lstar_dont_care

Note: the project was implemented and tested on Python 3.7.8

Installation and Usage

1. Clone the project :

git clone https://github.com/tomyaacov/lstar_dont_care.git

2. Create a virtual environment and activate it:

cd lstar_dont_care
python -m venv env 
source env/bin/activate

3. Update pip and install all dependencies:

pip install --upgrade pip
pip install -r requirements.txt

4. Run our initial algorithms experiment on the magento toy example:

• find a 3DFA using a standard l star (moore machine):
  • membership queries - for w: if w not in M return ?, else check if it is a prefix of P or in F. if not run on system and see result.
  • Equivalence queries - for a 3DFA:
    • Check that the failing tests in the list are indeed accepted by the 3DFA.
    • Check that the passing tests are rejected by the 3DFA.
    • Sample tests from L(3DFA)\cap M and check that they fail.
    • Sample words from M, and check that they produce the “right” results: pass → rejected, fail → accepted.
• After the 3DFA is built run this 2 options to get the final DFA:
  • run RPNI on a data set created from the 3DFA (observation table).
  • run the minimization algorithm from here to get the DFA.

python test_suite_based_lstar.py

5. Run algorithms experiment on the coffee\magento examples:

python rc_lstar.py