The visualizer can be used stand-alone or embedded in a Rails application.
The visualizer has an external dependency on coffee-script. It can be installed by:
# OS X $ brew install coffee-script # Linux $ apt-get install coffee-script
To use it standalone you must first get the code from the repository.
$ git clone email@example.com:AmanziTel/neo-viz.git $ cd neo-viz # To use it as is $ rackup [2011-06-23 07:48:34] INFO WEBrick 1.3.1 [2011-06-23 07:48:34] INFO ruby 1.8.7 (2011-05-23) [java] [2011-06-23 07:48:34] INFO WEBrick::HTTPServer#start: pid=64743 port=9292 # To install it as a gem $ rake install neo-viz 1.0.0 built to pkg/neo-viz-1.0.0.gem neo-viz (1.0.0) installed $ neo-viz [2011-06-23 07:51:10] INFO WEBrick 1.3.1 [2011-06-23 07:51:10] INFO ruby 1.8.7 (2011-05-23) [java] [2011-06-23 07:51:10] INFO WEBrick::HTTPServer#start: pid=64864 port=1666
Embedding in Rails
To embed the visualizer in Rails, you first have to add it to your
# Gemfile gem 'neo-viz', :git => 'firstname.lastname@example.org:AmanziTel/neo-viz.git'
Then you have to mount the route in
# config/routes.rb mount Neo::Viz::App => '/neo-viz'
You can now access it at e.g. http://localhost:3000/neo-viz.
When precompiling, you may need to specify these files manually
# production.rb config.assets.precompile += %w( jquery-1.6.1.min.js arbor.js neo-viz.js neo-viz.css )
The standalone version expects a
db directory containing the usual
Neo4j database files.
The embedded version uses the Neo4j database that has been configured for the embedding project.
There is a problem running the Jasmine specs (/run-specs) on Windows if the "therubyrhino" gem is installed and you are using JRuby. On Windows we want to use the built-in JScript runtime but ExecJS chooses Rhino if it is installed.
Typical errors returned over HTTP when requesting a .coffee file:
A workaround is to uninstall the therubyrhino gem, but ofcourse it will be re-installed on every call to bundle. I can't figure out how to tell bundler not to load a specific gem when I'm on JRuby and Windows.
- Make it possible to choose where that data comes from
- Implement a command line interface for starting the server and selecting the data source.
- Group nodes, if there are too many to show.
- bind and trigger with custom events
MIT, see LICENSE file
- 2011-12-20 Release 1.0.2 and fix of warnings (public folder)
- 2011-09-28 Added relationship filter
- 2001-09-28 Enabled embedding from AJAX-call (/embedded)
- 2011-06-23 Changed the namespace.
- 2011-06-23 Added depth to the traversal algorithm
- 2011-06-20 A query protocol for selecting only the relevant nodes
- 2011-06-20 Query protocol editor to select the relevant nodes to ask
- 2011-06-16 Views for showing only the relevant properties.
- 2011-06-16 Node filter
- 2011-06-16 Select number of nodes to show.
- 2011-06-14 Click to show detailed information.
- Connect with the Server to get the data.
- Fetch new data from the Server when a node is selected.
- Limit the amount of nodes shown on the screen.
- Show the properties of the nodes and relations.