首页 >
> 详细

COMP2022: Assignment 2

Due: 23:59pm Sunday 20th October 2019 (end week 10)

1 The grammar G [10%]

Consider the following grammar G, which represents a fragment of a simple programming language:

S → SL | ε

L → A; | E; | C;

E → (EBE) | N | V

A → let V =E

C → while E do S | while E do S else S

B → + | - | * | >

V → x | y | z

N → ND | N0 | D

D → 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 (you can denote this [1-9])

In this assignment, you can ignore whitespace in the strings (i.e. spaces and newlines are not part of the

language, but should be used for the sake of readability).

i) List the variables of G

ii) List the terminals of G

iii) Give a leftmost derivation of the string let x=(y-20); while 1 do y;;

iv) Draw a parse tree for the string: let x=(y-20); while 1 do y;;

2 Prove that L(G) is not regular [5%]

Prove that the language generated by G is not a regular language, by contradicting the Pumping Lemma.

3 Prove G is not LL(1) [5%]

Prove that G is not an LL(1) grammar.

4 Find an equivalent LL(1) grammar G′

[10%]

Find an equivalent grammar G′ which is LL(1), by using the grammar transformation techniques shown

in lectures, or otherwise. Describe the process and show your working.

1

5 LL(1) parse table [15%]

Complete the LL(1) parse table for G′

. Describe the process and show your working, including:

1. FIRST sets for all the production rules of G′

2. FOLLOW sets for variables of G′ only if they are needed

6 Implementation [25%]

Implement a program which parses strings using an LL(1) table driven parser, using the table you

determined for G′

in the previous exercise. You may use Python, Java, C, C++, or Lisp. If you’d like to

use a different language then please check with us first.

• Input:

The first command line argument is the filename of a file containing the string of characters to test.

• Output:

1. Print a trace of the execution, showing the steps followed by the program as it performs the

left-most derivation. This should look similar to parsing the string through a PDA. An example

of this is given in the appendices.

2. After parsing the whole input file, print ACCEPTED or REJECTED, depending on whether or not

the string could be derived by the grammar.

3. If there is a symbol in the input string which is not a terminal from the grammar, the program

should output ERROR_INVALID_SYMBOL (This could be during or before trying to parse the

input.)

• All whitespace in the input file should be ignored (line breaks, spaces, etc.). The output will be easier

to read if you remove the whitespace before starting the parse.

• Examples of the program output syntax are provided in the appendices.

7 Extension [30%]

You may pick one of the following extensions to improve your parser. Any one of the following ideas

would be sufficient (do not implement more than one). They are listed roughly in order of increasing

difficulty/effort, but are worth the same marks:

a) Use the FIRST and FOLLOW sets to implement an error recovery feature. This should give the user

suggestions on possible corrections which could change strings which could not be derived in G′ For

this extension you need to:

• Implementation: If a second command line argument error is given, then instead of rejecting

a string that is not in the language, it should make suggestions about how to correct it. The

user chooses one of the options, then the program should continue the parse. Some examples

are provided in the appendices. This should be included in the code you submit to PASTA.

• Report: Explain how your error recovery feature uses the FIRST and FOLLOW sets to work,

and show some useful examples.

b) Extend your program to be able to evaluate the input program, if a second command line argument

eval is given. Some examples are provided in the appendices, and some supplementary material will

be provided on Ed to explain the algorithm needed to build and evaluate the parse tree. This should

be included in the code you submit to PASTA.

• Implementation:

2

– Build a parse tree as it performs the leftmost derivation.

– Evaluate that parse tree.

• Report: Give some examples of usage

c) Implement a second parsing algorithm. You could implement:

• the CYK algorithm, or

• a LR(1) parser

• Note: a recursive descent parser would not be sufficient (too easy).

You will need to submit your code to PASTA, as well as including some details in your report

comparing your second parser to the first one (what are the advantages and disadvantages of this

parser compared to the LL(1) table driven parser.)

d) Something else? Check with your tutor to see if they think it’s appropriate.

8 Submission details

Due 23:59pm Sunday 20th October 2019. The late submission policy is detailed in the administrivia lecture

slides from week 1. Please notify us if you intend to make a late submission, or if you believe you will not

be able to submit, to make it easier for us to support you.

8.1 Canvas submission

You must submit a report as a single document (.pdf or .docx) to TurnItIn via Canvas. The written parts

of the report must be text, not images of hand-writing. Any diagrams can be images, of course. The report

should include:

• Task 1 – 5: Your answers, working, and explanations

• Task 6: A description of the testing runs you used, including examples of the output. Show enough

to convince the marker that your testing was comprehensive

• Task 7: A description of your extension, including documentation about how to use it and some

examples of input/output

8.2 PASTA submission

Exact submission details (such as requirements on the name of the program) will be provided next week.

• Task 6: submit your source code.

• Task 7: submit your source code, if relevant

The visible tests for task 6 (the implementation) will be just enough to show you that you have got

the input/output syntax correct. It will not test the correctness of your program (you are expected to test

that yourself!)

There will be no automatic testing of any extra functionality you added as an extension.

8.3 Marking criteria

The weight of marks for each question are noted next to each question.

• Tasks 1 – 5: The marks will be roughly evenly divided between correctness, and on your working

and explanations.

• Task 6: The marks will be roughly evenly divided between automatic marking (for correctness) and

hand marking (based on your testing, the quality of your explanations, code, etc.)

• Task 7: The extension will be hand marked.

9 Appendices

9.1 Example of string derivations

Suppose we had a program which parsed this grammar fragment (your grammar will be different!):

E → (F) | T (start variable)

F → +EE

T → 0 | 1 | 2 | 3

Example 1: For the input:

The output might look like this (note: it doesn’t need to line up neatly):

Example 2: For the input:

The output is the same, because we ignore all the

Example 3: For the input:

It is rejected because there would be no entry in the parse table for (E, +).

9.2 Extension option A (error recovery)

Reminder: your code should only run the extension if a second command line argument error is given.

Otherwise it should behave normally.

This appendix shows a few examples of what using an error recovery feature might look like. Your

output and features do not need to be identical to these examples, it is only a guideline to give you some

ideas.

Suppose we had a program which parsed this grammar fragment (your grammar will be different!):

E → (F) | T (start variable)

F → +EE

T → 0 | 1 | 2 | 3

There are three key steps to implementing error recovery. More marks will be awarded to more useful/sophisticated

features.

9.2.1 Identifying the error

The first step is to be able to describe the error.

Error: got +, but expected {0, 1, 2, 3, (}.

Error: got ), but expected $.

9.2.2 Recovering with user intervention

The next step would be to give the user the option to correct the error and allow the derivation to continue.

For example, you might have an option to delete the next input symbol:

It’s a bit more powerful if we also allow the user to insert a symbol:

Error: got $, but expected {)}.

Better still, would be to allow the user to choose to delete or insert (so they can replace a symbol)

9.2.3 Making useful suggestions

You might automatically suggest which is the ‘best’ correction. For example, you might have the program

explore each option automatically, then recommend the correction which leads to an accepted string using

a minimal number of changes. For performance reasons, it would be wise to limit the maximum number

of changes to some fairly small number (e.g. no more than 5 changes).

9.3 Extension option B (evaluation)

Reminder: your code should only run the extension if a second command line argument eval is given.

Otherwise it should behave normally.

9.3.1 Semantics of the language:

• S := a sequence of one or more commands to execute.

• E := an expression to evaluate, possibly including variables

• (EBE) := apply the two expressions to some binary operator.

– arithmetic is normal integer arithmetic (e.g. 5 − 3 is −2)

– > := 1 if the inequality is true, 0 otherwise

• N := a strictly positive integer.

• V := a variable label.

• let V = E; := assign the result of evaluating E, to V

• E; := output (print) the result of evaluating E

• while E do S; := as long as E evaluates to a non-zero number, do S

• while E do S else S; := as long as E evaluates to a non-zero number, do the first S. If E evaluates

to zero (possibly after some repetitions), do the second S.

• If there is more than one expression in the program, they should be evaluated in order. The only

output is the E; statements!

9.3.2 Examples of output:

Input Output Why

(5+3); 8

(5-3); -2

let x=1; there is no E; statement

let x=1;x; 1

Statements are evaluated in order

let x=3;

while x do let x=(x-1); x;;

while x do let x=(x-1); x;

else 9;;

while x do let x=(x-1); x;

while (x>1) do let y=(x*y); let x=(x-1);;y;

120 5! = 120

let x=5;

while (x>1) do let x=(x-1);;x;

1 Variables are not scoped

9.3.3 More information:

In the resources section on Ed, in the category “Supplementary Material”, there are some resources to help

you:

• Slides explaining the concepts

• A worked example of the process of building a parse tree following this algorithm

• Some code examples of evaluating a parse tree for a simpler grammar (in Python, Java and C++)

联系我们

- QQ：99515681
- 邮箱：99515681@qq.com
- 工作时间：8:00-23:00
- 微信：codinghelp2

- Ee6435 Programming Homework 2020-05-30
- Computer Architecture Homework 3 2020-05-30
- Infs7450作业代做、Media Analytics作业代写、Pytho 2020-05-29
- 代写stats 782作业、代做r编程设计作业、代写data留学生作业、R课 2020-05-29
- 代写math223作业、R课程设计作业代做、代写data课程作业、R程序语言 2020-05-28
- 代写kxo151留学生作业、代做programming作业、Java语言作业 2020-05-28
- Math 160作业代做、Data课程作业代写、Matlab程序设计作业调试 2020-05-28
- 代做csci 3120作业、Program课程作业代做、C/C++语言作业代 2020-05-28
- St2020课程作业代做、Statistics作业代做、R程序设计作业调试、 2020-05-28
- Com1005作业代写、Ai Techniques作业代做、Java语言作业 2020-05-28
- 代写sit216留学生作业、Python程序语言作业调试、Java/C++实 2020-05-28
- 代写artificial课程作业、Java，Python程序语言作业调试、C 2020-05-27
- Comp Sci 3306作业代写、Python编程语言作业调试、代做jav 2020-05-27
- Data留学生作业代写、代做r课程设计作业、Analytics作业代做、R编 2020-05-27
- Csci 3120作业代做、C++程序语言作业调试、代做c/C++课程作业、 2020-05-26
- 代写algorithms作业、Data留学生作业代做、代写java、Pyth 2020-05-26
- Data Science作业代写、C++程序设计作业代写、Programmi 2020-05-26
- Data课程作业代写、C++编程设计作业调试、C/C++语言作业代做、Alg 2020-05-26
- 代写r留学生作业、代做data课程作业、代写r编程语言作业代做r语言编程|调 2020-05-25
- Cosc473作业代做、Systems作业代写、Python编程设计作业调试 2020-05-25