Neography is a thin Ruby wrapper to the Neo4j Rest API, for more information:
If you want to the full power of Neo4j, you will want to use JRuby and the excellent Neo4j.rb gem at github.com/andreasronge/neo4j by Andreas Ronge
Complement to Neography are the:
-
Neo4j Active Record Adapter by Nikhil Lanjewar
-
Neology by Carlo Alberto Degli Atti
-
Neoid by Elad Ossadon
An alternative is the Architect4r Gem at github.com/namxam/architect4r by Maximilian Schulz
-
Neovigator fork it at github.com/maxdemarzi/neovigator
-
Neoflix fork it at github.com/maxdemarzi/neoflix
gem install 'neography'
After that, in your ruby script:
require 'rubygems' require 'neography'
in order to access the functionality.
for use: os rake json httparty for development: rspec net-http-spy
Just add gem ‘neography’ to your Gemfile and run bundle install.
The following tasks will be available to you:
rake neo4j:install # Install Neo4j to the neo4j directory under your project rake neo4j:install[community,1.6.M03] # Install Neo4j Community edition, version 1.6.M03 rake neo4j:install[advanced,1.5] # Install Neo4j Advanced edition, version 1.5 rake neo4j:install[enterprise,1.5] # Install Neo4j Enterprise edition, version 1.5 rake neo4j:start # Start Neo4j rake neo4j:stop # Stop Neo4j rake neo4j:restart # Restart Neo4j rake neo4j:reset_yes_i_am_sure # Wipe your Neo4j Database
Windows users will need to run in a command prompt with Administrative Privileges in order to install Neo4j as a Service.
If you are not using Rails, then add:
require 'neography/tasks'
to your Rakefile to have access to these tasks.
@neo = Neography::Rest.new({:protocol => 'http://', :server => 'localhost', :port => 7474, :directory => '', # use '/<my directory>' or leave out for default :authentication => 'basic', # 'basic', 'digest' or leave out for default :username => 'your username', #leave out for default :password => 'your password', #leave out for default :log_file => 'neography.log', :log_enabled => false, :max_threads => 20, :cypher_path => '/cypher', :gremlin_path => '/ext/GremlinPlugin/graphdb/execute_script'})
Quick initializer (assumes basic authorization if username is given):
@neo = Neography::Rest.new("http://username:[email protected]:7474/mydirectory")
To Use:
@neo = Neography::Rest.new # Inialize using all default parameters
@neo.get_root # Get the root node
@neo.create_node # Create an empty node
@neo.create_node("age" => 31, "name" => "Max") # Create a node with some properties
@neo.create_unique_node(index_name, key, unique_value, # Create a unique node
{"age" => 31, "name" => "Max"}) # this needs an existing index
@neo.get_node(node2) # Get a node and its properties
@neo.delete_node(node2) # Delete an unrelated node
@neo.delete_node!(node2) # Delete a node and all its relationships
@neo.reset_node_properties(node1, {"age" => 31}) # Reset a node's properties
@neo.set_node_properties(node1, {"weight" => 200}) # Set a node's properties
@neo.get_node_properties(node1) # Get just the node properties
@neo.get_node_properties(node1, ["weight","age"]) # Get some of the node properties
@neo.remove_node_properties(node1) # Remove all properties of a node
@neo.remove_node_properties(node1, "weight") # Remove one property of a node
@neo.remove_node_properties(node1, ["weight","age"]) # Remove multiple properties of a node
@neo.create_relationship("friends", node1, node2) # Create a relationship between node1 and node2
@neo.create_unique_relationship(index_name, key, value, # Create a unique relationship between nodes
"friends", new_node1, new_node2) # this needs an existing index
@neo.get_relationship(rel1) # Get a relationship
@neo.get_node_relationships(node1) # Get all relationships
@neo.get_node_relationships(node1, "in") # Get only incoming relationships
@neo.get_node_relationships(node1, "all", "enemies") # Get all relationships of type enemies
@neo.get_node_relationships(node1, "in", "enemies") # Get only incoming relationships of type enemies
@neo.delete_relationship(rel1) # Delete a relationship
@neo.reset_relationship_properties(rel1, {"age" => 31}) # Reset a relationship's properties
@neo.set_relationship_properties(rel1, {"weight" => 200}) # Set a relationship's properties
@neo.get_relationship_properties(rel1) # Get just the relationship properties
@neo.get_relationship_properties(rel1, ["since","met"]) # Get some of the relationship properties
@neo.remove_relationship_properties(rel1) # Remove all properties of a relationship
@neo.remove_relationship_properties(rel1, "since") # Remove one property of a relationship
@neo.remove_relationship_properties(rel1, ["since","met"]) # Remove multiple properties of a relationship
@neo.list_node_indexes # gives names and query templates for all defined indices
@neo.create_node_index(name, type, provider) # creates an index, defaults are "exact" and "lucene"
@neo.add_node_to_index(index, key, value, node1) # adds a node to the index with the given key/value pair
@neo.remove_node_from_index(index, key, value, node1) # removes a node from the index with the given key/value pair
@neo.remove_node_from_index(index, key, node1) # removes a node from the index with the given key
@neo.remove_node_from_index(index, node1) # removes a node from the index
@neo.get_node_index(index, key, value) # exact query of the node index with the given key/value pair
@neo.find_node_index(index, key, value) # advanced query of the node index with the given key/value pair
@neo.find_node_index(index, query ) # advanced query of the node index with the given query
@neo.list_relationship_indexes # gives names and query templates for relationship indices
@neo.create_relationship_index(name, "fulltext", provider) # creates a relationship index with "fulltext" option
@neo.add_relationship_to_index(index, key, value, rel1) # adds a relationship to the index with the given key/value pair
@neo.remove_relationship_from_index(index, key, value, rel1) # removes a relationship from the index with the given key/value pair
@neo.remove_relationship_from_index(index, key, rel1) # removes a relationship from the index with the given key
@neo.remove_relationship_from_index(index, rel1) # removes a relationship from the index
@neo.get_relationship_index(index, key, value) # exact query of the relationship index with the given key/value pair
@neo.find_relationship_index(index, key, value) # advanced query of the relationship index with the given key/value pair
@neo.find_relationship_index(index, query) # advanced query of the relationship index with the given query
@neo.execute_script("g.v(0)") # sends a Groovy script (through the Gremlin plugin)
@neo.execute_script("g.v(id)", {:id => 3}) # sends a parameterized Groovy script (optimized for repeated calls)
@neo.execute_query("start n=node(0) return n") # sends a Cypher query (through the Cypher plugin)
@neo.execute_query("start n=node(id) return n", {:id => 3}) # sends a parameterized Cypher query (optimized for repeated calls)
@neo.get_path(node1, node2, relationships, depth=4, algorithm="shortestPath") # finds the shortest path between two nodes
@neo.get_paths(node1, node2, relationships, depth=3, algorithm="allPaths") # finds all paths between two nodes
@neo.get_shortest_weighted_path(node1, node2, relationships, # find the shortest path between two nodes
weight_attr='weight', depth=2, # accounting for weight in the relationships
algorithm='dijkstra') # using 'weight' as the attribute
nodes = @neo.traverse(node1, # the node where the traversal starts
"nodes", # return_type "nodes", "relationships" or "paths"
{"order" => "breadth first", # "breadth first" or "depth first" traversal order
"uniqueness" => "node global", # See Uniqueness in API documentation for options.
"relationships" => [{"type"=> "roommates", # A hash containg a description of the traversal
"direction" => "all"}, # two relationships.
{"type"=> "friends", #
"direction" => "out"}], #
"prune evaluator" => {"language" => "javascript", # A prune evaluator (when to stop traversing)
"body" => "position.endNode().getProperty('age') < 21;"},
"return filter" => {"language" => "builtin", # "all" or "all but start node"
"name" => "all"},
"depth" => 4})
# "depth" is a short-hand way of specifying a prune evaluator which prunes after a certain depth.
# If not specified a depth of 1 is used and if a "prune evaluator" is specified instead of a depth, no depth limit is set.
@neo.batch [:get_node, node1], [:get_node, node2] # Gets two nodes in a batch
@neo.batch [:create_node, {"name" => "Max"}],
[:create_node, {"name" => "Marc"}] # Creates two nodes in a batch
@neo.batch [:set_node_property, node1, {"name" => "Tom"}],
[:set_node_property, node2, {"name" => "Jerry"}] # Sets the property of two nodes
@neo.batch [:create_unique_node, index_name, key, value,
{"age" => 33, "name" => "Max"}] # Creates a unique node
@neo.batch [:get_node_relationships, node1, "out",
[:get_node_relationships, node2, "out"] # Get node relationships in a batch
@neo.batch [:get_relationship, rel1],
[:get_relationship, rel2] # Gets two relationships in a batch
@neo.batch [:create_relationship, "friends",
node1, node2, {:since => "high school"}],
[:create_relationship, "friends",
node1, node3, {:since => "college"}] # Creates two relationships in a batch
@neo.batch [:create_unique_relationship, index_name,
key, value, "friends", node1, node2] # Creates a unique relationship
@neo.batch [:get_node_index, index_name, key, value] # Get node index
@neo.batch [:get_relationship_index, index_name, key, value] # Get relationship index
@neo.batch [:create_node, {"name" => "Max"}],
[:create_node, {"name" => "Marc"}], # Creates two nodes and index them
[:add_node_to_index, "test_node_index", key, value, "{0}"],
[:add_node_to_index, "test_node_index", key, value, "{1}"],
[:create_relationship, "friends", # and create a relationship for those
"{0}", "{1}", {:since => "college"}], # newly created nodes
[:add_relationship_to_index,
"test_relationship_index", key, value, "{4}"] # and index the new relationship
@neo.batch *[[:create_node, {"name" => "Max"}],
[:create_node, {"name" => "Marc"}]] # Use the Splat (*) with Arrays of Arrays
See docs.neo4j.org/chunked/milestone/rest-api-batch-ops.html for Neo4j Batch operations documentation.
Please see the specs for more examples.
Experimental:
nodes = @neo.create_nodes(5) # Create 5 empty nodes
nodes = @neo.create_nodes_threaded(5) # Create 5 empty nodes using threads
nodes = @neo.create_node_nodes([{"age" => 31, "name" => "Max"},
{"age" => 24, "name" => "Alex"}) # Create two nodes with properties
nodes = @neo.create_node_nodes_threaded([{"age" => 31, "name" => "Max"},
{"age" => 24, "name" => "Alex"}) # Create two nodes with properties threaded
nodes = @neo.get_nodes([17,86,397,33]) # Get four nodes by their id
one_set_nodes = @neo.create_nodes(3)
another_node = @neo.create_node("age" => 31, "name" => "Max")
nodes = @neo.get_nodes([one_set_nodes, another_node]) # Get four nodes
Trying to mimic the Neo4j.rb API.
Now we are returning full objects. The properties of the node or relationship can be accessed directly (node.name). The Neo4j ID is available by using node.neo_id .
@neo2 = Neography::Rest.new ({:server => '192.168.10.1'}) Neography::Node.create # Create an empty node Neography::Node.create("age" => 31, "name" => "Max") # Create a node with some properties Neography::Node.create(@neo2, {"age" => 31, "name" => "Max"}) # Create a node on the server defined in @neo2 Neography::Node.create({"age" => 31, "name" => "Max"}, @neo2) # Same as above, but different order Neography::Node.load(5) # Get a node and its properties by id Neography::Node.load(existing_node) # Get a node and its properties by Node Neography::Node.load("http://localhost:7474/db/data/node/2") # Get a node and its properties by String Neography::Node.load(@neo2, 5) # Get a node on the server defined in @neo2 Neography::Node.load(5, @neo2) # Same as above, but different order n1 = Node.create n1.del # Deletes the node n1.exist? # returns true/false if node exists in Neo4j n1 = Node.create("age" => 31, "name" => "Max") n1[:age] #returns 31 # Get a node property using [:key] n1.name #returns "Max" # Get a node property as a method n1[:age] = 24 # Set a node property using [:key] = n1.name = "Alex" # Set a node property as a method n1[:hair] = "black" # Add a node property using [:key] = n1.weight = 190 # Add a node property as a method n1[:name] = nil # Delete a node property using [:key] = nil n1.name = nil # Delete a node property by setting it to nil n2 = Neography::Node.create new_rel = Neography::Relationship.create(:family, n1, n2) # Create a relationship from my_node to node2 new_rel.start_node # Get the start/from node of a relationship new_rel.end_node # Get the end/to node of a relationship new_rel.other_node(n2) # Get the other node of a relationship new_rel.attributes # Get the attributes of the relationship as an array existing_rel = Neography::Relationship.load(12) # Get an existing relationship by id existing_rel.del # Delete a relationship Neography::Relationship.create(:friends, n1, n2) n1.outgoing(:friends) << n2 # Create outgoing relationship n1.incoming(:friends) << n2 # Create incoming relationship n1.both(:friends) << n2 # Create both relationships n1.outgoing # Get nodes related by outgoing relationships n1.incoming # Get nodes related by incoming relationships n1.both # Get nodes related by any relationships n1.outgoing(:friends) # Get nodes related by outgoing friends relationship n1.incoming(:friends) # Get nodes related by incoming friends relationship n1.both(:friends) # Get nodes related by friends relationship n1.outgoing(:friends).incoming(:enemies) # Get nodes related by one of multiple relationships n1.outgoing(:friends).depth(2) # Get nodes related by friends and friends of friends n1.outgoing(:friends).depth(:all) # Get nodes related by friends until the end of the graph n1.outgoing(:friends).depth(2).include_start_node # Get n1 and nodes related by friends and friends of friends n1.outgoing(:friends).prune("position.endNode().getProperty('name') == 'Tom';") n1.outgoing(:friends).filter("position.length() == 2;") n1.rel?(:friends) # Has a friends relationship n1.rel?(:outgoing, :friends) # Has outgoing friends relationship n1.rel?(:friends, :outgoing) # same, just the other way n1.rel?(:outgoing) # Has any outgoing relationships n1.rel?(:both) # Has any relationships n1.rel?(:all) # same as above n1.rel? # same as above n1.rels # Get node relationships n1.rels(:friends) # Get friends relationships n1.rels(:friends).outgoing # Get outgoing friends relationships n1.rels(:friends).incoming # Get incoming friends relationships n1.rels(:friends,:work) # Get friends and work relationships n1.rels(:friends,:work).outgoing # Get outgoing friends and work relationships n1.all_paths_to(n2).incoming(:friends).depth(4) # Gets all paths of a specified type n1.all_simple_paths_to(n2).incoming(:friends).depth(4) # for the relationships defined n1.all_shortest_paths_to(n2).incoming(:friends).depth(4) # at a maximum depth n1.path_to(n2).incoming(:friends).depth(4) # Same as above, but just one path. n1.simple_path_to(n2).incoming(:friends).depth(4) n1.shortest_path_to(n2).incoming(:friends).depth(4) n1.shortest_path_to(n2).incoming(:friends).depth(4).rels # Gets just relationships in path n1.shortest_path_to(n2).incoming(:friends).depth(4).nodes # Gets just nodes in path
See Neo4j API for:
A couple of examples borrowed from Matthew Deiters’s Neo4jr-social:
Phase 2 way of doing these:
To run testing locally you will need to have two instances of the server running. There is some good advice on how to set up the a second instance on the neo4j site. Connect to the second instance in your testing environment, for example:
if Rails.env.development? @neo = Neography::Rest.new({:port => 7474}) elsif Rails.env.test? @neo = Neography::Rest.new({:port => 7475}) end
Install the test-delete-db-extension plugin, as mentioned in the neo4j.org docs, if you want to use the Rest clean_database method to empty your database between tests. In Rspec, for example, put this in your spec_helper.rb:
config.before(:each) do @neo.clean_database("yes_i_really_want_to_clean_the_database") end
-
Batch functions
-
Phase 2 Index functionality
-
More Tests
-
More Examples
-
Mixins ?
<img src=“https://secure.travis-ci.org/maxdemarzi/neography.png” />
Please create a new issue if you run into any bugs. Contribute patches via pull requests.
If you are just starting out, or need help send me an e-mail at [email protected]. Check you my blog at maxdemarzi.com where I have more Neography examples.
-
Neography - MIT, see the LICENSE file github.com/maxdemarzi/neography/tree/master/LICENSE.
-
Lucene - Apache, see lucene.apache.org/java/docs/features.html
-
Neo4j - Dual free software/commercial license, see neo4j.org