-- ============================================
-- PastPaper Master — COMP2211 problem-level taxonomy backfill
-- Seed Date: 2026-03-24
-- ============================================
--
-- Purpose:
-- 1. Backfill coarse taxonomy for COMP2211 question rows after the paper has been
--    processed into `paper_questions`.
-- 2. Use the audited cover-page problem mapping as the initial analytics baseline.
-- 3. Only fill empty taxonomy fields, so later fine-grained per-question curation
--    can safely overwrite these defaults.

WITH mapping AS (
  SELECT *
  FROM (
    VALUES
      ('COMP2211-2022-fall-midterm', '1', 'True/False Questions', 'True/False', 'True/False', ARRAY['True/False']::TEXT[], ARRAY['concept_check', 'rapid_reasoning']::TEXT[], 'true_false'),
      ('COMP2211-2022-fall-midterm', '2', 'Python Fundamentals', 'Python Fundamentals', 'Python Fundamentals', ARRAY['Python Fundamentals']::TEXT[], ARRAY['code_tracing', 'implementation', 'debugging']::TEXT[], 'coding'),
      ('COMP2211-2022-fall-midterm', '3', 'Conditional Probability and Bayes Classifier', 'Probabilistic Models', 'Probabilistic Models', ARRAY['Probabilistic Models']::TEXT[], ARRAY['manual_computation', 'probability_reasoning', 'classification_decision']::TEXT[], 'long_question'),
      ('COMP2211-2022-fall-midterm', '4', 'K-Nearest Neighbors', 'KNN and Clustering', 'KNN and Clustering', ARRAY['KNN and Clustering']::TEXT[], ARRAY['manual_computation', 'distance_calculation', 'algorithm_tracing']::TEXT[], 'long_question'),
      ('COMP2211-2022-fall-midterm', '5', 'K-Means Clustering', 'KNN and Clustering', 'KNN and Clustering', ARRAY['KNN and Clustering']::TEXT[], ARRAY['manual_computation', 'cluster_update', 'algorithm_tracing']::TEXT[], 'long_question'),
      ('COMP2211-2022-fall-midterm', '6', 'Perceptron', 'Perceptron and MLP', 'Perceptron and MLP', ARRAY['Perceptron and MLP']::TEXT[], ARRAY['manual_computation', 'weight_update', 'formula_application']::TEXT[], 'long_question'),
      ('COMP2211-2022-fall-midterm', '7', 'Multilayer Perceptron', 'Perceptron and MLP', 'Perceptron and MLP', ARRAY['Perceptron and MLP']::TEXT[], ARRAY['forward_pass', 'backpropagation', 'derivation']::TEXT[], 'long_question'),

      ('COMP2211-2022-spring-midterm', '1', 'True/False Questions', 'True/False', 'True/False', ARRAY['True/False']::TEXT[], ARRAY['concept_check', 'rapid_reasoning']::TEXT[], 'true_false'),
      ('COMP2211-2022-spring-midterm', '2', 'Python Fundamentals', 'Python Fundamentals', 'Python Fundamentals', ARRAY['Python Fundamentals']::TEXT[], ARRAY['code_tracing', 'implementation', 'debugging']::TEXT[], 'coding'),
      ('COMP2211-2022-spring-midterm', '3', 'Conditional Probability and Bayes Classifier', 'Probabilistic Models', 'Probabilistic Models', ARRAY['Probabilistic Models']::TEXT[], ARRAY['manual_computation', 'probability_reasoning', 'classification_decision']::TEXT[], 'long_question'),
      ('COMP2211-2022-spring-midterm', '4', 'K-Nearest Neighbors', 'KNN and Clustering', 'KNN and Clustering', ARRAY['KNN and Clustering']::TEXT[], ARRAY['manual_computation', 'distance_calculation', 'algorithm_tracing']::TEXT[], 'long_question'),
      ('COMP2211-2022-spring-midterm', '5', 'K-Means Clustering', 'KNN and Clustering', 'KNN and Clustering', ARRAY['KNN and Clustering']::TEXT[], ARRAY['manual_computation', 'cluster_update', 'algorithm_tracing']::TEXT[], 'long_question'),
      ('COMP2211-2022-spring-midterm', '6', 'Perceptron', 'Perceptron and MLP', 'Perceptron and MLP', ARRAY['Perceptron and MLP']::TEXT[], ARRAY['manual_computation', 'weight_update', 'formula_application']::TEXT[], 'long_question'),
      ('COMP2211-2022-spring-midterm', '7', 'Perceptron and Multilayer Perceptron', 'Perceptron and MLP', 'Perceptron and MLP', ARRAY['Perceptron and MLP']::TEXT[], ARRAY['forward_pass', 'backpropagation', 'weight_update']::TEXT[], 'long_question'),

      ('COMP2211-2022-spring-final-part-a', '1', 'True/False Questions', 'True/False', 'True/False', ARRAY['True/False']::TEXT[], ARRAY['concept_check', 'rapid_reasoning']::TEXT[], 'true_false'),
      ('COMP2211-2022-spring-final-part-a', '2', 'Na¨ıve Bayes and K-Nearest Neighbors', NULL, 'Probabilistic Models', ARRAY['Probabilistic Models', 'KNN and Clustering']::TEXT[], ARRAY['manual_computation', 'probability_reasoning', 'distance_calculation']::TEXT[], 'long_question'),
      ('COMP2211-2022-spring-final-part-a', '3', 'Multilayer Perceptron (MLP)', 'Perceptron and MLP', 'Perceptron and MLP', ARRAY['Perceptron and MLP']::TEXT[], ARRAY['forward_pass', 'backpropagation', 'derivation']::TEXT[], 'long_question'),
      ('COMP2211-2022-spring-final-part-a', '4', 'Digital Image Processing', 'Vision and CNN', 'Vision and CNN', ARRAY['Vision and CNN']::TEXT[], ARRAY['manual_computation', 'filter_computation', 'architecture_reasoning']::TEXT[], 'long_question'),

      ('COMP2211-2022-spring-final-part-b', '1', 'Convolutional Neural Network (CNN)', 'Vision and CNN', 'Vision and CNN', ARRAY['Vision and CNN']::TEXT[], ARRAY['forward_pass', 'architecture_reasoning', 'manual_computation']::TEXT[], 'long_question'),
      ('COMP2211-2022-spring-final-part-b', '2', 'Python Programming: Convolutional Neural Network', 'Python Fundamentals', 'Python Fundamentals', ARRAY['Python Fundamentals', 'Vision and CNN']::TEXT[], ARRAY['implementation', 'code_tracing', 'debugging']::TEXT[], 'coding'),
      ('COMP2211-2022-spring-final-part-b', '3', 'Minimax and Alpha-Beta Pruning', 'Search and Games', 'Search and Games', ARRAY['Search and Games']::TEXT[], ARRAY['tree_search', 'pruning', 'manual_tracing']::TEXT[], 'long_question'),
      ('COMP2211-2022-spring-final-part-b', '4', 'Ethics of Artificial Intelligence', 'Ethics of AI', 'Ethics of AI', ARRAY['Ethics of AI']::TEXT[], ARRAY['concept_explanation', 'argumentation', 'comparison']::TEXT[], 'short_answer'),

      ('COMP2211-2023-spring-midterm', '1', 'True/False Questions', 'True/False', 'True/False', ARRAY['True/False']::TEXT[], ARRAY['concept_check', 'rapid_reasoning']::TEXT[], 'true_false'),
      ('COMP2211-2023-spring-midterm', '2', 'Python Fundamentals', 'Python Fundamentals', 'Python Fundamentals', ARRAY['Python Fundamentals']::TEXT[], ARRAY['code_tracing', 'implementation', 'debugging']::TEXT[], 'coding'),
      ('COMP2211-2023-spring-midterm', '3', 'Na¨ıve Bayes Classifier', 'Probabilistic Models', 'Probabilistic Models', ARRAY['Probabilistic Models']::TEXT[], ARRAY['manual_computation', 'probability_reasoning', 'classification_decision']::TEXT[], 'long_question'),
      ('COMP2211-2023-spring-midterm', '4', 'K-Nearest Neighbors', 'KNN and Clustering', 'KNN and Clustering', ARRAY['KNN and Clustering']::TEXT[], ARRAY['manual_computation', 'distance_calculation', 'algorithm_tracing']::TEXT[], 'long_question'),
      ('COMP2211-2023-spring-midterm', '5', 'K-Means Clustering', 'KNN and Clustering', 'KNN and Clustering', ARRAY['KNN and Clustering']::TEXT[], ARRAY['manual_computation', 'cluster_update', 'algorithm_tracing']::TEXT[], 'long_question'),
      ('COMP2211-2023-spring-midterm', '6', 'Perceptron', 'Perceptron and MLP', 'Perceptron and MLP', ARRAY['Perceptron and MLP']::TEXT[], ARRAY['manual_computation', 'weight_update', 'formula_application']::TEXT[], 'long_question'),
      ('COMP2211-2023-spring-midterm', '7', 'Multilayer Perceptron', 'Perceptron and MLP', 'Perceptron and MLP', ARRAY['Perceptron and MLP']::TEXT[], ARRAY['forward_pass', 'backpropagation', 'derivation']::TEXT[], 'long_question'),

      ('COMP2211-2024-spring-midterm', '1', 'True/False Questions', 'True/False', 'True/False', ARRAY['True/False']::TEXT[], ARRAY['concept_check', 'rapid_reasoning']::TEXT[], 'true_false'),
      ('COMP2211-2024-spring-midterm', '2', 'Advanced Python for Artificial Intelligence', 'Python Fundamentals', 'Python Fundamentals', ARRAY['Python Fundamentals']::TEXT[], ARRAY['code_tracing', 'implementation', 'data_manipulation']::TEXT[], 'coding'),
      ('COMP2211-2024-spring-midterm', '3', 'Model Evaluation & Advanced Python Programming', 'Evaluation and Validation', 'Evaluation and Validation', ARRAY['Evaluation and Validation', 'Python Fundamentals']::TEXT[], ARRAY['metric_computation', 'experimental_design', 'implementation']::TEXT[], 'coding'),
      ('COMP2211-2024-spring-midterm', '4', 'Na¨ıve Bayes Classifier', 'Probabilistic Models', 'Probabilistic Models', ARRAY['Probabilistic Models']::TEXT[], ARRAY['manual_computation', 'probability_reasoning', 'classification_decision']::TEXT[], 'long_question'),
      ('COMP2211-2024-spring-midterm', '5', 'K-Nearest Neighbors', 'KNN and Clustering', 'KNN and Clustering', ARRAY['KNN and Clustering']::TEXT[], ARRAY['manual_computation', 'distance_calculation', 'algorithm_tracing']::TEXT[], 'long_question'),
      ('COMP2211-2024-spring-midterm', '6', 'Leader Clustering', 'KNN and Clustering', 'KNN and Clustering', ARRAY['KNN and Clustering']::TEXT[], ARRAY['manual_computation', 'cluster_update', 'algorithm_tracing']::TEXT[], 'long_question'),
      ('COMP2211-2024-spring-midterm', '7', 'D-fold Cross Validation', 'Evaluation and Validation', 'Evaluation and Validation', ARRAY['Evaluation and Validation']::TEXT[], ARRAY['metric_computation', 'experimental_design', 'reasoning']::TEXT[], 'long_question'),

      ('COMP2211-2024-spring-final', '1', 'True/False Questions', 'True/False', 'True/False', ARRAY['True/False']::TEXT[], ARRAY['concept_check', 'rapid_reasoning']::TEXT[], 'true_false'),
      ('COMP2211-2024-spring-final', '2', 'Advanced Python: Image Processing with NumPy', 'Python Fundamentals', 'Python Fundamentals', ARRAY['Python Fundamentals', 'Vision and CNN']::TEXT[], ARRAY['implementation', 'data_manipulation', 'filter_computation']::TEXT[], 'coding'),
      ('COMP2211-2024-spring-final', '3', 'Na¨ıve Bayes, K-Nearest Neighbors and Perceptron', NULL, 'Probabilistic Models', ARRAY['Probabilistic Models', 'KNN and Clustering', 'Perceptron and MLP']::TEXT[], ARRAY['manual_computation', 'probability_reasoning', 'distance_calculation', 'weight_update']::TEXT[], 'long_question'),
      ('COMP2211-2024-spring-final', '4', 'Multi-layer Perceptron', 'Perceptron and MLP', 'Perceptron and MLP', ARRAY['Perceptron and MLP']::TEXT[], ARRAY['forward_pass', 'backpropagation', 'derivation']::TEXT[], 'long_question'),
      ('COMP2211-2024-spring-final', '5', 'Digital Image Processing', 'Vision and CNN', 'Vision and CNN', ARRAY['Vision and CNN']::TEXT[], ARRAY['manual_computation', 'filter_computation', 'architecture_reasoning']::TEXT[], 'long_question'),
      ('COMP2211-2024-spring-final', '6', 'Dilated Convolution and Dropout', 'Vision and CNN', 'Vision and CNN', ARRAY['Vision and CNN']::TEXT[], ARRAY['architecture_reasoning', 'forward_pass', 'comparison']::TEXT[], 'long_question'),
      ('COMP2211-2024-spring-final', '7', 'Convolutional Neural Network', 'Vision and CNN', 'Vision and CNN', ARRAY['Vision and CNN']::TEXT[], ARRAY['architecture_reasoning', 'forward_pass', 'implementation']::TEXT[], 'long_question'),
      ('COMP2211-2024-spring-final', '8', 'Minimax and Alpha-Beta Pruning', 'Search and Games', 'Search and Games', ARRAY['Search and Games']::TEXT[], ARRAY['tree_search', 'pruning', 'manual_tracing']::TEXT[], 'long_question'),
      ('COMP2211-2024-spring-final', '9', 'Ethics of Artificial Intelligence', 'Ethics of AI', 'Ethics of AI', ARRAY['Ethics of AI']::TEXT[], ARRAY['concept_explanation', 'argumentation', 'comparison']::TEXT[], 'short_answer')
  ) AS t (
    source_exam_key,
    problem_number,
    raw_topic,
    analytics_topic,
    topic_primary,
    topic_tags,
    skill_tags,
    default_question_format
  )
)
UPDATE paper_questions AS q
SET analytics_topic = COALESCE(q.analytics_topic, mapping.analytics_topic),
    topic_primary = COALESCE(q.topic_primary, mapping.topic_primary),
    topic_tags = CASE
      WHEN q.topic_tags IS NULL OR cardinality(q.topic_tags) = 0 THEN mapping.topic_tags
      ELSE q.topic_tags
    END,
    skill_tags = CASE
      WHEN q.skill_tags IS NULL OR cardinality(q.skill_tags) = 0 THEN mapping.skill_tags
      ELSE q.skill_tags
    END,
    topics = CASE
      WHEN q.topics IS NULL OR cardinality(q.topics) = 0 THEN mapping.topic_tags
      ELSE q.topics
    END,
    question_format = CASE
      WHEN (q.question_format IS NULL OR q.question_format = '')
        AND mapping.default_question_format IS NOT NULL
      THEN mapping.default_question_format
      ELSE q.question_format
    END
FROM papers AS p
JOIN mapping
  ON mapping.source_exam_key = p.source_exam_key
WHERE q.paper_id = p.id
  AND p.source_kind = 'course_library'
  AND p.course_code = 'COMP2211'
  AND (
    q.question_number = mapping.problem_number
    OR q.question_number ~ ('^' || mapping.problem_number || '([^0-9].*)?$')
  );