lab04 : Polynomials

num ready? description assigned due
lab04 true Polynomials Mon 10/07 05:00PM Fri 10/18 11:59PM

UPDATE 6:20PM, Thu October 17, 2019

If you get a warning about the summary attribute on the <table> tag when generating Javadoc for lab04, try changing this code in Polynomial.java from:

<table summary="each row ...">

to just:

<table>

I read in the JN7 text that <table> elements were supposed to have summary attributes for WAI compliance, but as it turns out, the version of javadoc we are using doesn’t like that attribute.

You may also need to change <sup>i<sup> to <sup>i</sup>. That’s a typo.

UPDATE: 7pm, Wed October 16, 2019

In order to get the Javadoc to publish correctly to GitHub Pages, you will need to add this to your pom.xml.

It should go:

  <!-- For use with github pages, to publish the site to the /docs subdirectory -->	
  <distributionManagement>	
    <site>	
      <id>website</id>	
      <url>file://${project.basedir}/docs/</url>	
    </site>	
  </distributionManagement>

You should also find this part of the pom.xml by searching for the string org.apache.maven.plugins

<plugin>
	<groupId>org.apache.maven.plugins</groupId>
	<artifactId>maven-site-plugin</artifactId>
	<version>3.3</version>
</plugin>

And add in these lines:

    <configuration>
       <dependencyLocationsEnabled>false</dependencyLocationsEnabled>
    </configuration>

This tells Maven to skip the “dependency report” when generating the site. This will make the site:deploy command go a lot faster.

Then, make sure that your .gitignore has this in it:

docs/jacoco

But that it does NOT have this in it:

docs

That should allow you to do:

mvn javadoc:javadoc site:site site:deploy

And then, if you do:

git add docs
git commit -m "add javadoc to repo"
git push origin master

And if you go to the settings for your repo and turn on GitHub pages for the docs folder of the master branch, you should be able to see the docs at the url:

In this lab:

Step-by-Step

Step 0: Set up your repo

This lab will build on the material we introduced about Maven and JUnit in previous labs.

You will also apply what you are learning from your reading about inheritance, object equality, and using the ArrayList class.

You may work as a pair, or individually on this lab (your choice). Name your repo as appropriate.

Create your repo the same way you did for a previous lab:

Then, create an empty directory for lab04, do a git init and then add remotes just as you did in previous labs.

The starter code is in https://github.com/ucsb-cs56-f19/STARTER-lab04. Visit that page for the approrpiate URL to add the starter remote.

If pairing, register your pair on Gradescope

On Gradescope, you can mark your submission as a pair submission. Be sure to do this if you are working in a pair.

Working in a pair? Switch navigator/driver frequently and tradeoff who commits

If you are in your repo directory, and type git log at the command line, you’ll see a list of the commits for your repo.

Record that you are pairing on each commit message by putting the initials of the pair partners at the start of the commit message.

E.g. If Selena Gomez is driving, and Justin Timberlake is navigating, and you fixed a bug in your getDanceMoves() method, your commit message should be SG/JT fixed bug in getDanceMoves()

Step 1: A note about the .git directory and the .gitignore file

If you aren’t already there, cd into your repo directory.

As you may know, in Unix, “hidden files” are those that start with a dot.

The -a flag of the Unix ls command can be used to list hidden files in directory listings.

What is the .git directory?

NOTE: The contents of this section may be something you are tested on on Midterm Exam 1 and subsequent exams, so READ CAREFULLY!

The .git directory is the magic ingredient that turns an ordinary directory into a git repository.

For now, that’s all you need to know about the contents of the .git directory. However, if you want to learn more, you are encouraged to read ahead in the GIT pocket guide. The internals of git are fascinating, and the more you know, the more power you will have to make git do all kinds of wonderful tricks.

What is the .gitignore file?

The .gitignore file is NOT inside the .git directory, but instead is considered a regular file in the repository.

BUT, the .gitignore file has special powers over the way that git operates.

Generally speaking, when we set up a repository, we want to store only SOURCE CODE:

We also don’t want to store backup files, such as the files that emacs creates that end in tilde, such as Foo.java~, Bar.java~, etc.

The .gitignore helps us do that. Use the more command to list the contents of your .gitignore file, which will look something like this (if you created a .gitignore for Java and it has more contents than this, that is fine):

-bash-4.2$ more .gitignore
*~
*.class
-bash-4.2$

You see that there are two lines in this file.

You will see that when you do “git status” in this repo, the backup and .class files don’t show up. They are ignore by git. And that’s the way we want it.

If you use emacs, you may want to make sure that this is in your .gitignore

*~

Summary of this Step

In short:

Step 3: Publish Javadoc and Fix your README!

These steps are needed to get full credit for the lab. Up to 10% may be deducted for not following them.

Step 4: Java Programming

Your programming task in this lab can be described very simply: replace all the stubs in Polynomial.java with code that

It is possible (though unlikely) that there may be typos/errors in the unit tests themselves. If you suspect this, please post something on the typos channel on Slack. The FIRST person/pair partner to CORRECTLY identify each incorrect unit test (supplying a correct test in its place) will receive extra credit.

If there are corrections to the tests, that information will be sent out to you by email and posted on the “Instructor Announcements” forum in Gauchospace.

Do NOT simply change the PolynomialTest.java file. We will test your program against the original PolynomialTest.java file (with any corrections that are distributed).

You may ADD tests to PolynomialTest.java, but you should not CHANGE the tests that are listed there or delete tests, unless you have been told to do so in an email from the instructor.

I suggest that you work in this order. There are also three hints for debugging that follow this suggested order of work. You may want to read though ALL of the instructions in Step 2 before starting anything.

Suggested order of work

  1. Start by commenting out the toString method temporarily.
    • That will cause the class to use the ArrayList `toString` method instead. That way, while you work on other methods, you'll be able to see the ground truth of what's inside the Arrays.
    • Come back and uncomment it when you are ready to work on it.
  2. You might want to do this so that you can focus on one method at a time:
    • cd src/test/java/edu/ucsb/cs56/polynomial
    • cp PolynomialTest.java PolynomialTest.java.keep
    • Then, delete the code inside PolynomialTest.java for all tests except the method you are working on.
    • Copy/paste the tests back into the file one at a time, and try to get one test at a time to pass.
    • Run the entire test suite each time, to make sure that when you get one test to pass, you don’t break something else.
  3. I suggest you then focus on these methods first, since they don’t depend on anything else, and they are useful when getting other things to work:
    • public static int degreeOfPolynomialCoeffsLowToHigh(int [] coeffsLowToHigh)
    • public static int degreeOfPolynomialCoeffsHighToLow(int [] coeffsHighToLow)
    • public static int [] lowToHigh(int [] coeffsHighToLow)
    • public static int [] highToLow(int [] coeffsLowToHigh)
  4. Then, tackle these as a group. There is a bit of a chicken and egg problem here, because the equals method is used by JUnit, and until it is correct, the JUnit test results may be misleading! But you can’t test equals unless you have a way to construct Polynomial objects. You’ll have to think about how to handle that.

    • public boolean equals(Object o)
    • public Polynomial(int [] coeffsHighToLow)
    • public int getDegree()
  5. Next, before tackling the plus, times, and minus, get toString working. Bugs in toString can be maddening because the JUnit output uses toString. You’ll be looking at test case output, and it won’t represent what’s really going on. So get this working and solid before moving on.

  6. Finally, I suggest doing plus, times and minus, in that order. There is a reason for doing it that way; it will save you work if you are clever.

  7. When you are all done, you can submit on Gradescope. Note that there may be extra tests on Gradescope for cases that you didn’t consider.

Helper functions are a good idea

If you find that a certain method is getting a bit too complex, or if you find that you are repeating yourself a lot, you may want to factor some code out into helper methods.

There are two approaches to dealing with helper methods, each of which can be argued:

My suggestion is: make them private UNLESS you write unit tests. And writing unit tests is a VERY GOOD IDEA. But, that’s up to you.

Additional debugging help

There are three additional things that can help while doing your debugging:

  1. There is a main program that you can run to test basic Polynomial operations and printing of Polynomials. Here are some examples of the Unix command line to run that program. Note that you need single quotes around the Polynomials, and that the * operator needs a backslash in front of it (so the Unix shell doesn’t turn into a “wildcard” for filenames.) You can use this main to see what happens when you take in a Polynomial and then print it back out with the toString() method. If you pass just ONE polynomial in the command line, it simply gets “constructed” from the string, and then printed back out with toString. If you are using “debug print statements” (see item 2 below) then it can be hard to make sense of those in the JUnit test output. But, running the main, you can isolate what’s happening with just ONE polynomial at a time, and see your debugging print statements in a less cluttered, sensible order.
    • java -cp build Polynomial '10x^2 - 20x + 3'
    • java -cp build Polynomial '2x + 3' + '4x - 5'
    • java -cp build Polynomial '2x + 3' \\\* '4x - 5'
    • java -cp build Polynomial '2x + 3' - '4x - 5'
  2. There is a public static boolean variable called debug inside the Polynomial class. You can set this to true, and then print debugging output. You’ll see examples of that in the constructor that creates a Polynomial object from a String.

  3. If you need to, for debugging purposes, you may add an additional class called PolynomialDebugging. That class might look something like this:

    
        public class PolynomialDebugging {
    
          public static void main (String [] args) {
            if (args[0]=="1") {
                // Put some code you want to test here
             } else if (args[0]=="2") {
                // Put some code you want to test here
             } else 
             .. // continue in this fashion
             }
          }
        } 
    

    The point of this is to be able to isolate one small thing that you are trying to debug, independent of all the confusing output of all the many test cases, and so you can see your debugging print statements more easily.

    If you do this, please DO git add, git commit, git push the PolynomialDebugging class. Note that we won’t grade that class—and you aren’t required to use one. This is just a suggestion in case you find it helpful.

Other Notes

By the way, the constructor that creates a Polynomial object from a String is already written for you, and is, to the best of our knowledge, correctly implemented. It should pass its tests once the other methods are correctly implemented. However, until you get toString, the constructor that takes an int [], and the .equals method working, however, some of its tests may fail.

When all the tests pass, you may then add additional tests for other kinds of polynomials (both well formed and badly formed) that are not covered by the tests that you see in PolynomialTest.java. The number of such tests is up to you. There is such as thing as too many tests, and there is such a thing as not enough tests. Finding the right balance is part of the exercise here.

The reason for adding the additional tests is to try to anticipate what the “secret” tests might be testing for, i.e cases not already covered by the tests in PolynomialTest.java.

When you feel you have tested sufficiently, and all your tests pass, you are ready for your final check over, then submission on Gauchospace.

Step 3: Preparing to submit: final code walkthrough against rubric

Now, you are almost done. The last four steps are to:

Checking over your code

First, check your code Check it against two things:

  1. ALL OF THE INSTRUCTIONS IN THIS LAB ASSIGNMENT. Start over at the very first step. If you are working a pair, have one pair partner read through the instructions, and have the other pair partner check the code (or in the case of the javadoc, check both the code, and what appears on the web.)
  2. ALL OF THE ITEMS IN THE GRADING RUBRIC. The grading rubric for this lab appears near the end of this file. It is the checklist that the TAs and instructor will use to determine your lab grade. Again, if working in a pair, divide up the responsibility (switching roles)—one person read the rubric items out loud, and the other person checks.

Even if you “solo programmed” this lab, you may want to see if you can find someone in the lab that also solo programmed, and ask him/her to be a “rubric buddy” with whom you can take turns doing this checklist step.

Updating javadoc

If you are working along, right before you do the final submission, do a “git commit”, “git push” and then run “mvn javadoc:javadoc”.

Then do the steps to publish the updated javadoc to your public repo.

Final git pull, git status, git add, git commit, git push origin master

Go through the git steps one last time in case you made any last minute changes. Be sure that your repo is up to date. When it is, you are ready to submit on Gauchospace.

Grading:

There may be additional deductions, however if you failed to follow any of the following instructions: