JanusGraph-Python extends Apache TinkerPop™'s Gremlin-Python with support for JanusGraph-specific types.

To connect to JanusGraph Server, a DriverRemoteConnection instance needs to be created and configured with a message serializer that adds support for JanusGraph specific types.

This can be done like this for GraphSON 3:

from gremlin_python.driver.driver_remote_connection import DriverRemoteConnection
from janusgraph_python.driver.serializer import JanusGraphSONSerializersV3d0

connection = DriverRemoteConnection(
  'ws://localhost:8182/gremlin', 'g',
  message_serializer=JanusGraphSONSerializersV3d0())

This can be done like this for GraphBinary:

from gremlin_python.driver.driver_remote_connection import DriverRemoteConnection
from janusgraph_python.driver.serializer import JanusGraphBinarySerializersV1

connection = DriverRemoteConnection(
 'ws://localhost:8182/gremlin', 'g',
 message_serializer=JanusGraphBinarySerializersV1())

Note that the client should be disposed on shut down to release resources and to close open connections with connection.close(). The connection can then be used to configure a GraphTraversalSource:

from gremlin_python.process.anonymous_traversal import traversal

g = traversal().with_remote(connection)
# Reuse 'g' across the application

The GraphTraversalSource g can now be used to spawn Gremlin traversals:

hercules_age = g.V().has("demigod", "name", "hercules").values("age").next()
print(f"Hercules is {hercules_age} years old.")

Refer to the chapter Gremlin Query Language in the JanusGraph docs for an introduction to Gremlin and pointers to further resources. The main syntactical difference for Gremlin-Python is that it follows Python naming conventions, e.g., method names use snake_case instead of camelCase. Other difference is that when Python reserved words (e.g. "is") overlap with Gremlin steps or tokens, those gets underscore suffix (e.g. "is_").

The Text class provides methods for full-text and string searches:

from janusgraph_python.process.traversal import Text

g.V().has("demigod", "name", Text.text_prefix("herc")).to_list()

The other text predicates can be used the same way.

The lowest supported JanusGraph version is 1.0.0. The following table shows the supported JanusGraph versions for each version of JanusGraph-Python:

While it should also be possible to use JanusGraph-Python with other versions of JanusGraph than mentioned here, compatibility is not tested and some functionality (like added Gremlin steps) will not work as it is not supported yet in case of an older JanusGraph version or was removed in a newer JanusGraph version.

JanusGraph-Python supports GraphSON 3 as well as GraphBinary.

Not all of the JanusGraph-specific types are already supported by the formats:

JanusGraph-Python uses the same communication channels as JanusGraph in general. So, please refer to the Community section in JanusGraph's main repository for more information about these various channels.

Please use GitHub issues only to report bugs or request features.

Please see CONTRIBUTING.md in JanusGraph's main repository for more information, including CLAs and best practices for working with GitHub.

JanusGraph-Python code is provided under the Apache 2.0 license and documentation is provided under the CC-BY-4.0 license. For details about this dual-license structure, please see LICENSE.txt.