Assignment 1

  +125

Prerequisite

First, set up your Git account Java development environment by following the instructions on the software setup page.

Task

Suppose we want to compile a language with a comment syntax similar to C/C++. A comment begins with the two characters (* and ends with the characters *). Those matching sequences and any characters between them should be ignored. Comments can contain newlines. They may not be nested. A program that ends with an unterminated comment is an error.

You should create a new Java project called assn1 in a fresh sub-directory of your Git repository. Within the project, create a class CommentLexer with a method stripComments defined like this:

    public static String stripComments(String source) {
      // TODO...
    }

The objective of stripComments is to remove the comments (as described above) from the source string, and return a new string containing only the non-comment portions of the source string. For example, stripComments("Hello(*cruel*)World") should return "HelloWorld".

You may want to think about the procedure in terms of a finite state automaton. What are the possible states and transitions? Are there any actions you need to take on certain transitions?

To submit your program, make sure the files in your assn1 directory are added to Git VCS, then commit and push. Some helpful hints follow.

Tips

  • I have set up a sample project at https://git.liucs.net/cs664s16/cs664pub/tree/master/assn1 (look in the src sub-directory). In addition to a template for the CommentLexer class, I provide a substantial test suite in CommentLexerTest and a class you can use as an exception to throw in the case of an unterminated comment. To try it out, you can clone the entire cs664pub project into your system (just like you did for your own project in the software setup) – just don’t do it within your own Git repository, save it somewhere else. The repository URL is git@git.liucs.net:cs664s16/cs664pub.git

  • The test suite uses jUnit 4. It’s very easy to add this library to an IntelliJ project. Go to File » Project Structure and select Libraries from the left panel. Then hit the green plus sign in the next column and select From Maven. In the search box, type junit:junit:4.12 and hit the search icon. Once it finds it, you can press OK. Confirm that you want to incorporate the library into your assn1 module.

  • To run the test suite, select Run » Run… and it should automatically show CommentLexerTest as an option. Once it’s configured, you can just use Run » Run CommentLexerTest or the green Play button.

  • The best way to build up a string incrementally in Java is to use StringBuilder. So here is an example of an implementation that keeps only the odd-numbered characters in the source string. So “banana” would become “aaa”. It won’t pass the tests, but it does show the correct usage of StringBuilder.

    public static String stripComments(String source) {
    StringBuilder buffer = new StringBuilder();
    for(int i = 0; i < source.length(); i++) {
        char current = source.charAt(i);
        if(i % 2 == 1) { // odd-numbered characters
            buffer.append(current);
        }
    }
    return buffer.toString(); // convert builder to regular string
}

    If you try to use this implementation in the test suite, the failures will show you what it does:

    org.junit.ComparisonFailure:
Expected :This (has balanced) parens
Actual   :hs(a aacd aes

  • The example from the language theory notes on how to implement an FSA in Java might also be helpful.

PDF