38 — Quiz — questions from a file

A quiz is a loop over data: each round shows one question, reads an answer, and adds to a score. Guide 14 saved a record; this guide keeps the questions in a JSON file so changing the quiz means editing data, not code. The menu loop is the one from guide 26, aimed at options instead of directions.

#Steps

1. Create questions.json — a list of question dicts. Each dict has the question, the options, and the answer as the position of the correct option, counting from 0:

   [
     {"question": "What is 5 + 3?", "options": ["6", "7", "8", "9"], "answer": 2},
     {"question": "What color is the sky?", "options": ["Green", "Blue", "Red", "Yellow"], "answer": 1},
     {"question": "How many legs does a dog have?", "options": ["2", "3", "4", "5"], "answer": 2}
   ]
   

   copyrun it →

   "answer": 2 means the third option, "8". Keeping the file data-only means a new question is just one new line in JSON.

2. Load the questions with json_load from guide 14. The file holds a list, so json_load returns a list of dicts:

   use file latest
   questions = json_load("questions.json")
   show f"Loaded {len(questions)} questions"
   

   copyrun it →

   Run nme r quiz; it prints Loaded 3 questions.

3. A for q in questions: loop visits each dict, and enumerate numbers the options for display. ask reads the answer, and the if compares it with the stored position:

   use file latest
   questions = json_load("questions.json")
   score = 0
   for q in questions:
       show q["question"]
       for i, option in enumerate(q["options"]):
           show str(i + 1) + ". " + option
       ask answer, "Your answer (number): "
       if answer == str(q["answer"] + 1):
           score = score + 1
   

   copyrun it →

   The answer is stored counting from 0, but the player counts from 1, so the stored answer needs a + 1 on both sides of the comparison.

4. The full game wraps that loop in while play_again == "yes": so a finished quiz can start over, and counts the missed questions. Save it as quiz.nme:

   # A multiple-choice quiz: the questions come from a JSON file.
   # Run: nme r quiz
   # The file questions.json must exist in the same folder.
   
   use file latest
   
   show "Welcome to the NME quiz!"
   ask name, "Your name: "
   questions = json_load("questions.json")
   
   play_again = "yes"
   while play_again == "yes":
       score = 0
       number = 0
       missed = []
   
       for q in questions:
           number = number + 1
           show ""
           show f"Question {number} of {len(questions)}: {q['question']}"
           for i, option in enumerate(q["options"]):
               show f"  {i + 1}. {option}"
           ask answer, "Your answer (number): "
           if answer == str(q["answer"] + 1):
               show "Correct!"
               score = score + 1
           else:
               show "Wrong. The correct answer was " + str(q["answer"] + 1) + "."
               missed.append(number)
   
       percent = score * 100 // len(questions)
       show ""
       show f"{name}: {score} of {len(questions)} correct ({percent}%)."
       if missed:
           show "Missed questions: " + str(missed)
       if percent == 100:
           show "Perfect score!"
       elif percent >= 70:
           show "Well done!"
       elif percent >= 50:
           show "Not bad. Keep practicing!"
       else:
           show "Try again - read each question carefully."
   
       ask play_again, "Play again? (yes/no): "
   
   show "Thanks for playing, " + name + "!"
   

   copyrun it →

5. Run it. The piped input plays one whole round — a name, then 1 for each question, then no:

   printf 'Mina\n1\n1\n1\nno\n' | nme r quiz
   

   copy

   Welcome to the NME quiz!
   Your name: 
   Question 1 of 3: What is 5 + 3?
     1. 6
     2. 7
     3. 8
     4. 9
   Your answer (number): Wrong. The correct answer was 3.
   
   Question 2 of 3: What color is the sky?
     1. Green
     2. Blue
     3. Red
     4. Yellow
   Your answer (number): Wrong. The correct answer was 2.
   
   Question 3 of 3: How many legs does a dog have?
     1. 2
     2. 3
     3. 4
     4. 5
   Your answer (number): Wrong. The correct answer was 3.
   
   Mina: 0 of 3 correct (0%).
   Missed questions: [1, 2, 3]
   Try again - read each question carefully.
   Play again? (yes/no): Thanks for playing, Mina!
   

   copy

   All three 1 answers are wrong here, so the score stays 0 and the missed list shows every question. Answering 3, 2, 3 would give a perfect score.

6. The Korean twin quiz.ko.nme writes the same game with 파일 사용 최신, json읽기, 물어봐, and 말해. The answers are numbers, so the same piped input works in both languages.

#Try it yourself

Change questions.json to hold four questions of your own, and make answer point at the right option for each. Then rerun quiz.nme and answer with the correct numbers; the score should reach 100%.

#What you learned

• A JSON list of dicts is a whole question bank; json_load returns it as a list.
• enumerate(q["options"]) numbers the options for display.
• The stored answer counts from 0; the player's choice counts from 1.
• while play_again == "yes": lets the player replay the whole quiz.