NeoMetaTracker 0.0.11

Track and visualize changes in a Neo4j database schema

neo4j-meta-logger

A python module to log and track your neo4j graph schema transformations.

call db.schema.visualization + ⏲️ a time machine + some helpful 🔧 tools and 📈 statistics

Works fast with neo4j graphs of any size 💪

Maintainer: Tim Bleimehl
Status: Pre-alpha

Example

Lets create a sample graph with python and Neo4J

import py2neo
from NeoMetaTracker import NeoMetaTracker

g = py2neo.Graph(name="test_graph")

g.run(
    "CREATE p = (:Human{name:'Amina Okujewa'})-[:LIVES_ON]->(:World {name: 'Earth'})"
)
g.run(
    "CREATE p = (:Cat{name:'Grumpy Cat'})-[:LIVES_ON]->(:World {name: 'Internet'})"
)
g.run(
    "MATCH (wI:World{name:'Internet'}),(wE:World{name:'Earth'}) CREATE (wI)-[:EXISTS_ON]->(wE)"
)

Our graph looks like this:

and the schema will look like this:

Lets capture the current status to analyse the changes later.

meta_logger = NeoMetaTracker(test_graph)
meta_logger.capture()

Now lets do some changes to our Graphs content

g.run(
    "MATCH (wI:World{name:'Internet'}),(wE:World{name:'Earth'}) CREATE (as:Human{name:'Aaron Swartz'})-[:LIVES_ON]->(wI), (as)-[:LIVES_ON]->(wE)"
)

Our graph now looks like this

The schema still looks the same. We wont be able to recognize changes without any help:

call db.schema.visualization

Here come NeoMetaTracker for the rescue

Lets do another capture to compare the changes we did:

meta_logger.capture()

Now we can analyze the changes in the graph:

meta_logger.get_numeric_last_changes()

This outputs:

{'labels': {'Human': 1}, 'relations': {'LIVES_ON': 2}}

We can see we created one new Node with the Label Human and 2 new relations named LIVES_ON.
This allready can be a valuable meta information, but wait there is more...

Lets visualize the schema changes in another graph.

changes_subgraph = meta_logger.get_schemagraph_last_changes()
schema_g = py2neo.Graph(name="test_graph")
schema_g.merge(changes_subgraph)

Now we only see the part of our schema that changed:

We can also recall any old state of our schema on any point in time where we did a capture

meta_logger.capture_points[0].schema

This will return a py2neo.Subgraph of the schema from the beginning of our script. Same as call db.schema.visualization but with a timemachine :)