---

Problem 1 (3 points)

Background

If you've ever used a web browser before, there's a good chance that you've seen a Uniform Resource Locator (URL) before. URLs provide a way to identify resources on the web — as well as other places online, as they're actually a more generic concept than you might at first think. For example, the URL for this page is below.

• https://www.ics.uci.edu/~thornton/ics32/Exercises/Set3/index.html

Because I've rendered the URL, on this page, as a link you can click, you'll be able to click that link and wind up right back where you started: on this page. But the URL is really just the text. One way you could store that into a variable in Python would like this.

url = 'https://www.ics.uci.edu/~thornton/ics32/Exercises/Set3/index.html'

However, not every string is a valid URL; only some strings are. URLs, as it turns out, are a much more complex topic than they might appear at first blush. The original specification defining URLs, from 1994 (when I was about a year ahead of where you are in your studies now), weighs in at about 25 pages, which is the original definition of what is and isn't a valid URL, as well as what various URLs are intended to mean. We'll talk a lot more in detail about URLs and the web in the not-too-distant future in this course, but, for now, the basic idea suffices: A URL tells you where something is on the web, and there are rules governing whether a URL is valid and what it means.

The problem

Let's start with a simplification of the URL, which we'll call an SRL (Simplified Resource Locator). (There's no such thing as an SRL; we're making this up as we go along.) Suppose that these are the rules governing SRLs:

• SRLs are sequences of text characters. Only some sequences are correct, and parts of that sequence have different meanings.
• An SRL begins with a sequence of one or more lowercase letters (from 'a' through 'z'), denoting a kind, which specifies how to get whatever the SRL refers to.
• The kind is followed by a colon and two slashes. Those must be there.
• Next, there should be a location, which specifies the location of a machine where the resource is.
  • A location is a sequence of one or more location parts, separated by dot ('.') characters. (Note that "separated" means that a dot cannot appear at the beginning or the end of the location.)
  • Each part is a sequence of one or more characters that are either uppercase letters ('A' through 'Z'), lowercase letters ('a' through 'z'), or digits ('0' through '9').
• After the location is a resource, which specifies what resource on that machine we're interested in.
  • A resource is a sequence of one or more resource parts, each preceded by a slash ('/') character. (Note that "preceded" means that a slash must come before each resource part, but a slash will not appear on the end.)
  • Each part is a sequence of one or more characters that are either uppercase letters ('A' through 'Z'), lowercase letters ('a' through 'z'), or digits ('0' through '9').

If we wanted to write a program in Python that used or manipulated SRLs, we'd have two somewhat related problems to solve:

1. Use our understanding of the structure of SRLs mean to ensure that we never have any that are invalid.
2. Given a valid SRL, use our understanding of what SRLs mean to be able to access its various component parts.

We could use strings to store SRLs, but strings don't solve either of the most important problems we need to solve. As a result, we'd have to spread that knowledge throughout our program: Any time we built one, we'd have to be careful that it be valid, while any time we used one, we'd have to slice it up to understand what it means.

What would be better is a Python class that solves these problems. If we had a special type that knew about all of these details, we could have enforcement of the rules around SRLs, as well as one-stop-shopping for understanding what's inside of them, similar to what pathlib did for us in Project #1.

Write a class called Locator. A Locator object represents an SRL, and provides the following methods.

• def __init__(self, srl: str), which takes a string (an SRL) as a parameter and does one of two things:
  • If the given string is a valid SRL, according to the rules above, initializes the Locator object so that subsequent calls to methods work as they should.
  • If the given string is not a valid SRL, according to the rules above, a ValueError is raised instead.
• def srl(self) -> str, which returns the entire SRL represented by a locator.
• def kind(self) -> str, which returns the locator's "kind."
• def location(self) -> str, which returns a string representing the locator's "location." Note that this is the entire location (e.g., www.ics.uci.edu).
• def location_parts(self) -> list[str], which returns a list of strings representing the parts of a locator's "location." If the location is www.ics.uci.edu, this method would return ['www', 'ics', 'uci', 'edu'].
• def resource(self) -> str, which returns the locator's "resource." Note that this is the entire resource (e.g., /abc/def/ghi).
• def resource_parts(self) -> list[str], which returns a list of strings representing the parts of a locator's "resource." If the resource is /abc/def/ghi, this method would return ['abc', 'def', 'ghi'].
• def parent(self) -> 'Locator', which returns a new Locator object where the last resource part has been removed. For example, given a Locator object representing w://x.y.z/a/b/c, calling parent() on it would return a Locator object representing w://x.y.z/a/b.
  • Note that the parent of a Locator whose resource has one resource part is the Locator itself.
• def within(self, resource_part: str) -> 'Locator', which returns a new Locator object where the given resource part is appended to the end of the SRL. For example, given a Locator object representing w://x.y.z/a/b/c, calling within('d') on it would return a Locator object representing w://x.y.z/a/b/c/d.

Along with your class, you are required to write assert-based tests that exercise its functionality.

Safely testing with variables

As we've seen, assert-based tests are best written at the "top level" of a module, which is to say that they appear outside of any functions, interspersed with the code they test. For "pure" functions, which are those whose outputs are determined solely by their inputs, each assert statement can be self-contained, because we won't need any variables to store values; we just write expressions that are evaluated in place, checked by assert, and then, if they succeed, can safely be thrown away.

def duplicate(s: str) -> str:
    return s + s

assert duplicate('Boo') == 'BooBoo'
assert duplicate('z') == 'zz'
assert duplicate('') == ''

But what do we do when we want to test classes like the one we've asked you to build in this problem? In this case, we need to be able to create an object, store it in a variable, then call methods on it and see what their results are; it's not one function call we're testing, but sequences of behavior that we're testing. Before too long, we'll start using testing tools that allow us to do this kind of thing more cleanly, but all is not lost, even if all we're using is assert; we just need a way to isolate our tests, so that any variables we use aren't global.

The way we'll solve that problem is to write test functions. The test functions will contain assert statements, but then we'll make sure there are calls to these functions at the top level — ideally, right after each test function, so we have an obvious place to look for them. This way, any variable we create will be scoped to a function, which means it'll only last long enough to run our tests; that's a lot better than letting it be global to our module, where it will be visible to any code that uses our module.

# Imagine this is the Counter class from lecture
class Counter:
    ...


def test_counters_count_upward() -> None:
    c1 = Counter()
    assert c1.count() == 1
    assert c1.count() == 2
    assert c1.count() == 3

test_counters_count_upward()

def test_peek_leaves_counters_unchanged() -> None:
    c1 = Counter
    c1.count()
    c1.count()
    assert c1.peek() == 2
    assert c1.peek() == 2

test_peek_leaves_counters_unchanged()

# I'd write more tests than this, but this demonstrates the technique

If you're thinking that this is a bit of a clunky process to follow — having to write functions and then remember to call each one — you're right! That's why we'll soon learn about some tools built into Python's standard library that automate some of the things we're doing by hand here. But, as we've seen all quarter, you can do a lot with meager skills; even where we are now, we have a pretty reasonable way to isolate our the effect of our testing from everything else, so we should use what we know, then allow our gradually-improving knowledge to make things better, a little bit at a time. (I'm still doing that, after all these years; I still learn new techniques that improve my work, and I expect I will for as long as I do this kind of work.)

What to submit

Submit one Python script named problem1.py, which contains the implementation of your Locator class, along with your assert-based tests of that class.

Docstrings are not required, since we've already fully agreed on what problem we're solving here.

---

Problem 2 (1 point)

One of the things we established during our discussion about Classes is the need to be strategic in our design. When we write a class, we're designing a new type, and the choices we make will influence whether objects of that type will be easy to use, or whether it will be too easy to make mistakes with them that lead to program misbehavior or crashes. While there are many things that Python will allow us to do wherever we want, there's value in isolating those things in the "right place." We've already learned a few design recipes (i.e., basic rules we can follow so that our designs will be simpler, easier to grow into much larger programs, and so on). As we learn the design recipes in any programming language, it's worth considering the forces that shaped those recipes; in other words, we should want to know why those recipes say what they say. One way to understand why is to imagine that they said something different, then consider what the impact would be.

A design recipe we saw when we talked about Classes was the idea that we should initialize all of an object's attributes in the __init__ method of that object's class. In a sentence or two, explain why we made that decision by assuming that we made the opposite decision — attributes can be initialized anywhere within the class instead — and describing how that could lead to trouble.

What to submit

Submit one PDF file named problem2.pdf, which contains your answer to this question.

---

Problem 3 (4 points)

Suppose that we had an interface shape, which represents the concept of a geometric shape in two-dimensional space, such as a circle or a rectangle. Of course, there are substantial differences between the different shapes that we might want to represent in a system like this, but there are also substantial similarities. In particular, there are things that all shapes ought to be able to do, even if the way they do those things might be quite different in each case. If we design an interface for those similarities (i.e., if the way we ask different kinds of shapes to do these jobs is always the same, even if there differences in how the jobs are done), then we should be able to use shapes for many purposes interchangeably.

Let's agree that all shapes have the following common behavior:

• A shape knows what its centroid is, a point in two-dimensional space. Note that this isn't something your shape would need to calculate; it's information that your shape should store.
  • Centroid is a fancy term, but a simple concept: If you were to balance your shape on the head of a pin, the centroid would be the point where the pin would contact your shape so that the shape wouldn't fall off.
  • Assume that all points in two-dimensional spaces consist of an x- and y-coordinate, and that, like in algebra, x-coordinates grow as you move to the right, while y-coordinates grow as you move up.
• A shape can be moved, which means that its centroid is replaced with a new one.
• A shape can be asked what its area is, which is the total amount of space within the boundaries of the shape.
• A shape can be asked what its perimeter is, which is the distance around the shape. (By picking a starting point and following a path around the shape that leads back to where you started, how far did you go?).
• A shape can be asked whether it contains a particular point, which is to say that you can give it a point in two-dimensional space, and it can tell you whether that point lies somewhere within the area encompassed by that shape.

Your task here is do three things:

1. Design a Python interface that allows all of the common behavior above. All I mean by "design a Python interface" is to decide what the names of the methods would be, what parameters you'd need to pass to those methods, and so on. One important thing to realize, though, is that your design has to be one that allows all kinds of shapes to solve each problem, e.g., the parameters for a method in your interface have to be the same for all kinds of shapes. (This echoes a concept we talked a lot about in the Duck Typing and Interfaces lecture.)
2. Choose any three of the following shapes and implement a class for each. All of them should implement the interface you designed, and include whatever else they would need (e.g., attributes, additional public or private methods) that you think are necessary to be able to provide that interface.
   • Circle
   • Right triangle, oriented so that each of the two sides other than the hypotenuse run parallel to the x and y axes, respectively, and so that the hypotenuse runs from the top-rightmost point to the bottom-leftmost point within the shape, with the hypotenuse lying above the side parallel to the y-axis.
   • Rectangle, oriented so that it has two sides running parallel to the x axis and the other two sides running parallel to the y axis.
   • Regular pentagon, oriented so that its bottommost side runs parallel to the x axis.
   • Regular hexagon, oriented so that two of its sides run parallel to the x axis but none of its sides run parallel to the y axis.
3. Write a function that, given a sequence of shapes of any kind — including any that you've not yet implemented, but that you might later write that share the same interface — calculates the sum of their areas.

We're leaving you some design flexibility here (purposefully), so you'll be permitted to choose your own names for classes, methods, and so on. Use that flexibility wisely; that's not to say that we're fishing for whether you'll choose a particular name for something, but we're at least requiring that you choose a name that accurately reflects something's purpose.

Limitations

Your prior background in other programming languages (e.g., Java) might lead you to want to use a design technique called inheritance in this problem. While it's true that Python supports inheritance, its meaning isn't quite what you would expect if you've seen it in other programming languages, and we've not had a chance to discuss it yet, so inheritance is not permitted in your solution.

What to submit

Submit one Python module named problem3.py, which contains comments at the top that describe your interface, followed by your three classes, your function that calculates the sum of the areas of a shape, and assert-based tests for your classes and area-summing function. If you need variables in your assert-based tests, use the technique described in Problem 1 for isolating your tests from your module's global scope.

Docstrings are not required, but the names of your functions, methods, parameters, and attributes need to be well-chosen to reflect their purpose. If we can't tell that you've solved the problem we've assigned, we won't be able to give you credit for it.

---

Problem 4 (2 points)

When we discussed Duck Typing and Interfaces, one of the central themes was that of an interface. An interface is a concept whose power belies its simplicity; while interfaces are specified differently in Python than they do in many other programming languages, the idea is one that crosscuts programming languages. One way or another, all an interface really does is specify a set of characteristics a type needs to have in order to be compatible with some function, some class, or whatever.

In each of the following functions, there are (parts of) type annotations that have been specified opaquely as '???'. In each case, the variable or parameter they annotate have also been given intentionally non-descriptive names. For each function, propose a name for an interface (i.e., what you might replace '???' with) that describes each of the types that have been left unspecified, and specify what characteristics would be necessary to make a type compatible with that interface.

1. def render_scene(display: Display, x: list['???']) -> None:
       for obj in x:
           obj.draw(display)
   
   

2. def aggregate_length(x: list['???']) -> int:
       total = 0
   
       for obj in x:
           total += len(obj)
   
       return total
   
   

3. def find_largest_component(x: '???') -> int:
       max_size: int = 0
   
       y: '???' = x.all_components()
   
       while True:
           try:
               z: '???' = y.next_component()
   
               if z.size() > max_size:
                   max_size = z.size()
           except NoMoreComponentsError:
               break
   
       return max_size
   
   

What to submit

Submit one PDF file named problem4.pdf, which contains your answers to these questions.

---

Deliverables

In Canvas, you'll find a separate submission area for each problem. Submit your solution to each problem into the appropriate submission area. Be sure that you're submitting the correct file into the correct area (i.e., submitting your Problem 1 solution to the area for Problem 1, and so on). Under no circumstances will we offer credit for files submitted in the incorrect area.

Submit each file as-is, without putting it into a Zip file or arranging it in any other way. If we asked for a PDF, for example, all we want is a PDF; no more, no less. If you submit something other than what we asked for, we will not be offering you any credit on the submission. There are no exceptions to this rule.

Of course, you should also be aware that you're responsible for submitting precisely the version of your work that you want graded. We won't regrade an exercise simply because you submitted the wrong version accidentally, and we won't be able to offer any credit on exercises that you forgot to submit.

Can I submit after the deadline?

Unlike some of the projects in this course, the reinforcement exercises cannot be submitted after the deadline; there is no late policy for these. Each is worth only 3% of your grade, so it's not a disaster if you miss one of them along the way.