CSE333 Homework 1

Out:   Friday, January 10, 2020
Due:   Thursday, January 23, 2020 by 11:59 pm


Goals

You will finish our implementation of two C data structures: a doubly-linked list (Part A) and a chained hash table (Part B). You will also maintain a bug journal as you debug your project.

Please read through this entire document before beginning the assignment, and please start early! This assignment involves messy pointer manipulation and malloc/free puzzles, and these can cause arbitrarily awful bugs that take time and patience to find and fix.

All CSE 333 homeworks are only supported on attu and the CSE VM. We do not support building and running this assignment in any other work environments.


Bug Journal

You are expected maintain a bug journal as you work on each homework this quarter. This is a lightweight way to help you focus on how you approach debugging and should not add much time, assuming that you fill it out as you encounter bugs. In fact, its intent is to save you time overall by improving your debugging skills!

In each homework, you will find a file in the directory called bugjournal.md with a skeleton for the prompts that you will find below for you to fill out. You may treat this file as a standard text file.

Bug Journal Prompts

For at least three bugs that you encounter during this homework, track your progress in solving them by responding to the following prompts:

  1. What is the current (incorrect) behavior and what is the expected behavior?
    Examples:
    • I am getting a segault at line 100, which is supposed to move a pointer to the next node in a linked list.
    • My output from function_name() doesn't match the solution binaries.
    • I am getting a memory leak of 128 bytes allocated in function_name(), line 140.
  2. Brainstorm a few possible causes of this bug. We just want ideas; you don't have to investigate these yet. Possible causes often start with phrases like "I think...", "Maybe", or "If X happens, then it would cause Y to fail." Try to be as specific as you can.
    Examples:
    • I think that foo() returns NULL when it should be returning a valid heap-allocated pointer.
    • When we reach the if-statement at line 130, it should take the else branch, but instead it takes if branch.
    • When foo() calls bar(), a heap-allocated pointer is returned, but it is never deallocated.
  3. Briefly describe your fix of the error AND explain why it was necessary. First, investigate the possible causes from part B. If none of these yields a solution, try brainstorming more with a peer (no code!) or TA. You may need to be more specific about the cause of the error.
    Examples:
    • Removed the call to free() at line 108 because the pointer was returned and cannot be deallocated yet.
    • Changed the while-loop condition from (ptr->next != NULL) to (ptr != NULL), because we weren't processing the last element of the list with the other condition.

Important: While we would prefer if you focused on "major" bugs, you may journal about smaller bugs (quickly fixed or without the use of a debugger) if you encountered fewer than three major bugs.

What if I encountered fewer than three bugs?

That's great! Instead of journaling your bugs, reflect on your programming process by describing two processes that you used to help you avoid bugs, writing a few sentences for each.

Example:

  • I wrote pseudocode using inline comments to draft out the implementation for each function. This helped me organize my thoughts and helped me spot potential problems before I wrote the code instead of spending time debugging later.

C Data Structures

Part A: Doubly-Linked List

If you've programmed in Java, you're used to having a fairly rich library of elemental data structures upon which you can build, such as vectors and hash tables. In C, you don't have that luxury: the C standard library provides you with very little. In this assignment, you will add missing pieces of code in our implementation of a generic doubly-linked list.

At a high-level, a doubly-linked list looks like this:


Each node in a doubly-linked list has three fields: a payload, a pointer to the previous element in the list (or NULL if there is no previous element), and a pointer to the next element in the list. If the list is empty, there are no nodes. If the list has a single element, both of its next and previous pointers are NULL.

So, what makes implementing this in C tricky? Quite a few things:

  • First, we want to make the list useful for storing arbitrary kinds of payloads. In practice, this means the payload element in a list node needs to be a pointer supplied by the customer of the list implementation. Given that the pointer might point to something malloc'ed by the customer, this means we might need to help the customer free the payload when the list is destroyed.

  • Second, we want to hide details about the implementation of the list by exposing a high-level, nicely abstracted API. In particular, we don't want our customers to fiddle with next and previous pointers in order to navigate through the list, and we don't want our customers to have to stitch up pointers in order to add or remove elements from the list. Instead, we'll offer our customers nice functions for adding and removing elements and a Java-like iterator abstraction for navigating through the list.

  • Third, C is not a garbage-collected language: you're responsible for managing memory allocation and deallocation yourself. This means we need to be malloc'ing structures when we add nodes to a list, and we need to be free'ing structures when we remove nodes from a list. We also might need to malloc and free structures that represent the overall list itself.

Given all of these complications, our actual linked list data structure ends up looking like this:


Specifically, we define the following types and structures:

  • LinkedList: The structure containing our linked list's metadata, such as head and tail pointers. When our customer asks us to allocate a new, empty linked list, we malloc and initialize an instance then return a pointer to that malloc'ed structure to the customer.

  • LinkedListNode: This structure represents a node in a doubly-linked list. It contains a field for stashing away (a pointer to) the customer-supplied payload and fields pointing to the previous and next LinkedListNode in the list. When a customer requests that we add an element to the linked list, we malloc a new LinkedListNode to store the pointer to that element, do surgery to splice the LinkedListNode into the data structure, and update our LinkedList's metadata.

  • LLIterator: Sometimes customers want to navigate through a linked list; to help them do that, we provide them with an iterator. LLIterator contains bookkeeping associated with an iterator. In particular, it tracks the list that the iterator is associated with and the node in the list that the iterator currently points to. Note that there is a consistency problem here: if a customer updates a linked list by removing a node, it's possible that some existing iterator becomes inconsistent because it referenced the deleted node. So, we make our customers promise that they will free any live iterators before mutating the linked list. (Since we are generous, we do allow a customer to keep an iterator if the mutation was done using that iterator.) When a customer asks for a new iterator, we malloc an instance and return a pointer to it to the customer.

Instructions

You should follow these steps to do this assignment:

  1. Make sure you are comfortable with C pointers, structures, malloc, and free. We will cover them in detail in lecture, but you might need to brush up and practice a bit on your own; you should have no problem Googling for practice programming exercises on the Web for each of these topics.

  2. Get the source files for hw1. Navigate to a directory that contains a checked-out copy of your cse333 Git repository and run the command git pull. (See the CSE 333 Git Tutorial for some tips if the pull command fails because you have unstaged changes or other problems.) After the pull command finishes you should see the following directories and files in your repository:
    bash$ ls
    clint.py  gtest  hw0  hw1
    

  3. Look inside the hw1 directory. You'll see a number of files and subdirectories, including these that are relevant to Part A:

    • Makefile: a makefile you can use to compile the assignment using the Linux command make all on attu and the VM.

    • LinkedList.h: a header file that defines and documents the API to the linked list. A customer of the linked list includes this header file and uses the functions defined within in. Read through this header file very carefully to understand how the linked list is expected to behave.

    • LinkedList_priv.h & LinkedList.c: LinkedList_priv.h is a private header file included by LinkedList.c; it defines the structures we diagrammed above. These implementation details are typically withheld from the client by placing its contents directly in LinkedList.c; however, we have opted to place them in a "private .h" instead so that our unittest code can verify the correctness of the linked list's internals. LinkedList.c contains the partially completed implementation of our doubly-linked list. Your task will be to finish the implementation. Take a minute and read through both files; note that there are a bunch of places in LinkedList.c that say "STEP X:" these labels identify the missing pieces of the implementation that you will finish.

    • example_program_ll.c: this is a simple example of how a customer might use the linked list; in it, you can see the customer allocating a linked list, adding elements to it, creating an iterator, using the iterator to navigate a bit, and then cleaning up.

    • test_linkedlist.cc: this file contains unit tests that we wrote to verify that the linked list implementation works correctly. The unit tests are written to use the Google Test unit testing framework, which has similarities to Java's JUnit testing framework. As well, this test driver will assist the TA in grading your assignment: as you add more pieces to the implementation, the test driver will make it further through the unit tests, and it will print out a cumulative score along the way. You don't need to understand what's in the test driver for this assignment, though if you peek inside it, you might get hints for what kinds of things you should be doing in your implementation!

    • solution_binaries: in this directory, you'll find some Linux executables, including example_program_ll and test_suite. These binaries were compiled with a complete, working version of LinkedList.c; you can run them to explore what should be displayed when your assignment is working!

  4. Run make on attu or the VM to verify that you can build your own versions of example_program_ll and test_suite. make should print out a few things, and you should end up with new binaries inside the hw1 directory.

  5. Since you haven't yet finished the implementation of LinkedList.c, the binaries you just compiled won't work correctly yet. Try running them, and note that example_program_ll halts with an assertion error or a segfault and test_suite prints out some information indicating failed tests, and may crash before terminating.

  6. This is the hard step: finish the implementation of LinkedList.c. Go through LinkedList.c, find each comment that says "STEP X", and place working code there (please keep the "STEP X" for your graders' sanity, though!). The initial steps are meant to be relatively straightforward, and some of the later steps are trickier. You will probably find it helpful to read through the code from top to bottom to figure out what's going on. You will also probably find it helpful to recompile frequently to see what compilation errors you've introduced and need to fix. When compilation works again, try running the test driver to see if you're closer to being finished.

    • Note: You may not modify any header files or interfaces in this or later project assignments. We may test your code by extracting your implementations and compiling them with the original header files or in some other test harness where they are expected to behave as specified. You certainly are free, of course, to add additional private (i.e. static) helper functions in your implementation, and you should do that when it improves modularity.

    • Debugging hint: Verify333 is used in many places in the code to check for errors and terminate execution if something is wrong. You might find it helpful to discover the function that is called when this happens so you can place a debugger breakpoint there.

  7. We'll also be testing whether your program has any memory leaks. We'll be using Valgrind to do this. To try out Valgrind for yourself, do this:

    • from the hw1 directory run the following command:
      valgrind --leak-check=full ./solution_binaries/example_program_ll
      Note that Valgrind prints out that no memory leaks were found. Similarly, try running the test driver under Valgrind:
      valgrind --leak-check=full ./solution_binaries/test_suite
      and note that Valgrind again indicates that no memory leaks were found.

    • While still in the hw1 directory, compile your versions of the example_program_ll and test_suite binaries, and try running them under Valgrind. If you have no memory leaks and the test_suite runs the linked list tests to completion, you're done with Part A!


Part B: Chained Hash Table

A chained hash table is a data structure that consists of an array of buckets, with each bucket containing a linked list of elements. When a user inserts a key/value pair into the hash table, the hash table uses a hash function to map the key into one of the buckets, and then adds the key/value pair onto the linked list. (As an important corner case, if the key of the inserted key/value pair already exists in the hash table, our implementation of a hash table replaces the existing key/value pair with the new one, and returns the old key/value pair to the customer.)

Over time, as more and more elements are added to the hash table, the linked lists hanging off of each bucket will start to grow. As long as the number of elements is a small multiple of the number of buckets, lookup time is small: you hash the key to find the bucket, then iterate through the chain (linked list) hanging off the bucket until you find the key. As the number of elements gets longer, lookup gets less efficient, so our hash table includes logic to resize itself to maintain short chains.

As with the linked list in Part A, we've given you a partial implementation of a hash table. Our hash table implementation looks approximately like this:


Specifically, we defined the following types and structures:

  • HashTable: The structure containing our hash table's metadata, such as the number of elements and the bucket array. When our customer asks us to allocate a new, empty hash table, we malloc and initialize an instance (including malloc'ing space for the bucket array inside it, and allocating LinkedLists for each bucket), and return a pointer to that malloc'ed structure to the customer.

  • HTIterator (not shown in the diagram): Sometimes customers want to iterate through all elements in the hash table; to help them do that, we provide them with an iterator. HTIterator points to a structure that contains bookkeeping associated with an iterator. As before, it tracks the hash table that the iterator is associated with as well as a linked list iterator. When a customer asks for a new iterator, we malloc an HTIterator and return a pointer to it.

Instructions

You should follow these steps to do this assignment:

  1. The code you fetched in Part A also contains the files you'll need to complete your hash table implementation and test it. Similar to the linked list, the hash table implementation is split across a few files: HashTable.c contains the implementation you need to finish, HashTable.h contains the public interface to the hash table and documents all of the functions & structures that customers see, and HashTable_priv.h contains some private, internal structures that HashTable.c uses.

  2. Read through HashTable.h first to get a sense of what the hash table interface semantics are. Then, take a look at example_program_ht.c; this is a program that uses the hash table interface to insert/lookup/remove elements from a hash table, and uses the iterator interface to iterate through the elements of the hash table.

  3. As before, test_hashtable.cc contains our Google Test unittests for the hash table. Run this – on its own, and using valgrind – to see how close you are to finishing your hash table implementation.

  4. Look through HashTable.c, find all of the missing pieces (identified by "STEP X" comments, as before), and implement them.

  5. As before, in solution_binaries, we've provided you with linux executables (i.e. example_program_ht and the same test_suite) that were compiled with our complete, working version of HashTable.c. You can run them to explore what should be displayed when your part B implementation is working and look at the source code for examples of how to use the data structures.


Bonus: Code Coverage Statistics

You'll note that we provided a second Makefile called Makefile.coverage. You can use it to invoke the gcov code coverage generation tool. Figure how to (a) use it to generate code coverage statistics for LinkedList.c and HashTable.c, (b) note that the code coverage for HashTable is worse than that for the LinkedList, and (c) write additional HashTable unit tests to improve HashTable's code coverage.

The bonus task is simple, but we're (deliberately) providing next to no detailed instructions on how to do it – figuring out how is part of the bonus task!

Please make sure your additional unit tests don't change the scoring mechanism that we use, obviously (we'll be checking that). Place your additional unit tests in a separate file from the original test suite. That will make it easier for us to find and evaluate your tests.


Submission

When you are ready to turn in your assignment, you should follow exactly the same procedures you used in hw0, except this time tag the repository with hw1-final (instead of hw0-final). Remember to clean up and commit and push all necessary files to your repository before you add and push the tag.

After you have created and pushed the tag, be absolutely sure to test everything ON ATTU OR THE VM by creating a new clone of the repository in a separate, empty directory, checkout the hw1-final tag, and verify that everything works as expected. Refer to the hw0 submission instructions for details and follow those steps carefully.

If you fail to check your work and your project doesn't build properly when the same steps are done by the course staff to grade it, you may lose a huge amount of the possible credit for the assignment even if almost absolutely everything is actually correct.


Grading

We will be basing your grade on several elements:

  • The degree to which your code passes the unit tests contained in test_linkedlist.cc and test_hashtable.cc. If your code fails a test, we won't attempt to understand why: we're planning on just including the number of points that the test drivers print out.

  • We have some additional unit tests that test a few additional cases that aren't in the supplied test drivers. We'll be checking to see if your code passes these as well.

  • The quality of your code. We'll be judging this on several qualitative aspects, including whether you've sufficiently factored your code and whether there is any redundancy in your code that could be eliminated.

  • The readability of your code. While we don't have detailed, formal coding style guidelines that you must follow, you attempt to mimic the style of code that we've provided you. Aspects you should mimic are conventions you see for capitalization and naming of variables, functions, and arguments, the use of comments to document aspects of the code, and how code is indented. We strongly suggest using the clint.py tool to check your code for style issues. You also will find it useful to refer to the Google C++ Style Guide (linked from the course web site). Much of this guide applies equally well to C.

  • Reasonable effort has been put into the completion of your bug journal.