Alchemist

Given a Cogswell state machine definition, including test vectors that generate events that pass and fail the transition guards, Alchemist examines the shortest traversal that covers all the guards, and uses this to build a sequence of Touchstone tests.

The shortest traversal is found by the solution to the directed Chinese postman problem (Mei-Ko Kwan, Graphic Programming Using Odd or Even Points, Chinese Math., 1:273-277, 1962), as found by Harold Thimbleby’s Java code at www.uclic.ucl.ac.uk/harold/cpp.

Models

State machines defined using Cogswell can be partitioned into multiple scenarios. These state machines have test vectors that generate events which pass or fail guards. Alchemist generates test definitions from the state machines. The test definitions are run by Touchstone. The tests verify the option of a System which is assembled by Trimurti from components.

Alchemist holds one state machine that it uses to define tests. It also holds a Touchstone::System, where the tests are put. The Alchemist delegates determination of the shortest complete test sequence to a DirectedPostman.

Status

Alchemist and documentation are currently previews, which have not yet undergone internal reviews. Alchemist is actively being used.

Installation

Alchemist can be downloaded in several forms:

A RubyGem, containing only Alchemist. Due to lack of resources, the RubyGem is untested.
A "combo package" containing all code locally managed by Amber, but not required libraries. This is all you need if you want to install the code and required libraries yourself. The Amber build system is not included and not required. Due to a lack of resources, this has not been tested.
A development snapshot containing Alchemist and nearly everything it depends on, in a .tgz or .zip archive. The snapshot contains a complete Amber development environment. These snapshots are very large. Before using the development snapshot, you need to:
1. Setup Amber (included in the snapshot): see ReadMe_Amber.txt
2. Install libraries required by Touchstone:
```
    cd $DevRoot/Projects/Alchemist
    rake install_lib
```

Usage

You should start by defining a Cogswell state machine describing the system you want to test. This can be done in YAML or in code: please see the Cogswell documentation. The state machine Transitions need to include test vectors that pass or fail the guards. You can then build the tests, and execute them immediately, or save them in YAML form for later execution.

Example

  require 'nist/alchemist'

  machine = nil
  File.open('Cogswell_Telephone_Default.yml') {|f|
     machine = Cogswell::Machine.load(f)
     }
  builder = Alchemist.new(machine)
  # Build the tests
  builder.build('Telewerks', 'Complete')
  # Save a graph of the tests (view it with GraphViz):
  builder.save_dot('Telephone_Test.dot')
  # Save the tests for execution by Touchstone:
  builder.save_yaml('Telephone_Test.yml')

Room For Improvement

Integration with Trimurti could be improved.

Currently mock objects can be used, but there are no helpers to simplify this. If you are interested in mocks, you might want to look into:

Mocha: mocha.rubyforge.org/
Mockr: mockr.rubyforge.org/
Cmock: rubyforge.org/projects/cmock/
Mockery: rubyforge.org/projects/mockery/
SchMock: rubyforge.org/projects/schmock/

Author: A. Griesser

Official contribution of the National Institute of Standards and Technology; not subject to copyright in the United States. Certain commercial equipment, instruments, or materials are identified in this paper to foster understanding. Such identification does not imply recommendation or endorsement by the National Institute of Standards and Technology, nor does it imply that the materials or equipment identified are necessarily the best available for the purpose.

ReadMe.txt