError - If context is invalid.

Examples:

graph.get_vertex_by_id(vertex_id)

vertices

@property
def vertices() -> Vertices

Get all vertices in the graph.

Access to a Vertex is only valid during a single execution of a procedure in a query. You should not globally store the returned Vertex instances.

Returns:

Vertices that contained in the graph.

Raises:

• InvalidContextError - If context is invalid.

Examples:

Iteration over all graph Vertices.

graph = context.graph
for vertex in graph.vertices:

is_mutable()

def is_mutable() -> bool

Check if the graph is mutable. Thus it can be used to modify vertices and edges.

Returns:

A bool value that depends if the graph is mutable or not.

Examples:

graph.is_mutable()

create_vertex()

def create_vertex() -> Vertex

Create an empty vertex. When the first parameter to a procedure is a projected graph, the vertex is also added to the projected graph view.

Returns:

Created Vertex.

Raises:

• ImmutableObjectError - If graph is immutable.
• UnableToAllocateError - If unable to allocate a vertex.

Examples

vertex = graph.create_vertex()

delete_vertex()

def delete_vertex(vertex: Vertex) -> None

Delete a vertex if there are no edges. When the first parameter to a procedure is a projected graph, the vertex must also exist in the projected graph.

Arguments:

• vertex - Vertex to be deleted

Raises:

• ImmutableObjectError - If graph is immutable.
• LogicErrorError - If vertex has edges.
• SerializationError - If vertex has been modified by another transaction.

Examples:

graph.delete_vertex(vertex)

detach_delete_vertex()

def detach_delete_vertex(vertex: Vertex) -> None

Delete a vertex and all of its edges. When the first parameter to a procedure is a projected graph, such an operation is not possible.

Arguments:

• vertex - Vertex to be deleted with all of its edges

Raises:

• ImmutableObjectError - If graph is immutable.
• SerializationError - If vertex has been modified by another transaction.

Examples:

graph.detach_delete_vertex(vertex)

create_edge()

def create_edge(from_vertex: Vertex, to_vertex: Vertex,
                edge_type: EdgeType) -> None

Create an edge. When the first parameter is a projected graph, it will create a new directed edge with a specified label only if both vertices are a part of the projected graph.

Arguments:

• from_vertex - Vertex from where edge is directed.
• to_vertex - Vertex to where edge is directed.
• edge_type - EdgeType defines the type of edge.

Raises:

• ImmutableObjectError - If graph is immutable.
• UnableToAllocateError - If unable to allocate an edge.
• DeletedObjectError - If from_vertex or to_vertex has been deleted.
• SerializationError - If from_vertex or to_vertex has been modified by another transaction.

Examples:

graph.create_edge(from_vertex, vertex, edge_type)

delete_edge()

def delete_edge(edge: Edge) -> None

Delete an edge. When the first parameter to a procedure is a projected graph, the edge must also exist in the projected graph.

Arguments:

• edge - Edge to be deleted

Raises:

• ImmutableObjectError: If graph is immutable.
• Raise SerializationError: If edge, its source or destination vertex has been modified by another transaction.

Examples

graph.delete_edge(edge)

AbortError Objects

class AbortError(Exception)

Signals that the procedure was asked to abort its execution.

ProcCtx Objects

class ProcCtx()

Context of a procedure being executed.

Access to a ProcCtx is only valid during a single execution of a procedure in a query. You should not globally store a ProcCtx instance.

graph

@property
def graph() -> Graph

Access to Graph object.

Returns:

Graph object.

Raises:

• InvalidContextError - If context is invalid.

Examples:

context.graph

Logger Objects

class Logger()

Class for logging.

info()

def info(out: str) -> None

Logs a message out on INFO log level.

Arguments:

• out - str to be logged

Examples

logger.info("message")

debug()

def debug(out: str) -> None

Logs a message out on DEBUG log level.

Arguments:

• out - str to be logged

Examples

logger.debug("message")

error()

def error(out: str) -> None

Logs a message out on ERROR log level.

Arguments:

• out - str to be logged

Examples

logger.error("message")

trace()

def trace(out: str) -> None

Logs a message out on TRACE log level.

Arguments:

• out - str to be logged

Examples

logger.trace("message")

warning()

def warning(out: str) -> None

Logs a message out on WARNING log level.

Arguments:

• out - str to be logged

Examples

logger.warning("message")

critical()

def critical(out: str) -> None

Logs a message out on CRITICAL log level.

Arguments:

• out - str to be logged

Examples

logger.critical("message")

UnsupportedTypingError Objects

class UnsupportedTypingError(Exception)

Signals a typing annotation is not supported as a _mgp.CypherType.

Deprecated Objects

class Deprecated()

Annotate a resulting Record's field as deprecated.

read_proc()

def read_proc(func: typing.Callable[..., Record])

Register func as a read-only procedure of the current module.

The decorator read_proc is meant to be used to register module procedures. The registered func needs to be a callable which optionally takes ProcCtx as its first argument. Other arguments of func will be bound to values passed in the cypherQuery. The full signature of func needs to be annotated with types. The return type must be Record(field_name=type, ...) and the procedure must produce either a complete Record or None. To mark a field as deprecated, use Record(field_name=Deprecated(type), ...). Multiple records can be produced by returning an iterable of them. Registering generator functions is currently not supported.

write_proc()

def write_proc(func: typing.Callable[..., Record])

Register func as a writeable procedure of the current module.

The decorator write_proc is meant to be used to register module procedures. The registered func needs to be a callable which optionally takes ProcCtx as the first argument. Other arguments of func will be bound to values passed in the cypherQuery. The full signature of func needs to be annotated with types. The return type must be Record(field_name=type, ...) and the procedure must produce either a complete Record or None. To mark a field as deprecated, use Record(field_name=Deprecated(type), ...). Multiple records can be produced by returning an iterable of them. Registering generator functions is currently not supported.

InvalidMessageError Objects

class InvalidMessageError(Exception)

Signals using a message instance outside of the registered transformation.

Message Objects

class Message()

Represents a message from a stream.

is_valid()

def is_valid() -> bool

Return True if self is in valid context and may be used.

source_type()

def source_type() -> str

Supported in all stream sources

Raise InvalidArgumentError if the message is from an unsupported stream source.

payload()

def payload() -> bytes

Supported stream sources:

• Kafka
• Pulsar

Raise InvalidArgumentError if the message is from an unsupported stream source.

topic_name()

def topic_name() -> str

Supported stream sources:

• Kafka
• Pulsar

Raise InvalidArgumentError if the message is from an unsupported stream source.

key()

def key() -> bytes

Supported stream sources:

• Kafka

Raise InvalidArgumentError if the message is from an unsupported stream source.

timestamp()

def timestamp() -> int

Supported stream sources:

• Kafka

Raise InvalidArgumentError if the message is from an unsupported stream source.

offset()

def offset() -> int

Supported stream sources:

• Kafka

Raise InvalidArgumentError if the message is from an unsupported stream source.

InvalidMessagesError Objects

class InvalidMessagesError(Exception)

Signals using a messages instance outside of the registered transformation.

Messages Objects

class Messages()

Represents a list of messages from a stream.

is_valid()

def is_valid() -> bool

Return True if self is in valid context and may be used.

message_at()

def message_at(id: int) -> Message

Raise InvalidMessagesError if context is invalid.

total_messages()

def total_messages() -> int

Raise InvalidContextError if context is invalid.

TransCtx Objects

class TransCtx()

Context of a transformation being executed.

Access to a TransCtx is only valid during a single execution of a transformation. You should not globally store a TransCtx instance.

graph

@property
def graph() -> Graph

Raise InvalidContextError if context is invalid.

FuncCtx Objects

class FuncCtx()

Context of a function being executed.

Access to a FuncCtx is only valid during a single execution of a function in a query. You should not globally store a FuncCtx instance. The graph object within the FuncCtx is not mutable.

function()

def function(func: typing.Callable)

Register func as a user-defined function in the current module.

The decorator function is meant to be used to register module functions. The registered func needs to be a callable which optionally takes FuncCtx as its first argument. Other arguments of func will be bound to values passed in the Cypher query. Only the function arguments need to be annotated with types. The return type doesn't need to be specified, but it has to be supported by mgp.Any. Registering generator functions is currently not supported.

InvalidContextError Objects

class InvalidContextError(Exception)

Signals using a graph element instance outside of the registered procedure.

UnknownError Objects

class UnknownError(_mgp.UnknownError)

Signals unspecified failure.

UnableToAllocateError Objects

class UnableToAllocateError(_mgp.UnableToAllocateError)

Signals failed memory allocation.

InsufficientBufferError Objects

class InsufficientBufferError(_mgp.InsufficientBufferError)

Signals that some buffer is not big enough.

OutOfRangeError Objects

class OutOfRangeError(_mgp.OutOfRangeError)

Signals that an index-like parameter has a value that is outside its possible values.

LogicErrorError Objects

class LogicErrorError(_mgp.LogicErrorError)

Signals faulty logic within the program such as violating logical preconditions or class invariants and may be preventable.

DeletedObjectError Objects

class DeletedObjectError(_mgp.DeletedObjectError)

Signals accessing an already deleted object.

InvalidArgumentError Objects

class InvalidArgumentError(_mgp.InvalidArgumentError)

Signals that some of the arguments have invalid values.

KeyAlreadyExistsError Objects

class KeyAlreadyExistsError(_mgp.KeyAlreadyExistsError)

Signals that a key already exists in a container-like object.

ImmutableObjectError Objects

class ImmutableObjectError(_mgp.ImmutableObjectError)

Signals modification of an immutable object.

ValueConversionError Objects

class ValueConversionError(_mgp.ValueConversionError)

Signals that the conversion failed between python and cypher values.

SerializationError Objects

class SerializationError(_mgp.SerializationError)

Signals serialization error caused by concurrent modifications from different transactions.