Newsgroups: ut.cdf.csc324h
From: shiva@cs.toronto.edu (Shiva Nejati)
Subject: Marking, Assignment 1
Date: 3 Oct 2002 19:11:22 GMT


Hi all,

Here is a brief explanation about the marking Scheme of
the first assignment and the most common errors:

Question 1:

Part a: (2)

Part b: (2)

Part c: (3)

You should point out that it is impossible because regular expressions
cannot count (I didn't deduct any mark for some other similar answers).

Part d: (3)

A portion of marks was given to the answers with some minor mistakes.

For example:

+ instead of * (-1)

Question 2:

Part a: (2)
One mark was deducted for the answers that <boolean-expr> was not
surrounded by ().

Part b: (4)

Your answer should produce the following string:

username@domainPart1.domainPart2.dominaPart3. ... . domainPartN

username and domain Parts are strings of lower case letters.
(It was explained on the newsgroup)

- Two marks were deducted, for the CFG's which only produce two
domainParts.
- One mark was deducted, for the CFG's which produce a dot separated
string
  for username.
- One or Two marks were deducted, for the CFG's which produce a one
character username or
  domainPart instead of a string.

Part c: (4)

A portion of mark was given to some good answers that were not quite
correct.

Question 3:

Part a: (6)

1 - One mark was deducted for the answers that do not produce a list of
statements.
    Most of the answers had this problem:

  <case> -> case <int-expr> <stmt> ... (wrong, produces only one
statement)

  <case> -> case <int-expr> <stmt-list> .. (right).
  <stmt-list> -> <stmt><stmt-list>|empty

2 - Two marks were deducted for the answers that can produce more than one
default statement    not necessarily at the end of the switch statement.

  <switch-stmt> -> switch ( <int-expr> ) { <case-list> } (wrong)
  <case-list> -> case ...
                |default ...

Part b: (4)

Part c: (4)

The errors that were mentioned for Part a, were repeated in this part.
For error 1 and error 2, each of them, one mark was deducted.

Question 4:

Part a: (2)

Part b: (2)

You should point out the main property  of this Language. One of the
acceptable answers
is:
 For all the prefixes of these strings, # a's are more than or equal to #
b's.

Or you should at least describe this language by mathematical notations.

- One mark was given to the answers which say # a's is equal to # b's.

Part c: (2)

Part b: (2)

Your description should be sound and complete. One should be able to
rewrite the
grammars by reading your descriptions.

All the answers that were not complete but described some features of the
languages were given one mark (one mark for each part).


Question 5:

Part a: (2)
Part b: (2)
Part c: (2)
Part d: (2)


Let me know if you have any problem.

Thanks,
Shiva.


Newsgroups: ut.cdf.csc324h
From: shiva@cs.toronto.edu (Shiva Nejati)
Subject: Marking, Assignment 1
Date: 3 Oct 2002 23:33:24 GMT


Hi all,

Just to clarify more, for question 2, Part a, one mark was deducted for
answers that <boolean-exp> was not surrounded by (), and
the answers that used {} around the <stmt>.

Shiva.

P.S. In the newsgroup it was mentioned that <stmt> can be one statement,
or a group of statements surrounded by {}. Therefore, {<stmt>} is not
correct.