views.py
24.3 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
import operator
import time
from collections import defaultdict
from functools import reduce
from itertools import chain, product
import simplejson
from django.db.models import Q
from django.http import JsonResponse, QueryDict
from django.shortcuts import render
from django.template.context_processors import csrf
from django.utils.translation import gettext as _
from crispy_forms.utils import render_crispy_form
from connections.models import Entry, Subentry, ArgumentConnection, RealisationDescription
from syntax.models import NaturalLanguageDescription, Schema
from semantics.models import Frame
from common.decorators import ajax_required
from .forms import (
EntryForm,
SchemaFormFactory,
FrameFormFactory,
PositionFormFactory,
PhraseAttributesFormFactory,
LexFormFactory,
LemmaFormFactory,
ArgumentFormFactory,
PredefinedPreferenceFormFactory,
RelationalPreferenceFormFactory,
SynsetPreferenceFormFactory,
)
from .polish_strings import SCHEMA_OPINION, FRAME_OPINION, EXAMPLE_SOURCE, EXAMPLE_OPINION, RELATION
from .phrase_descriptions.descriptions import position_prop_description
MAX_LAST_VISITED = 10
# TODO remove!!!
#def test(request):
# return render(request, 'test.html', {})
def entries(request):
# TODO make this automatic by subclassing/configuring session object
if 'last_visited' not in request.session:
request.session['last_visited'] = []
return render(request, 'entries.html', { 'entries_form' : EntryForm() })
FORM_TYPES = {
#'entry-main' : MainEntryForm,
'entry' : EntryForm,
}
FORM_FACTORY_TYPES = {
'schema' : SchemaFormFactory,
'position' : PositionFormFactory,
'phrase_lex' : LexFormFactory,
'lemma' : LemmaFormFactory,
'frame' : FrameFormFactory,
'argument' : ArgumentFormFactory,
'predefined' : PredefinedPreferenceFormFactory,
'relational' : RelationalPreferenceFormFactory,
'synset' : SynsetPreferenceFormFactory,
}
def make_form(form_type, data=None, unique_number=None):
#print('======== MAKE FORM', form_type)
if form_type in FORM_FACTORY_TYPES:
return FORM_FACTORY_TYPES[form_type].get_form(data=data, unique_number=unique_number)
if form_type in FORM_TYPES:
return FORM_TYPES[form_type](data=data)
elif form_type.startswith('phrase_'):
phrase_type = form_type[7:]
#print('================ MAKE PHRASE FORM', phrase_type, unique_number)
return PhraseAttributesFormFactory.get_form(phrase_type, data=data, unique_number=unique_number)
return None
@ajax_required
def get_subform(request):
if request.method == 'GET':
ctx = {}
ctx.update(csrf(request))
form_type = request.GET['subform_type']
form = make_form(form_type)
try:
form_html = render_crispy_form(form, context=ctx)
except:
print('******************', form_type)
raise
return JsonResponse({'form_html' : form_html})
#TODO clean this code bordello up
def filter_objects(objects, queries, tab=''):
#print(tab + '===================================================================')
for query in queries:
#print(tab + '***', query)
objects = objects.filter(query).distinct()
#print(tab + '---------------------------------------------------------------')
#print(tab, objects)
#print('\n')
#print(tab + '===================================================================')
return objects.distinct()
def collect_forms(forms_json, errors_collector_dict, tab=' '):
data = simplejson.loads(forms_json)
form_type = data['formtype']
form_number = data.get('formnumber', 0)
if form_type in ('or', 'other'):
return form_type
else:
#print(tab, 'FORM:', data['form'])
#print(tab, 'TYPE:', form_type, 'NUMBER:', form_number)
#print(tab, 'DATA:', data)
query_params = QueryDict(data['form'])
#print(tab, 'PARAMS:', query_params)
form = make_form(form_type, data=query_params, unique_number=form_number)
#print(tab, 'FORM TYPE:', type(form))
if not form.is_valid():
errors_collector_dict.update(form.errors)
#print(tab, '{} CHILDREN GROUP(S)'.format(len(data['children'])))
# a form may have one or more children forms, organised into and-or
# (e.g. an entry form has child schema forms, frame forms etc.)
subform_groups = []
for subforms_json in data['children']:
subform_group = simplejson.loads(subforms_json)
subform_type, subforms = subform_group['formtype'], subform_group['subforms']
children = [[]]
conjunctions = set()
for child in subforms:
child_form = collect_forms(child, errors_collector_dict, tab + ' ')
if child_form in ('or', 'other'):
children.append([])
conjunctions.add(child_form)
else:
children[-1].append(child_form)
assert(len(conjunctions) <= 1)
conjunction = 'or' if not conjunctions else conjunctions.pop()
subforms = list(filter(None, children))
if subforms:
subform_groups.append((subform_type, conjunction, subforms))
return (form, data['negated'], subform_groups)
def reduce_ids_and(ids):
# sum the negated ids
ids[False] = reduce(lambda x, y: x.union(y), ids[False], set())
if ids[True]:
# intersect the non-negated and subtract the sum of negated
return (True, reduce(lambda x, y: x.intersection(y), ids[True]) - ids[False])
else:
# negate the sum of negated
return (False, ids[False])
def other_operator(tab, form, children_group, parent_objects):
name, conjunction, children = children_group
print(tab, '==============', name, conjunction)
print(tab, 'PARENT OBJECTS:', parent_objects)
prefixes = set()
# matches[id][j] -> set of child ids of ‹id› parent satisfying ‹j›th specification
matches = defaultdict(lambda: defaultdict(set))
for i, conj_children in enumerate(children):
print(tab, ' ', conjunction)
child_ids_and = { True: [], False: [] }
for child in conj_children:
child_form = child[0]
child_objects = get_filtered_objects(child, tab=tab + ' ')
prefix = form.get_child_form_prefix(child_form)
prefixes.add(prefix)
child_ids = set(co.id for co in child_objects)
child_ids_and[not child_form.is_negated()].append(child_ids)
child_ids_and = reduce_ids_and(child_ids_and)
print(tab, ' ===>', '[]'.format(i), child_ids_and[0], len(child_ids_and[1]), sorted(child_ids_and[1])[:10])
# TODO enable negations?
assert(child_ids_and[0] == True)
for child_id in child_ids_and[1]:
assert(prefix.endswith('__in'))
for parent in parent_objects.filter(Q((prefix, [child_id]))):
matches[parent.id][i].add(child_id)
assert(len(prefixes) == 1)
N = len(children)
matching_parent_ids = set()
for parent_id, mtchs in matches.items():
print(tab, parent_id, mtchs)
if len(mtchs) < len(children):
print(tab, 'not all matched')
continue
for x in product(*mtchs.values()):
if len(x) == len(set(x)):
# found N different children objects, each satisfying different specification
matching_parent_ids.add(parent_id)
print(tab, 'MATCH:', x)
continue
if parent_id not in matching_parent_ids:
print(tab, 'no match')
return parent_objects.filter(id__in=matching_parent_ids)
def get_filtered_objects(forms, initial_objects=None, tab=' '):
form, negated_attrs, children = forms
objects = form.model_class.objects.all() if initial_objects is None else initial_objects.all()
queries = form.get_queries(negated_attrs)
#print(tab, type(form), 'FOR FILTERING:', form.model_class)
#print(tab, queries)
objects = filter_objects(objects, queries, tab=tab)
#print(tab, 'OK')
for children_group in children:
if children_group[1] == 'other':
objects = other_operator(tab, form, children_group, objects)
continue
#print(tab, 'CHILD FORMS')
object_ids_or = []
prefixes = set()
for or_children in children_group[2]:
objects_and = form.model_class.objects.all() if initial_objects is None else initial_objects.all()
for child in or_children:
child_form = child[0]
child_objects = get_filtered_objects(child, tab=tab + ' ')
prefix = form.get_child_form_prefix(child_form)
prefixes.add(prefix)
child_ids = [co.id for co in child_objects]
q = Q((prefix, child_ids))
if child_form.is_negated():
objects_and = objects_and.exclude(q)
else:
objects_and = objects_and.filter(q)
object_ids_or.append({o.id for o in objects_and})
assert(len(prefixes) == 1)
object_ids = reduce(operator.or_, object_ids_or)
objects = objects.filter(id__in=object_ids)
objects = objects.distinct()
#print(tab, 'FILTERED:', form.model_class)
return objects
# forms – an ‘or’ list of ‘and’ lists of forms, the forms are flattened and treated as one ‘or’ list.
# The function is used for filtering out schemata/frames. E.g. if the user chooses entries with a schema
# safisfying X AND a schema satisfying Y, schemata satisfying X OR Y should be displayed (and all other
# schemata should be hidden).
def get_filtered_objects2(objects, forms):
print(forms)
filtered_ids = [{ schema.id for schema in get_filtered_objects(form, initial_objects=objects) } for form in chain.from_iterable(forms)]
filtered_ids = reduce(operator.or_, filtered_ids)
return objects.filter(id__in=filtered_ids)
@ajax_required
def send_form(request):
if request.method == 'POST':
errors_dict = dict()
forms = collect_forms(request.POST['forms[]'], errors_dict)
if errors_dict:
return JsonResponse({ 'success' : 0, 'errors' : errors_dict })
del request.session['forms']
else:
request.session['forms'] = request.POST['forms[]']
return JsonResponse({ 'success' : 1 })
return JsonResponse({})
def get_scroller_params(POST_data):
order = (int(POST_data['order[0][column]']), POST_data['order[0][dir]'])
return {
'draw' : int(POST_data['draw']),
'start' : int(POST_data['start']),
'length' : int(POST_data['length']),
'order' : order,
'filter' : POST_data['search[value]']
}
@ajax_required
def get_entries(request):
if request.method == 'POST':
errors_dict = dict()
forms = collect_forms(request.session['forms'], errors_dict)
# form should already be validated if it passed through send_form
assert(not errors_dict)
scroller_params = get_scroller_params(request.POST)
entries = get_filtered_objects(forms).filter(import_error=False)
total = entries.count()
if scroller_params['filter']:
entries = entries.filter(name__startswith=scroller_params['filter'])
filtered = entries.count()
i, j = scroller_params['start'], scroller_params['start'] + scroller_params['length']
order_field, order_dir = scroller_params['order']
if order_field != 0 or order_dir != 'asc':
# ordering other than lemma ascending (database default)
order_field = 'name' if order_field == 0 else 'status__key'
if order_dir == 'desc':
order_field = '-' + order_field
entries = entries.order_by(order_field)
entries_list = list(entries.values('id', 'name', 'status__key'))
result = {
'draw' : scroller_params['draw'],
'recordsTotal': total,
'recordsFiltered': filtered,
'data': entries_list[i:j],
}
return JsonResponse(result)
return JsonResponse({})
def subentry2str(subentry):
ret = subentry.entry.name
if subentry.inherent_sie.name == 'true':
ret += ' się'
elems = []
if subentry.aspect:
elems.append(subentry.aspect.name)
if subentry.negativity:
elems.append(subentry.negativity.name)
if elems:
ret += ' ({})'.format(', '.join(elems))
if subentry.predicativity.name == 'true':
ret += ' pred.'
return ret
def position_prop2dict(prop):
return {
'str' : prop.name,
'desc' : position_prop_description(prop.name),
} if prop else {
'str' : '',
'desc' : '',
}
def get_phrase_desc(phrase, position, negativity, lang):
return NaturalLanguageDescription.objects.get(
phrase_str=phrase.text_rep,
function=position.function,
control=position.control,
pred_control=position.pred_control,
negativity=negativity,
lang=lang).description
def schema2dict(schema, negativity, lang):
return {
'opinion' : SCHEMA_OPINION()[schema.opinion.key],
'opinion_key' : schema.opinion.key,
'id' : str(schema.id),
'positions' : [
{
'func' : position_prop2dict(p.function),
'control' : position_prop2dict(p.control),
'p_control' : position_prop2dict(p.pred_control),
'id' : '{}-{}'.format(schema.id, p.id),
'phrases' : [
{
'str' : str(pt),
'id' : '{}-{}-{}'.format(schema.id, p.id, pt.id),
'desc' : get_phrase_desc(pt, p, negativity, lang),
} for pt in p.phrase_types.all()
],
} for p in schema.positions.all()
],
}
def get_rel_pref_desc(pref):
relation = pref.relation.key
if relation == 'RELAT':
desc = _('Realizacja tego argumentu w zdaniu powinna być powiązana jakąkolwiek relacją')
else:
desc = _('Realizacja tego argumentu w zdaniu powinna być powiązana relacją <i>{}</i>').format(RELATION()[relation])
return desc + ' ' + _('z realizacją argumentu <i>{}</i>.').format(pref.to)
def make_ul(items):
return '<ul>{}</ul>'.format(''.join(map('<li>{}</li>'.format, items)))
def get_synset_def(synset):
ret = []
if synset.definition:
ret.append(_('definicja:') + make_ul([synset.definition]))
#ret.append(_('jednostki leksykalne: ') + ', '.join(map(str, synset.lexical_units.all())))
hypernyms = list(synset.hypernyms.all())
if hypernyms:
ret.append(_('hiperonimy:') + make_ul(map(str, hypernyms)))
return ' '.join(ret)
def get_prefs_list(argument):
return sorted(
({
'str' : str(p),
} for p in argument.predefined.all()),
key=lambda x: x['str']
) + sorted(
({
'str' : str(s),
# can be a new synset
'url' : None if s.id < 0 else 'http://plwordnet21.clarin-pl.eu/synset/{}'.format(s.id),
'info' : s.lexical_units.all()[0].gloss if s.id < 0 else get_synset_def(s),
} for s in argument.synsets.all()),
key=lambda x: x['str']
) + sorted(
({
'str' : str(r),
'info' : get_rel_pref_desc(r),
} for r in argument.relations.all()),
key=lambda x: x['str']
)
def frame2dict(frame, entry_meanings):
return {
'opinion' : FRAME_OPINION()[frame.opinion.key],
'opinion_key' : frame.opinion.key,
'id' : str(frame.id),
'lexical_units' : [
{
'str' : lu.text_rep,
'id' : lu.id,
'entry_meaning' : lu in entry_meanings,
'definition' : lu.definition,
'gloss' : lu.gloss,
'url' : None if lu.luid is None else 'http://plwordnet21.clarin-pl.eu/lemma/{}/{}'.format(lu.base, lu.sense)
} for lu in frame.lexical_units.all()
],
'arguments' : [
{
'str' : str(a),
'id' : '{}-{}'.format(frame.id, a.id),
'role' : '{}{}'.format(a.role.role.role.lower(), ' ' + a.role.attribute.attribute.lower() if a.role.attribute else ''),
'preferences' : get_prefs_list(a),
} for a in sorted(frame.arguments.all(), key=lambda a: a.role.role.priority + (a.role.attribute.priority if a.role.attribute else 2))
],
}
# returns two dicts:
# (1) {
# frame_id_1 : {
# schema_id_1 : { [alt_1*, ..., alt_l] },
# schema_id_k : {...}
# }
# ...
# frame_id_n : {...}
# }
# *alternation is a dict: {
# key: extended argument id (frame_id-arg_id)
# val: list of extended phrase ids (schema_id-position_id-phr_id)
# (2) {
# extended_arg_id_1 : {
# alt_1 : {
# extended_phr_id_1 : psedo_natural_language_phrase_1_1_1,
# extended_phr_id_l : psedo_natural_language_phrase_1_1_l,
# }
# ...
# alt_k : {...}
# }
# ...
# extended_arg_idn : {...}
# }
def get_alternations(schemata, frames):
# TODO czy alternacja może być podpięta do całej pozycji, bez konkretnej frazy?
alternations = defaultdict(lambda: defaultdict(lambda: defaultdict(lambda: defaultdict(list))))
phrases = defaultdict(lambda: defaultdict(dict))
realisation_descriptions = defaultdict(lambda: defaultdict(lambda: defaultdict(dict)))
for schema in schemata:
for hook in schema.schema_hooks.all():
arg_conns = hook.argument_connections.all()
assert (len(arg_conns) < 2)
if (arg_conns):
argument = arg_conns[0].argument
frame = argument.frame
if frame not in frames:
continue
phr_id = '{}-{}-{}'.format(schema.id, hook.position.id, hook.phrase_type.id)
arg_id = '{}-{}'.format(frame.id, argument.id)
alternations[frame.id][schema.id][hook.alternation][arg_id].append(phr_id)
phrases[arg_id][hook.alternation - 1][phr_id] = hook.description
alt_dict = defaultdict(lambda: defaultdict(list))
for frame_id, frame_schema_alternations in alternations.items():
for schema_id, schema_alternations in frame_schema_alternations.items():
for alt_no in sorted(schema_alternations.keys()):
alt_dict[frame_id][schema_id].append(schema_alternations[alt_no])
realisation_descriptions[frame_id][schema_id][alt_no - 1] = RealisationDescription.objects.get(frame__id=frame_id, schema__id=schema_id, alternation=alt_no).description
return alt_dict, phrases, realisation_descriptions
def get_examples(entry):
examples = []
for example in entry.examples.all():
frame_ids, argument_ids, lu_ids, schema_ids, phrases, phrases_syntax, positions = set(), set(), set(), set(), set(), set(), set()
for connection in example.example_connections.all():
for argument in connection.arguments.all():
frame_ids.add(argument.frame.id)
argument_ids.add('{}-{}'.format(argument.frame.id, argument.id))
if connection.lexical_unit:
lu_ids.add(connection.lexical_unit.id)
for hook in connection.schema_connections.all():
schema_ids.add(hook.schema.id);
phrases.add('{}-{}-{}-{}'.format(hook.schema.id, hook.position.id, hook.phrase_type.id, hook.alternation - 1))
phrases_syntax.add('{}-{}-{}'.format(hook.schema.id, hook.position.id, hook.phrase_type.id))
positions.add('{}-{}'.format(hook.schema.id, hook.position.id))
examples.append({
'id' : str(example.id),
'sentence' : example.sentence,
'source' : EXAMPLE_SOURCE()[example.source.key],
'opinion' : EXAMPLE_OPINION()[example.opinion.key],
'note' : example.note,
'frame_ids' : sorted(frame_ids),
'argument_ids' : sorted(argument_ids),
'lu_ids' : sorted(lu_ids),
'schema_ids' : sorted(schema_ids),
'phrases' : sorted(phrases),
'phrases_syntax' : sorted(phrases_syntax),
'positions' : sorted(positions),
})
return sorted(examples, key=lambda x: x['sentence'])
@ajax_required
def get_entry(request):
if request.method == 'POST':
form = EntryForm(request.POST)
eid = request.POST['entry']
if eid.isdigit() and form.is_valid():
eid = int(eid)
# TODO check that Entry has no import errors
entry = Entry.objects.get(id=eid)
errors_dict = dict()
entry_form, _, children_forms = collect_forms(request.POST['forms[]'], errors_dict)
# form should already be validated if it passed through send_form
assert(not errors_dict)
filter_schemata, filter_frames = entry_form.cleaned_data['filter_schemata'], entry_form.cleaned_data['filter_frames']
if filter_schemata:
schema_forms = [frms[2] for frms in children_forms if frms[0] == 'schema']
assert (len(schema_forms) <= 1)
if schema_forms:
schema_forms = schema_forms[0]
else:
filter_schemata = False
if filter_frames:
frame_forms = [frms[2] for frms in children_forms if frms[0] == 'frame']
assert (len(frame_forms) <= 1)
if frame_forms:
frame_forms = frame_forms[0]
else:
filter_frames = False
subentries = []
all_schema_objects = []
for subentry in entry.subentries.all():
schemata = []
schema_objects = subentry.schemata.all()
if filter_schemata:
schema_objects = get_filtered_objects2(schema_objects, schema_forms)
for schema in schema_objects:
schemata.append(schema2dict(schema, subentry.negativity, request.LANGUAGE_CODE))
if schemata:
all_schema_objects += list(schema_objects)
subentries.append({ 'str' : subentry2str(subentry), 'schemata' : schemata })
frame_objects = Frame.objects.filter(arguments__argument_connections__schema_connections__subentry__entry=entry).distinct()
if filter_frames:
frame_objects = get_filtered_objects2(frame_objects, frame_forms)
frames = [frame2dict(frame, entry.lexical_units.all()) for frame in frame_objects]
alternations, realisation_phrases, realisation_descriptions = get_alternations(all_schema_objects, frame_objects)
examples = get_examples(entry)
if 'last_visited' not in request.session:
request.session['last_visited'] = []
if 'show_reals_desc' not in request.session:
request.session['show_reals_desc'] = True
# https://docs.djangoproject.com/en/2.2/topics/http/sessions/#when-sessions-are-saved
if [entry.name, entry.id] in request.session['last_visited']:
request.session['last_visited'].remove([entry.name, entry.id])
request.session['last_visited'].insert(0, (entry.name, entry.id))
request.session['last_visited'] = request.session['last_visited'][:(MAX_LAST_VISITED + 1)]
request.session.modified = True
return JsonResponse({ 'subentries' : subentries, 'frames' : frames, 'alternations' : alternations, 'realisation_phrases' : realisation_phrases, 'realisation_descriptions' : realisation_descriptions, 'examples' : examples })
return JsonResponse({})
@ajax_required
def change_show_reals_desc(request):
if request.method == 'POST':
val = simplejson.loads(request.POST['val'])
request.session['show_reals_desc'] = val
return JsonResponse({ 'success' : 1 })
return JsonResponse({})