ska-referee-test/ref-test/common/data_tools.py

import os
import pathlib
from json import dump, loads
from datetime import datetime, timedelta
from glob import glob
from random import shuffle
from werkzeug.utils import secure_filename

from .security.database import decrypt_find_one

def check_data_folder_exists():
    from main import app
    if not os.path.exists(app.config['DATA_FILE_DIRECTORY']):
        pathlib.Path(app.config['DATA_FILE_DIRECTORY']).mkdir(parents='True', exist_ok='True')

def check_default_indicator():
    from main import app
    if not os.path.isfile(os.path.join(app.config['DATA_FILE_DIRECTORY'], '.default.txt')):
        open(os.path.join(app.config['DATA_FILE_DIRECTORY'], '.default.txt'),'w').close()

def get_default_dataset():
    check_default_indicator()
    from main import app
    default_file_path = os.path.join(app.config['DATA_FILE_DIRECTORY'], '.default.txt')
    with open(default_file_path, 'r') as default_file:
        default = default_file.read()
    return default

def available_datasets():
    from main import app
    files = glob(os.path.join(app.config["DATA_FILE_DIRECTORY"],'*.json'))
    default = get_default_dataset()
    output = []
    for file in files:
        filename = file.rsplit('/')[-1]
        label = f'{filename[:-5]} (Default)' if filename == default else filename[:-5]
        element = (filename, label)
        output.append(element)
    output.reverse()
    return output

def check_json_format(file):
    if not '.' in file.filename:
        return False
    if not file.filename.rsplit('.', 1)[-1] == 'json':
        return False
    return True

def validate_json_contents(file):
    file.stream.seek(0)
    data = loads(file.read())
    if not type(data) is dict:
        return False
    elif not all( key in data for key in ['meta', 'questions']):
        return False
    elif not type(data['meta']) is dict:
        return False
    elif not type(data['questions']) is list:
        return False
    return True

def store_data_file(file, default:bool=None):
    from admin.views import get_id_from_cookie
    from main import app
    check_default_indicator()
    timestamp = datetime.utcnow()
    filename = '.'.join([timestamp.strftime('%Y%m%d%H%M%S'),'json'])
    filename = secure_filename(filename)
    file_path = os.path.join(app.config['DATA_FILE_DIRECTORY'], filename)
    file.stream.seek(0)
    data = loads(file.read())
    data['meta']['timestamp'] = timestamp.strftime('%Y-%m-%d %H%M%S')
    data['meta']['author'] = get_id_from_cookie()
    data['meta']['tests'] = []
    with open(file_path, 'w') as _file:
        dump(data, _file, indent=2)
    if default:
        with open(os.path.join(app.config['DATA_FILE_DIRECTORY'], '.default.txt'), 'w') as _file:
            _file.write(filename)
    return filename

def randomise_list(list:list):
    _list = list.copy()
    shuffle(_list)
    return(_list)

def generate_questions(dataset:dict):
    questions_list = dataset['questions']
    output = []
    for block in randomise_list(questions_list):
        if block['type'] == 'question':
            question = {
                'type': 'question',
                'q_no': block['q_no'],
                'question_header': '',
                'text': block['text']
            }
            if block['q_type'] == 'Multiple Choice':
                question['options'] = randomise_list(block['options'])
            else:
                question['options'] = block['options'].copy()
            output.append(question)
        if block['type'] == 'block':
            for key, _question in enumerate(randomise_list(block['questions'])):
                question = {
                    'type': 'block',
                    'q_no': _question['q_no'],
                    'question_header': block['question_header'] if 'question_header' in block else '',
                    'block_length': len(block['questions']),
                    'block_q_no': key,
                    'text': _question['text']
                }
                if _question['q_type'] == 'Multiple Choice':
                    question['options'] = randomise_list(_question['options'])
                else:
                    question['options'] = _question['options'].copy()
                output.append(question)
    return output

def evaluate_answers(dataset: dict, answers: dict):
    score = 0
    max = 0
    tags = {}
    for block in dataset['questions']:
        if block['type'] == 'question':
            max += 1
            q_no = block['q_no']
            if str(q_no) in answers:
                correct = block['correct']
                correct_answer = block['options'][correct]
                submitted_answer = answers[str(q_no)]
                submitted_answer = submitted_answer.replace('‘', '‘').replace('’', '’').replace('—', '—')
                if submitted_answer == correct_answer:
                    score += 1
                    for tag in block['tags']:
                        if tag not in tags:
                            tags[tag] = {
                                'scored': 1,
                                'max': 1
                            }
                        else:
                            tags[tag]['scored'] += 1
                            tags[tag]['max'] += 1
                else:
                    for tag in block['tags']:
                        if tag not in tags:
                            tags[tag] = {
                                'scored': 0,
                                'max': 1
                            }
                        else:
                            tags[tag]['max'] += 1
        if block['type'] == 'block':
            for question in block['questions']:
                max += 1
                q_no = question['q_no']
                if str(q_no) in answers:
                    correct = question['correct']
                    correct_answer = question['options'][correct]
                    if answers[str(q_no)] == correct_answer:
                        score += 1
                        for tag in question['tags']:
                            if tag not in tags:
                                tags[tag] = {
                                    'scored': 1,
                                    'max': 1
                                }
                            else:
                                tags[tag]['scored'] += 1
                                tags[tag]['max'] += 1
                else:
                    for tag in question['tags']:
                        if tag not in tags:
                            tags[tag] = {
                                'scored': 0,
                                'max': 1
                            }
                        else:
                            tags[tag]['max'] += 1

    grade = 'merit' if score/max >= .85 else 'pass' if score/max >= .70 else 'fail'
    return {
        'grade': grade,
        'tags': tags,
        'score': score,
        'max': max
    }

def get_tags_list(dataset:dict):
    output = []
    blocks = dataset['questions']
    for block in blocks:
        if block['type'] == 'question':
            output = list(set(output) | set(block['tags']))
        if block['type'] == 'block':
            for question in block['questions']:
                output = list(set(output) | set(question['tags']))
    return output           

def get_time_options():
    time_options = [
        ('none', 'None'),
        ('60', '1 hour'),
        ('90', '1 hour 30 minutes'),
        ('120', '2 hours')
    ]
    return time_options

def get_datasets():
    from main import app, db
    files = glob(os.path.join(app.config["DATA_FILE_DIRECTORY"],'*.json'))
    data = []
    if files:
        for file in files:
            filename = file.rsplit('/')[-1]
            with open(file) as _file:
                load = loads(_file.read())
            _author = load['meta']['author']
            author = decrypt_find_one(db.users, {'_id': _author})['username']
            data_element = {
                'filename': filename,
                'timestamp': datetime.strptime(load['meta']['timestamp'], '%Y-%m-%d %H%M%S'),
                'author': author,
                'use': len(load['meta']['tests'])
            }
            data.append(data_element)
    return data

def get_correct_answers(dataset:dict):
    output = {}
    blocks = dataset['questions']
    for block in blocks:
        if block['type'] == 'question':
            output[str(block['q_no'])] = block['options'][block['correct']]
        if block['type'] == 'block':
            for question in block['questions']:
                output[str(question['q_no'])] = question['options'][question['correct']]
    return output