questions.py
23.8 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
# python standard library
import asyncio
import random
import re
from os import path
import logging
from typing import Any, Dict, NewType
import uuid
# this project
from .tools import run_script, run_script_async
# setup logger for this module
logger = logging.getLogger(__name__)
QDict = NewType('QDict', Dict[str, Any])
class QuestionException(Exception):
pass
# ============================================================================
# Questions derived from Question are already instantiated and ready to be
# presented to students.
# ============================================================================
class Question(dict):
'''
Classes derived from this base class are meant to instantiate questions
for each student.
Instances can shuffle options or automatically generate questions.
'''
def __init__(self, q: QDict) -> None:
super().__init__(q)
# add required keys if missing
self.set_defaults(QDict({
'title': '',
'answer': None,
'comments': '',
'solution': '',
'files': {},
}))
def correct(self) -> None:
self['comments'] = ''
self['grade'] = 0.0
async def correct_async(self) -> None:
self.correct()
def set_defaults(self, d: QDict) -> None:
'Add k:v pairs from default dict d for nonexistent keys'
for k, v in d.items():
self.setdefault(k, v)
# ============================================================================
class QuestionRadio(Question):
'''An instance of QuestionRadio will always have the keys:
type (str)
text (str)
options (list of strings)
correct (list of floats)
discount (bool, default=True)
answer (None or an actual answer)
shuffle (bool, default=True)
choose (int) # only used if shuffle=True
'''
# ------------------------------------------------------------------------
def __init__(self, q: QDict) -> None:
super().__init__(q)
n = len(self['options'])
self.set_defaults(QDict({
'text': '',
'correct': 0,
'shuffle': True,
'discount': True,
'max_tries': (n + 3) // 4 # 1 try for each 4 options
}))
# check correct bounds and convert int to list,
# e.g. correct: 2 --> correct: [0,0,1,0,0]
if isinstance(self['correct'], int):
if not (0 <= self['correct'] < n):
msg = (f'Correct option not in range 0..{n-1} in '
f'"{self["ref"]}"')
raise QuestionException(msg)
self['correct'] = [1.0 if x == self['correct'] else 0.0
for x in range(n)]
elif isinstance(self['correct'], list):
# must match number of options
if len(self['correct']) != n:
msg = (f'Incompatible sizes: {n} options vs '
f'{len(self["correct"])} correct in "{self["ref"]}"')
raise QuestionException(msg)
# make sure is a list of floats
try:
self['correct'] = [float(x) for x in self['correct']]
except (ValueError, TypeError):
msg = (f'Correct list must contain numbers [0.0, 1.0] or '
f'booleans in "{self["ref"]}"')
raise QuestionException(msg)
# check grade boundaries
if self['discount'] and not all(0.0 <= x <= 1.0
for x in self['correct']):
msg = (f'Correct values must be in the interval [0.0, 1.0] in '
f'"{self["ref"]}"')
raise QuestionException(msg)
# at least one correct option
if all(x < 1.0 for x in self['correct']):
msg = (f'At least one correct option is required in '
f'"{self["ref"]}"')
raise QuestionException(msg)
# If shuffle==false, all options are shown as defined
# otherwise, select 1 correct and choose a few wrong ones
if self['shuffle']:
# lists with indices of right and wrong options
right = [i for i in range(n) if self['correct'][i] >= 1]
wrong = [i for i in range(n) if self['correct'][i] < 1]
self.set_defaults(QDict({'choose': 1+len(wrong)}))
# try to choose 1 correct option
if right:
r = random.choice(right)
options = [self['options'][r]]
correct = [self['correct'][r]]
else:
options = []
correct = []
# choose remaining wrong options
nwrong = self['choose'] - len(correct)
wrongsample = random.sample(wrong, k=nwrong)
options += [self['options'][i] for i in wrongsample]
correct += [self['correct'][i] for i in wrongsample]
# final shuffle of the options
perm = random.sample(range(self['choose']), k=self['choose'])
self['options'] = [str(options[i]) for i in perm]
self['correct'] = [correct[i] for i in perm]
# ------------------------------------------------------------------------
# can assign negative grades for wrong answers
def correct(self) -> None:
super().correct()
if self['answer'] is not None:
x = self['correct'][int(self['answer'])] # get grade of the answer
n = len(self['options'])
x_aver = sum(self['correct']) / n # expected value of grade
# note: there are no numerical errors when summing 1.0s so the
# x_aver can be exactly 1.0 if all options are right
if self['discount'] and x_aver != 1.0:
x = (x - x_aver) / (1.0 - x_aver)
self['grade'] = x
# ============================================================================
class QuestionCheckbox(Question):
'''An instance of QuestionCheckbox will always have the keys:
type (str)
text (str)
options (list of strings)
shuffle (bool, default True)
correct (list of floats)
discount (bool, default True)
choose (int)
answer (None or an actual answer)
'''
# ------------------------------------------------------------------------
def __init__(self, q: QDict) -> None:
super().__init__(q)
n = len(self['options'])
# set defaults if missing
self.set_defaults(QDict({
'text': '',
'correct': [1.0] * n, # Using 0.0 breaks (right, wrong) options
'shuffle': True,
'discount': True,
'choose': n, # number of options
'max_tries': max(1, min(n - 1, 3))
}))
# must be a list of numbers
if not isinstance(self['correct'], list):
msg = 'Correct must be a list of numbers or booleans'
raise QuestionException(msg)
# must match number of options
if len(self['correct']) != n:
msg = (f'Incompatible sizes: {n} options vs '
f'{len(self["correct"])} correct in "{self["ref"]}"')
raise QuestionException(msg)
# make sure is a list of floats
try:
self['correct'] = [float(x) for x in self['correct']]
except (ValueError, TypeError):
msg = (f'Correct list must contain numbers or '
f'booleans in "{self["ref"]}"')
raise QuestionException(msg)
# check grade boundaries
if self['discount'] and not all(0.0 <= x <= 1.0
for x in self['correct']):
msg0 = ('+-------------- BEHAVIOR CHANGE NOTICE --------------+')
msg1 = ('| Correct values must be in the interval [0.0, 1.0]. |')
msg2 = ('| I will convert to the new behavior, but you should |')
msg3 = ('| fix it in the question. |')
msg4 = ('+----------------------------------------------------+')
logger.warning(msg0)
logger.warning(msg1)
logger.warning(msg2)
logger.warning(msg3)
logger.warning(msg4)
logger.warning(f'-> please fix "{self["ref"]}"')
# normalize to [0,1]
self['correct'] = [(x+1)/2 for x in self['correct']]
# if an option is a list of (right, wrong), pick one
options = []
correct = []
for o, c in zip(self['options'], self['correct']):
if isinstance(o, list):
r = random.randint(0, 1)
o = o[r]
if r == 1:
# c = -c
c = 1.0 - c
options.append(str(o))
correct.append(c)
# generate random permutation, e.g. [2,1,4,0,3]
# and apply to `options` and `correct`
if self['shuffle']:
perm = random.sample(range(n), k=self['choose'])
self['options'] = [options[i] for i in perm]
self['correct'] = [correct[i] for i in perm]
else:
self['options'] = options[:self['choose']]
self['correct'] = correct[:self['choose']]
# ------------------------------------------------------------------------
# can return negative values for wrong answers
def correct(self) -> None:
super().correct()
if self['answer'] is not None:
x = 0.0
if self['discount']:
sum_abs = sum(abs(2*p-1) for p in self['correct'])
for i, p in enumerate(self['correct']):
x += 2*p-1 if str(i) in self['answer'] else 1-2*p
else:
sum_abs = sum(abs(p) for p in self['correct'])
for i, p in enumerate(self['correct']):
x += p if str(i) in self['answer'] else 0.0
try:
self['grade'] = x / sum_abs
except ZeroDivisionError:
self['grade'] = 1.0 # limit p->0
# ============================================================================
class QuestionText(Question):
'''An instance of QuestionText will always have the keys:
type (str)
text (str)
correct (list of str)
answer (None or an actual answer)
'''
# ------------------------------------------------------------------------
def __init__(self, q: QDict) -> None:
super().__init__(q)
self.set_defaults(QDict({
'text': '',
'correct': [], # no correct answers, always wrong
'transform': [], # transformations applied to the answer, in order
}))
# make sure its always a list of possible correct answers
if not isinstance(self['correct'], list):
self['correct'] = [str(self['correct'])]
else:
# make sure all elements of the list are strings
self['correct'] = [str(a) for a in self['correct']]
for f in self['transform']:
if f not in ('remove_space', 'trim', 'normalize_space', 'lower',
'upper'):
msg = (f'Unknown transform "{f}" in "{self["ref"]}"')
raise QuestionException(msg)
# check if answers are invariant with respect to the transforms
if any(c != self.transform(c) for c in self['correct']):
logger.warning(f'in "{self["ref"]}", correct answers are not '
'invariant wrt transformations => never correct')
# ------------------------------------------------------------------------
# apply optional filters to the answer
def transform(self, ans):
for f in self['transform']:
if f == 'remove_space': # removes all spaces
ans = ans.replace(' ', '')
elif f == 'trim': # removes spaces around
ans = ans.strip()
elif f == 'normalize_space': # replaces multiple spaces by one
ans = re.sub(r'\s+', ' ', ans.strip())
elif f == 'lower': # convert to lowercase
ans = ans.lower()
elif f == 'upper': # convert to uppercase
ans = ans.upper()
else:
logger.warning(f'in "{self["ref"]}", unknown transform "{f}"')
return ans
# ------------------------------------------------------------------------
def correct(self) -> None:
super().correct()
if self['answer'] is not None:
answer = self.transform(self['answer']) # apply transformations
self['grade'] = 1.0 if answer in self['correct'] else 0.0
# ============================================================================
class QuestionTextRegex(Question):
'''An instance of QuestionTextRegex will always have the keys:
type (str)
text (str)
correct (str or list[str])
answer (None or an actual answer)
The correct strings are python standard regular expressions.
Grade is 1.0 when the answer matches any of the regex in the list.
'''
# ------------------------------------------------------------------------
def __init__(self, q: QDict) -> None:
super().__init__(q)
self.set_defaults(QDict({
'text': '',
'correct': ['$.^'], # will always return false
}))
# make sure its always a list of regular expressions
if not isinstance(self['correct'], list):
self['correct'] = [self['correct']]
# converts patterns to compiled versions
try:
self['correct'] = [re.compile(a) for a in self['correct']]
except Exception:
msg = f'Failed to compile regex in "{self["ref"]}"'
raise QuestionException(msg)
# ------------------------------------------------------------------------
def correct(self) -> None:
super().correct()
if self['answer'] is not None:
self['grade'] = 0.0
for r in self['correct']:
try:
if r.match(self['answer']):
self['grade'] = 1.0
return
except TypeError:
logger.error(f'While matching regex {r.pattern} with '
f'answer "{self["answer"]}".')
# ============================================================================
class QuestionNumericInterval(Question):
'''An instance of QuestionTextNumeric will always have the keys:
type (str)
text (str)
correct (list [lower bound, upper bound])
answer (None or an actual answer)
An answer is correct if it's in the closed interval.
'''
# ------------------------------------------------------------------------
def __init__(self, q: QDict) -> None:
super().__init__(q)
self.set_defaults(QDict({
'text': '',
'correct': [1.0, -1.0], # will always return false
}))
# if only one number n is given, make an interval [n,n]
if isinstance(self['correct'], (int, float)):
self['correct'] = [float(self['correct']), float(self['correct'])]
# make sure its a list of two numbers
elif isinstance(self['correct'], list):
if len(self['correct']) != 2:
msg = (f'Numeric interval must be a list with two numbers, in '
f'{self["ref"]}')
raise QuestionException(msg)
try:
self['correct'] = [float(n) for n in self['correct']]
except Exception:
msg = (f'Numeric interval must be a list with two numbers, in '
f'{self["ref"]}')
raise QuestionException(msg)
# invalid
else:
msg = (f'Numeric interval must be a list with two numbers, in '
f'{self["ref"]}')
raise QuestionException(msg)
# ------------------------------------------------------------------------
def correct(self) -> None:
super().correct()
if self['answer'] is not None:
lower, upper = self['correct']
try: # replace , by . and convert to float
answer = float(self['answer'].replace(',', '.', 1))
except ValueError:
self['comments'] = ('A resposta tem de ser numérica, '
'por exemplo `12.345`.')
self['grade'] = 0.0
else:
self['grade'] = 1.0 if lower <= answer <= upper else 0.0
# ============================================================================
class QuestionTextArea(Question):
'''An instance of QuestionTextArea will always have the keys:
type (str)
text (str)
correct (str with script to run)
answer (None or an actual answer)
'''
# ------------------------------------------------------------------------
def __init__(self, q: QDict) -> None:
super().__init__(q)
self.set_defaults(QDict({
'text': '',
'timeout': 5, # seconds
'correct': '', # trying to execute this will fail => grade 0.0
'args': []
}))
self['correct'] = path.join(self['path'], self['correct'])
# ------------------------------------------------------------------------
def correct(self) -> None:
super().correct()
if self['answer'] is not None: # correct answer and parse yaml ouput
out = run_script(
script=self['correct'],
args=self['args'],
stdin=self['answer'],
timeout=self['timeout']
)
if out is None:
logger.warning(f'No grade after running "{self["correct"]}".')
self['grade'] = 0.0
elif isinstance(out, dict):
self['comments'] = out.get('comments', '')
try:
self['grade'] = float(out['grade'])
except ValueError:
logger.error(f'Output error in "{self["correct"]}".')
except KeyError:
logger.error(f'No grade in "{self["correct"]}".')
else:
try:
self['grade'] = float(out)
except (TypeError, ValueError):
logger.error(f'Invalid grade in "{self["correct"]}".')
# ------------------------------------------------------------------------
async def correct_async(self) -> None:
super().correct()
if self['answer'] is not None: # correct answer and parse yaml ouput
out = await run_script_async(
script=self['correct'],
args=self['args'],
stdin=self['answer'],
timeout=self['timeout']
)
if out is None:
logger.warning(f'No grade after running "{self["correct"]}".')
self['grade'] = 0.0
elif isinstance(out, dict):
self['comments'] = out.get('comments', '')
try:
self['grade'] = float(out['grade'])
except ValueError:
logger.error(f'Output error in "{self["correct"]}".')
except KeyError:
logger.error(f'No grade in "{self["correct"]}".')
else:
try:
self['grade'] = float(out)
except (TypeError, ValueError):
logger.error(f'Invalid grade in "{self["correct"]}".')
# ============================================================================
class QuestionInformation(Question):
# ------------------------------------------------------------------------
def __init__(self, q: QDict) -> None:
super().__init__(q)
self.set_defaults(QDict({
'text': '',
}))
# ------------------------------------------------------------------------
def correct(self) -> None:
super().correct()
self['grade'] = 1.0 # always "correct" but points should be zero!
# ============================================================================
#
# QFactory is a class that can generate question instances, e.g. by shuffling
# options, running a script to generate the question, etc.
#
# To generate an instance of a question we use the method generate().
# It returns a question instance of the correct class.
# There is also an asynchronous version called gen_async(). This version is
# synchronous for all question types (radio, checkbox, etc) except for
# generator types which run asynchronously.
#
# Example:
#
# # make a factory for a question
# qfactory = QFactory({
# 'type': 'radio',
# 'text': 'Choose one',
# 'options': ['a', 'b']
# })
#
# # generate synchronously
# question = qfactory.generate()
#
# # generate asynchronously
# question = await qfactory.gen_async()
#
# # answer one question and correct it
# question['answer'] = 42 # set answer
# question.correct() # correct answer
# grade = question['grade'] # get grade
#
# ============================================================================
class QFactory(object):
# Depending on the type of question, a different question class will be
# instantiated. All these classes derive from the base class `Question`.
_types = {
'radio': QuestionRadio,
'checkbox': QuestionCheckbox,
'text': QuestionText,
'text-regex': QuestionTextRegex,
'numeric-interval': QuestionNumericInterval,
'textarea': QuestionTextArea,
# -- informative panels --
'information': QuestionInformation,
'success': QuestionInformation,
'warning': QuestionInformation,
'alert': QuestionInformation,
}
def __init__(self, qdict: QDict = QDict({})) -> None:
self.question = qdict
# ------------------------------------------------------------------------
# generates a question instance of QuestionRadio, QuestionCheckbox, ...,
# which is a descendent of base class Question.
# ------------------------------------------------------------------------
async def gen_async(self) -> Question:
logger.debug(f'generating {self.question["ref"]}...')
# Shallow copy so that script generated questions will not replace
# the original generators
q = self.question.copy()
q['qid'] = str(uuid.uuid4()) # unique for each generated question
# If question is of generator type, an external program will be run
# which will print a valid question in yaml format to stdout. This
# output is then yaml parsed into a dictionary `q`.
if q['type'] == 'generator':
logger.debug(f' \\_ Running "{q["script"]}".')
q.setdefault('args', [])
q.setdefault('stdin', '')
script = path.join(q['path'], q['script'])
out = await run_script_async(script=script, args=q['args'],
stdin=q['stdin'])
q.update(out)
# Get class for this question type
try:
qclass = self._types[q['type']]
except KeyError:
logger.error(f'Invalid type "{q["type"]}" in "{q["ref"]}"')
raise
# Finally create an instance of Question()
try:
qinstance = qclass(QDict(q))
except QuestionException as e:
# logger.error(e)
raise e
return qinstance
# ------------------------------------------------------------------------
def generate(self) -> Question:
return asyncio.get_event_loop().run_until_complete(self.gen_async())