****OWL Pizzeria
****Project 2 - CMSC 498w

****Due April 14, 2004

The overall performance of your system is easy -- you will write a program
(in Python) that will take a list of pizzas with the requested ingredients
(i.e., toppings) for each pizza.  At the end, you will print out a bill
itemizing which type of pizza each one is, and how much they each cost
(and a total).  You must charge the MINIMUM that makes sense.

The trick, of course, is in how your pizza ordering system works -- you
will define an OWL ontology that encodes the following:

Ingredients are: 
	Anchovy, Ham, MeatBall, Mushroom, Olive, Onion, 
	Pepperoni, Pepper, Sausage

The following are vegetables: 
	Mushroom, Olive, Onion, Pepper

The following are meats: 
	Ham, Meatball, Pepperoni, Sausage

Meats and vegetables are disjoint classes.

Note that Anchovy is neither a vegetable or a meat!

A small pizza is any pizza with 2 or less ingredients on it.
A medium pizza is any pizza with 3 to 5 ingredients on it.
A large pizza is any pizza with 6 or more ingredients.

A vegetarian pizza is any pizza with no meat on it.
A meatOnly pizza is any pizza with no vegetables on it.

Prices are as follow:
 Small vegetarian pizza - $2
 Any other small pizza - $3
 MeatOnly medium pizza - $4
 Any other Medium pizza - $5
 Any large pizza - $6
 Anything else that might be a pizza but not one of the above - $10

****Implementation details:

Reasoning will be done by the online version of the Pellet OWL-DL reasoner. 
We've provided a support script that wraps the basic calls for you:
	http://www.mindswap.org/2004/cmsc498w/pellet-submit.py

There is one function, checkString(), which you can import (or just incorporate)
into your program. It take one required argument and one optional 
argument, both strings which must be RDF/XML documents. (Feel free to
write some convenience wrappers, e.g., pass in a TripleStore.) If only
one argument is given, then that string will be checked for consistency.
If two arguments are given, then the second will be checked to see if it 
is an entailment of the first.

This, of course, requires a network connection. You could also download 
and install Pellet locally, but then you'd have to write a slightly 
different wrapper (e.g., some system() calls).

Note that currently, the results are in a fragment of HTML, e.g.:

	<H1>Results</H1>
	<b>Input file: </b>Text area<br>
	<b>Level: </b>Lite<br>
	<b>Consistent: </b>Yes<br>

You have to scrape this for the results you want (e.g., that the string 
you submitted is consistent) at the moment. I'll try to quickly upgrade 
the script to wrap this (i.e., it should return true or false), but feel 
free to do it yourself :)

****Submissions:

This follows the usual format, email to bparsia@isr.umd.edu with a subject 
line of "[CMCS498w] PA2".

There should be two files attached:
	yourinitials-pizza.owl  Your OWL ontology for pizzas
	yourinitials-pizzaOrders.py Your processer for Pizzas

The input to pizzaOrders.py will be a text file describing the order.

Each pizza description will appear on a distinct line. 

Each line a space delimited list of tokens.

The first token is the Pizza "label", e.g, "A", "B", "Table2"

The remaining tokens will be ingredients from the lists above (case 
insignificant).

So:
----

table1 Anchovy Mushroom Olive
table2 Pepperoni
table3 Pepper onion  Sausage

----

The result printout should look like:

----
table1: PizzaType Price
...

TotalPrice: $...
----