PyMogwai

PyMogwai is a Python-based implementation of the Gremlin graph traversal language, designed to create and handle knowledge graphs entirely in Python without the need for an external Gremlin server.

Features

• Supports knowledge graph creation and manipulation
• Supports the import of arbitrary knowledge graphs with GraphML
• Implements a variety of traversal steps
• Enables one to traverse a graph with these steps
• Ability to integrate data from various sources like Excel, PDF and PowerPoint
• Simple and Pythonic API for graph operations

Demo

nicegui based demo

Getting started

Creating a Knowledge Graph

To create a graph using PyMogwai

from mogwai.core import MogwaiGraph
graph = MogwaiGraph()

# Add nodes and edges
n1 = graph.add_labeled_node("person", name="Alice", properties={"Age": 30})
n2 = graph.add_labeled_node("person", name="Bob", properties={"Age": 28})
graph.add_labeled_edge(n1, n2, "knows")

Import graphs

from mogwai.parser.graphml_converter import graphml_to_mogwaigraph

graph = graphml_to_mogwaigraph(path, node_label_key="node_label", node_name_key="node_name", edge_label_key="edge_label")

Performing Traversals

To perform any traversal of the graph, create a TraversalSource. Traversals start with a start step (V() or E()), which is followed by a sequence of steps. Note that a traversal can be executed by calling .run() on it.

from mogwai.core.traversal import MogwaiGraphTraversalSource

g = MogwaiGraphTraversalSource(graph)

# Example traversal that returns every person in the graph as a list
res = g.V().has_label("person").to_list().run()
print(res)

In order to use anonymous traversal in complex queries, import the statics module:

from mogwai.core.traversal import MogwaiGraphTraversalSource
from mogwai.core.steps.statics import *

g = MogwaiGraphTraversalSource(graph)

# Example traversal that returns every person in the graph as a list
query = g.V().has_label("person").filter_(properties('age').is_(gte(30))).to_list().by('name')
res = query.run()
print(res)

History

This project started as part of the RWTH Aachen i5 Knowledge Graph Lab SS2024 The original source is hosted at https://git.rwth-aachen.de/i5/teaching/kglab/ss2024/pymogwai 2024-08-15 the repository moved to github for better pypi integration