Assignment 1

C♯

Clone this repository and bring the code pieces you need into your BDSA Assignments GitHub repository.

Generics

Compare the following two methods:

int GreaterCount<T, U>(IEnumerable<T> items, T x)
    where T : IComparable<T>;

int GreaterCount<T, U>(IEnumerable<T> items, T x)
    where T : U
    where U : IComparable<U>;

Both methods returns the amount of elements in items which are greater than x.

Explain in your own words what the type constraints mean for both methods.

Optional

Implement and test the latter of the two methods including a type hierarchy which supports the given type constraints.

Iterators

Implement and test the following methods:

IEnumerable<T> Flatten<T>(IEnumerable<IEnumerable<T>> items);

IEnumerable<T> Filter<T>(IEnumerable<T> items, Predicate<T> predicate);

1. Flatten takes as argument a stream of a stream of T's. It should return a stream of T's.

2. Filter takes as arguments a stream of T's and a function which returns true or false when applied to an instance of T. It returns a stream of only the Ts where the predicate returns true.

Notes

You must yield elements and not use a temporary in-memory collection.

You can declare a Predicate likes this:

Predicate<int> even = Even;

bool Even(int i) => i % 2 == 0;

Regular Expressions

Implement and test the following methods:

IEnumerable<string> SplitLine(IEnumerable<string> lines);

IEnumerable<(int width, int height)> Resolutions(IEnumerable<string> resolutions);

IEnumerable<string> InnerText(string html, string tag);

IEnumerable<(Uri url, string title)> Urls(string html);

1. SplitLine takes as argument a stream of lines (strings) and returns a stream of the words on those lines (also strings). A 'word' is a non-empty contiguous sequence of the letters a–z or A–Z or the digits 0–9. Use a regular expression to split the lines into words.

2. Resolutions takes a string containing resolutions. A resolution could be 1920x1080. It returns a stream of resolutions as tuples, e.g. (1920, 1080). The solution must use named capture groups.

   Sample input (one line represents an element in the IEnumerable):

   1920x1080
   1024x768, 800x600, 640x480
   320x200, 320x240, 800x600
   1280x960

   Sample output:

   (1920, 1080)
   (1024, 768)
   (800, 600)
   (640, 480)
   (320, 200)
   (320, 240)
   (800, 600)
   (1280, 960)

3. InnerText takes as arguments a string containing HTML and a specific tag name. It returns the inner text of each of those tags. Use a regular expression with a back reference to match tags.

4. Urls takes as argument a string containing HTML and returns all urls with their titles if any, otherwise innerText.

Notes

You must yield elements and not use a temporary in-memory collection.

Given the following html and the tag a:

<div>
    <p>A <b>regular expression</b>, <b>regex</b> or <b>regexp</b> (sometimes called a <b>rational expression</b>) is, in <a href="https://en.wikipedia.org/wiki/Theoretical_computer_science" title="Theoretical computer science">theoretical computer science</a> and <a href="https://en.wikipedia.org/wiki/Formal_language" title="Formal language">formal language</a> theory, a sequence of <a href="https://en.wikipedia.org/wiki/Character_(computing)" title="Character (computing)">characters</a> that define a <i>search <a href="https://en.wikipedia.org/wiki/Pattern_matching" title="Pattern matching">pattern</a></i>. Usually this pattern is then used by <a href="https://en.wikipedia.org/wiki/String_searching_algorithm" title="String searching algorithm">string searching algorithms</a> for "find" or "find and replace" operations on <a href="https://en.wikipedia.org/wiki/String_(computer_science)" title="String (computer science)">strings</a>.</p>
</div>

The InnerText method should return:

• theoretical computer science
• formal language
• characters
• pattern
• string searching algorithms
• strings

You should support nested html tags such that given the following html and the tag p:

<div>
    <p>The phrase <i>regular expressions</i> (and consequently, regexes) is often used to mean the specific, standard textual syntax for representing <u>patterns</u> that matching <em>text</em> need to conform to.</p>
</div>

The InnerText method should return:

The phrase regular expressions (and consequently, regexes) is often used to mean the specific, standard textual syntax for representing patterns that matching text need to conform to.

---

Software Engineering

Let's say that the following is a natural language specification of a version control system:

I want a version control system that records changes to a file or set of files over time so that I can recall specific versions later. This system should work on any kind of files may they contain source code, configuration data, diagrams, binaries, etc.

I want to use such a system to be able to revert selected files back to a previous state, revert the entire project back to a previous state, to compare changes over time, to see who last modified something that might be causing a problem, who introduced an issue and when, etc.

Exercise 1

Use the noun/verb technique from "Objects First with Java: A Practical Introduction Using BlueJ" chapter 15, to analyze the given description.

1. Explain in which domain nouns and verbs that you identified are located.
2. The implementation in libgit2sharp does neither contain a class File nor a class State. Explain how that can be when libgit2sharp is an implementation of Git which is certainly a version control system as described above.

Exercise 2

In class we discussed so far the Coronapas App and Git as cases for software systems.

1. Categorize each of the two systems into Sommervilles types of applications. Note, the systems may not fall cleanly into a single category.
2. Argue for why you choose certain categories for each system.

Exercise 3

Sommerville describes that there are two kinds of software products. Describe for the Coronapas App and Git what kind of software product they are and provide arguments for your believes.

Exercise 4

Sommerville discusses quality of professional software, non-functional quality attributes, or product characteristics. Compare the Coronapas App, Git, and the Insulin pump control system (see SE chap 1.3.1) with respect to the quality attributes dependability, security, efficiency, and maintainability.

Do they all share the same characteristics with regards to these quality attributes or are these of varying importance to the three systems? Give examples for each of the three systems with regards to each if the quality attributes above.

Exercise 5

Inspect the implementations Gitlet and Git a bit more thoroughly than in class.

1. Explain why is there likely no architecture for Gitlet.
2. How could you infer the architecture of Git that was depicted in class without any more documentation, i.e., only the available source code?
3. Gitlet has a particular design that mimics the architecture of Git but that is implemented differently. Can you describe it in words?
4. Git and Gitlet are designed with respect to different quality attributes (product characteristics). Name some of the most prominent quality attributes that influence the design of each of the two systems.

Exercise 6

Look at the following two cases of issues with health care software systems:

• "Softwareproblemer skadede mere end 100 patienter på amerikansk hospital"
• "Kodefejl i Sundhedsplatformen: Fem patienter har fået forkert dosis medicin"

1. Based on the articles, describe the reasons that caused the respective issues.
2. Describe potential solutions to the problem.
3. Compare your solutions to the proposed solutions in the articles, which one would you as a software engineer recommend? Argue for your recommendations.
4. Discuss ethical dilemas in case you were developing either of these health care systems.

---

Submitting the assignment

To submit the assignment you need to create a PDF document using LaTeX that contains the answers to the questions and a link to a public GitHub repository that contains a fork of the assignments repository with the completed code.

The PDF file must conform to the following naming convention: group_<x>_<id1>_<id2>_<id3>_assignment_01.pdf, where <x> is replaced by the number of your group from README_GROUPS.md and <id1>, <id2>, and <id3> are your respective ITU identifiers.

You submit via LearnIT.