Problem Solving — CS & DS

Assignment 2 — Trace, Test, Fix (Block II)

Due: week 14 · Weight: see assessment plan · Covers LO4, LO5, LO6 · Submit code + report.

This assignment is about evidence: tracing what code really does, designing tests that could fail, and fixing defects at the mechanism. Python is required; every claim must be backed by a run you can reproduce.

The scenario

The registry stores student IDs as 8-character strings: exactly 8 characters, first character a letter, last character a digit. A junior assistant wrote this checker:

def check(sid):
    if len(sid) == 8:
        if sid[0].isalpha() and sid[-1].isdigit():
            return "ok"
    return "bad"

Registry staff report: "Sometimes valid IDs are called bad, and it never explains why anything is bad."

Task 1 — Trace and explain (5 marks)

For each input, give the returned value and the reason in one sentence:

(a) "CS101" (b) "cs101abc" (that is 8 characters) (c) "1CS0 1AB9" (d) "CS0 1AB9" (contains a space)

Task 2 — Test table from requirements (6 marks)

Design a test table of at least 12 rows from the requirements (not from reading the code first — write the table before looking again at check). Columns: input, expected verdict, why this row exists. Mark the three rows you predict will fail against the given code, with one line of predicted mechanism each.

Task 3 — Better contract (4 marks)

The staff want the verdict to name the first failed rule, in the order length, first character, last character. Write the pseudocode (or Python) for verdict(sid) with that contract, and state what it returns for the empty string.

Task 4 — Fix and prove (5 marks)

(a) Implement verdict in Python. (2) (b) Run your task-2 table against both check and verdict; include the full output of both runs (12+ rows each). (2) (c) In three sentences: which rows distinguished the two implementations, and why a suite in which every row passes proves less than one with a deliberate failure? (1)

Submission checklist