def get_form(self, form_class):
"""
Returns an instance of the form to be used in this view.
"""
self.model_name = self.kwargs["model_name"]
model = _get_model_from_string(self.model_name)
return form_class(model, **self.get_form_kwargs())
def get_context_data(self, **kwargs):
"""Add nodes and links to the context"""
context = super(NgramPOSView, self).get_context_data(**kwargs)
model = _get_model_from_string(self.kwargs["model_name"])
ct = ContentType.objects.get_for_model(model, for_concrete_model=False)
rules_dict = self._parse_rules()
relevant_ids = set(
TaggedCollection.objects.filter(content_type=ct, is_relevant=True).values_list("object_id", flat=True)
)
irrelevant_ids = set(
TaggedCollection.objects.filter(content_type=ct, is_relevant=False).values_list("object_id", flat=True)
)
all_tags = set()
correct_data = defaultdict(lambda: 0)
incorrect_data = defaultdict(lambda: 0)
unjudged_data = defaultdict(lambda: 0)
for obj in model.objects.all():
max_pos_tag = " ".join(max(obj.pos_tag, key=lambda x: x[1])[0])
tag = str(scores.compress_pos_tag(max_pos_tag, rules_dict))
all_tags.add(tag)
if obj.id in relevant_ids:
correct_data[tag] += 1
elif obj.id in irrelevant_ids:
incorrect_data[tag] += 1
else:
unjudged_data[tag] += 1
all_tags = sorted(all_tags)
context["categories"] = all_tags
context["top_relevant_data"] = [
(tag, correct_data[tag], incorrect_data[tag], unjudged_data[tag])
for tag in all_tags
if int(correct_data[tag] / (incorrect_data[tag] + 0.1)) > 10
]
context["top_irrelevant_data"] = [
(tag, correct_data[tag], incorrect_data[tag], unjudged_data[tag])
for tag in all_tags
if int(incorrect_data[tag] / (correct_data[tag] + 0.1)) > 10
]
context["correct_data"] = [correct_data[tag] for tag in all_tags]
context["incorrect_data"] = [incorrect_data[tag] for tag in all_tags]
context["unjudged_data"] = [unjudged_data[tag] for tag in all_tags]
# Add regex groups to populate forms
context["regex_groups"] = self.regex_groups
return context
def form_valid(self, form):
"""
clear cache
"""
attribute = form.cleaned_data["attr"]
model = _get_model_from_string(self.model_name)
for obj in model.objects.all():
fields = obj.extra_fields
if attribute in fields:
del fields[attribute]
obj.extra_fields = fields
obj.save()
next_url = self.request.GET.get("next", reverse("testcollection_model", args=[self.model_name]))
return HttpResponseRedirect(next_url)
def get_context_data(self, **kwargs):
context = super(AttributeFilterView, self).get_context_data(**kwargs)
if not self.queryset:
try:
model = _get_model_from_string(self.kwargs[self.model_name])
self.queryset = model.objects.all()
except:
raise ImproperlyConfigured(
"AttributeFilterView requires either a definition of "
"'queryset' or a 'model_name' kwarg attribute")
fields = getattr(self.queryset.model, self.model_fields_attr)
form = self._FilterForm(fields)(self.request.POST or None)
if form.is_valid():
filter_values = dict([(field, value) for field, value in form.cleaned_data.iteritems()])
self.queryset = self.queryset.filter(**filter_values)
context[self.context_form_name] = form
return context
def get_context_data(self, **kwargs):
"""Add data to context"""
context = super(ConceptIndexStats, self).get_context_data(**kwargs)
model = _get_model_from_string(self.kwargs["model_name"])
global_word_set = cache.get(WORDS_SET)
counts = defaultdict(lambda: 0)
for word in global_word_set:
counts[len(cache.get(CONCEPT_PREFIX + word))] += 1
context["histogram_data"] = str(counts.items()).replace("(", "[").replace(")", "]")
context["word_count"] = len(global_word_set)
context["concept_count"] = model.objects.count()
word_counts = defaultdict(lambda: 0)
for collocation in model.objects.values_list("ngram", flat=True):
word_counts[len(collocation.split())] += 1
context["col_word_len_hist"] = str(word_counts.items()).replace("(", "[").replace(")", "]")
return context
