EventVisualizer¶

Module: eventspype.viz.visualizer

EventVisualizer¶

class EventVisualizer:
    def __init__(self) -> None

Generates graphviz directed graphs showing the architecture of an EventsPype system. Works with MultiPublisher and MultiSubscriber subclasses by reading their class-level EventPublication and EventSubscription attributes.

Note

The graphviz Python package is a core dependency. The Graphviz system binaries must also be installed separately (brew install graphviz or apt-get install graphviz).

Methods¶

`add_publisher`¶

def add_publisher(self, publisher_class: type[MultiPublisher]) -> None

Register a MultiPublisher subclass for inclusion in the diagram. Reads get_event_definitions() to discover publications.

Raises: ValueError if publisher_class is not a MultiPublisher subclass.

`add_subscriber`¶

def add_subscriber(self, subscriber_class: type[MultiSubscriber]) -> None

Register a MultiSubscriber subclass for inclusion in the diagram. Reads get_event_definitions() to discover subscriptions.

Raises: ValueError if subscriber_class is not a MultiSubscriber subclass.

`generate_graph`¶

def generate_graph(
    self,
    graph_name: str = "EventSystem",
    graph_format: str = "png",
) -> graphviz.Digraph

Build and return a graphviz.Digraph object. Does not write any files.

Parameters:

Parameter	Type	Default	Description
`graph_name`	`str`	`"EventSystem"`	Name embedded in the graph
`graph_format`	`str`	`"png"`	Output format for subsequent renders

Returns: graphviz.Digraph

`render`¶

def render(
    self,
    output_path: str,
    graph_name: str = "EventSystem",
    graph_format: str = "png",
    view: bool = False,
) -> str

Render the diagram to a file and return the path.

Parameters:

Parameter	Type	Default	Description
`output_path`	`str`	—	File path without extension
`graph_name`	`str`	`"EventSystem"`	Name embedded in the graph
`graph_format`	`str`	`"png"`	Output format: `png`, `pdf`, `svg`, `dot`, etc.
`view`	`bool`	`False`	Open the rendered file with the system default viewer

Returns: str — path to the rendered file.

`clear`¶

def clear(self) -> None

Remove all registered publishers and subscribers.

Diagram layout¶

The graph uses a left-to-right (rankdir=LR) layout with orthogonal splines:

Publisher nodes (blue, #E3F2FD / #1976D2) — show class name and all EventPublication attributes with their original tags.
Subscriber nodes (purple, #F3E5F5 / #7B1FA2) — show class name and all EventSubscription attributes with tags and source publisher class name.
Edges (green, #4CAF50) — drawn from each publisher to each subscriber where a subscription tag matches a publication tag.

Example¶

from eventspype import EventVisualizer, MultiPublisher, MultiSubscriber
from eventspype import EventPublication, EventSubscription
from dataclasses import dataclass
import logging

@dataclass
class OrderPlacedEvent:
    order_id: int

@dataclass
class OrderCancelledEvent:
    order_id: int

class OrderService(MultiPublisher):
    ORDER_PLACED = EventPublication("order_placed", OrderPlacedEvent)
    ORDER_CANCELLED = EventPublication("order_cancelled", OrderCancelledEvent)

class OrderHandler(MultiSubscriber):
    on_placed = EventSubscription(
        publisher_class=OrderService,
        event_tag="order_placed",
        callback=lambda self, e, t, c: None,
    )
    on_cancelled = EventSubscription(
        publisher_class=OrderService,
        event_tag="order_cancelled",
        callback=lambda self, e, t, c: None,
    )

    def logger(self) -> logging.Logger:
        return logging.getLogger(__name__)

viz = EventVisualizer()
viz.add_publisher(OrderService)
viz.add_subscriber(OrderHandler)

# Render to order_system.png
path = viz.render("order_system", graph_format="png")
print(f"Diagram: {path}")

# Or get the Digraph object for customization
graph = viz.generate_graph("OrderSystem", "svg")
graph.attr(bgcolor="white")
graph.render("custom_output")

EventVisualizer¶