def test_new_search(self):
"""Tests that GET /collectionbackups/new_search returns the search parameters for searching the collection backups resource."""
query_builder = SQLAQueryBuilder('CollectionBackup')
response = self.app.get(url('/collectionbackups/new_search'), headers=self.json_headers,
extra_environ=self.extra_environ_view)
resp = json.loads(response.body)
assert resp['search_parameters'] == h.get_search_parameters(query_builder)
def edit(self, id):
"""GET /formsearches/id/edit: Return the data necessary to update an existing
OLD form search.
"""
"""Return a form search and the data needed to update it.
:URL: ``GET /formsearches/edit`` with optional query string parameters
:param str id: the ``id`` value of the form search that will be updated.
:returns: a dictionary of the form::
{"form_search": {...}, "data": {...}}
where the value of the ``form_search`` key is a dictionary
representation of the form search and the value of the ``data`` key
is a dictionary containing the data necessary to update a form
search.
"""
form_search = h.eagerload_form_search(
Session.query(FormSearch)).get(id)
if form_search:
data = {
'search_parameters':
h.get_search_parameters(self.query_builder)
}
return {'data': data, 'form_search': form_search}
else:
response.status_int = 404
return {'error': 'There is no form search with id %s' % id}
def edit(self, id):
"""GET /formsearches/id/edit: Return the data necessary to update an existing
OLD form search.
"""
"""Return a form search and the data needed to update it.
:URL: ``GET /formsearches/edit`` with optional query string parameters
:param str id: the ``id`` value of the form search that will be updated.
:returns: a dictionary of the form::
{"form_search": {...}, "data": {...}}
where the value of the ``form_search`` key is a dictionary
representation of the form search and the value of the ``data`` key
is a dictionary containing the data necessary to update a form
search.
"""
form_search = h.eagerload_form_search(Session.query(FormSearch)).get(id)
if form_search:
data = {'search_parameters': h.get_search_parameters(self.query_builder)}
return {'data': data, 'form_search': form_search}
else:
response.status_int = 404
return {'error': 'There is no form search with id %s' % id}
def test_new_search(self):
"""Tests that GET /languages/new_search returns the search parameters for searching the languages resource."""
query_builder = SQLAQueryBuilder('Language')
response = self.app.get(url('/languages/new_search'), headers=self.json_headers,
extra_environ=self.extra_environ_view)
resp = json.loads(response.body)
assert resp['search_parameters'] == h.get_search_parameters(query_builder)
def new_search(self):
"""Return the data necessary to search the phonology resources.
:URL: ``GET /phonologies/new_search``
:returns: ``{"search_parameters": {"attributes": { ... }, "relations": { ... }}``
"""
return {'search_parameters': h.get_search_parameters(self.query_builder)}
def new_search(self):
"""Return the data necessary to search the morpheme language model
resources.
:URL: ``GET /morphemelanguagemodels/new_search``
:returns: ``{"search_parameters": {"attributes": { ... }, "relations": { ... }}``
"""
return {'search_parameters': h.get_search_parameters(self.query_builder)}
def new(self):
"""GET /formsearches/new: Return the data necessary to create a new OLD
form search.
"""
"""Return the data necessary to create a new form search.
:URL: ``GET /formsearches/new`` with optional query string parameters
:returns: A dictionary of lists of resources
"""
return {'search_parameters': h.get_search_parameters(self.query_builder)}
def new_search(self):
"""Return the data necessary to search the morpheme language model
resources.
:URL: ``GET /morphemelanguagemodels/new_search``
:returns: ``{"search_parameters": {"attributes": { ... }, "relations": { ... }}``
"""
return {
'search_parameters': h.get_search_parameters(self.query_builder)
}
def new(self):
"""GET /formsearches/new: Return the data necessary to create a new OLD
form search.
"""
"""Return the data necessary to create a new form search.
:URL: ``GET /formsearches/new`` with optional query string parameters
:returns: A dictionary of lists of resources
"""
return {
'search_parameters': h.get_search_parameters(self.query_builder)
}
def new_search_corpora(self):
"""Return the data necessary to search across corpus resources.
:URL: ``GET /corpora/new_search_corpora``
:returns: ``{"search_parameters": {"attributes": { ... }, "relations": { ... }}``
.. note::
Contrast this action with `new_search`, which returns the data
needed to search across the forms of a corpus.
"""
return {'search_parameters':
h.get_search_parameters(self.query_builder_for_ordering)}
def new_search_corpora(self):
"""Return the data necessary to search across corpus resources.
:URL: ``GET /corpora/new_search_corpora``
:returns: ``{"search_parameters": {"attributes": { ... }, "relations": { ... }}``
.. note::
Contrast this action with `new_search`, which returns the data
needed to search across the forms of a corpus.
"""
return {
'search_parameters':
h.get_search_parameters(self.query_builder_for_ordering)
}
def test_search(self):
"""Tests that corpora search works correctly.
"""
# Create a corpus defined by ``content`` that contains all sentences
# with five or more words.
# Get ids of all sentences with more than 5 words.
long_sentences = Session.query(model.Form).\
filter(and_(
model.Form.syntactic_category.has(model.SyntacticCategory.name==u'S'),
model.Form.transcription.op('regexp')(u'^([^ ]+ ){5}[^ ]+'))).all()
long_sentence = long_sentences[0]
len_long_sentences = len(long_sentences)
long_sentence_ids = [f.id for f in long_sentences]
long_sentences = u','.join(map(str, long_sentence_ids))
# Restrict one of the forms that will be in the corpus.
restricted_tag = h.get_restricted_tag()
long_sentence.tags.append(restricted_tag)
Session.add(long_sentence)
Session.commit()
# Create the corpus
name = u'Sentences with 6 or more words.'
params = self.corpus_create_params.copy()
params.update({
'name': name,
'content': long_sentences
})
params = json.dumps(params)
original_corpus_count = Session.query(Corpus).count()
response = self.app.post(url('corpora'), params, self.json_headers,
self.extra_environ_admin)
resp = json.loads(response.body)
corpus_id = resp['id']
new_corpus_count = Session.query(Corpus).count()
corpus = Session.query(Corpus).get(corpus_id)
corpus_dir = os.path.join(self.corpora_path, 'corpus_%d' % corpus_id)
corpus_dir_contents = os.listdir(corpus_dir)
assert new_corpus_count == original_corpus_count + 1
assert resp['name'] == name
assert corpus_dir_contents == []
assert response.content_type == 'application/json'
assert resp['content'] == long_sentences
# The ``forms`` attribute is a collection, no repeats, that's why the following is true:
assert len(corpus.forms) == len_long_sentences
# Search the corpus for forms beginning in vowels.
query = json.dumps({"query": {"filter": ['Form', 'transcription', 'regex', '^[AEIOUaeiou]']},
"paginator": {'page': 1, 'items_per_page': 10}})
response = self.app.post(url('/corpora/%d/search' % corpus_id), query,
self.json_headers, self.extra_environ_admin)
resp = json.loads(response.body)
matches = resp['items']
assert not set([f['id'] for f in matches]) - set(long_sentence_ids)
assert not filter(
lambda f: f['transcription'][0].lower() not in ['a', 'e', 'i', 'o', 'u'], matches)
assert not filter(lambda f: len(f['transcription'].split(' ')) < 6, matches)
# Vacuous search of the corpus returns everything.
query = json.dumps({"query": {"filter": ['Form', 'transcription', 'like', '%']}})
response = self.app.post(url('/corpora/%d/search' % corpus_id), query,
self.json_headers, self.extra_environ_admin)
resp = json.loads(response.body)
assert set([f['id'] for f in resp]) == set(long_sentence_ids)
# Vacuous search as the viewer returns everything that is not restricted.
query = json.dumps({"query": {"filter": ['Form', 'transcription', 'like', '%']}})
response = self.app.post(url('/corpora/%d/search' % corpus_id), query,
self.json_headers, self.extra_environ_view)
resp2 = json.loads(response.body)
# Viewer will get 1 or 2 forms fewer (2 are restricted, 1 assuredly a long sentence.)
assert len(resp) > len(resp2)
# Failed search with an invalid corpus id
query = json.dumps({"query": {"filter": ['Form', 'transcription', 'like', '%']}})
response = self.app.post(url('/corpora/123456789/search'), query,
self.json_headers, self.extra_environ_admin, status=404)
resp = json.loads(response.body)
assert resp['error'] == u'There is no corpus with id 123456789'
# Failed search with an invalid query
query = json.dumps({"query": {"filter": ['Form', 'thingamafracasicle', 'like', '%']}})
response = self.app.post(url('/corpora/%d/search' % corpus_id), query,
self.json_headers, self.extra_environ_admin, status=400)
resp = json.loads(response.body)
assert resp['errors']['Form.thingamafracasicle'] == 'There is no attribute thingamafracasicle of Form'
# Request GET /corpora/new_search
response = self.app.get(url(controller='corpora', action='new_search'),
headers=self.json_headers, extra_environ=self.extra_environ_admin)
resp = json.loads(response.body)
assert resp == {'search_parameters': h.get_search_parameters(SQLAQueryBuilder('Form'))}
def test_search(self):
"""Tests that corpora search works correctly.
"""
# Create a corpus defined by ``content`` that contains all sentences
# with five or more words.
# Get ids of all sentences with more than 5 words.
long_sentences = Session.query(model.Form).\
filter(and_(
model.Form.syntactic_category.has(model.SyntacticCategory.name==u'S'),
model.Form.transcription.op('regexp')(u'^([^ ]+ ){5}[^ ]+'))).all()
long_sentence = long_sentences[0]
len_long_sentences = len(long_sentences)
long_sentence_ids = [f.id for f in long_sentences]
long_sentences = u','.join(map(str, long_sentence_ids))
# Restrict one of the forms that will be in the corpus.
restricted_tag = h.get_restricted_tag()
long_sentence.tags.append(restricted_tag)
Session.add(long_sentence)
Session.commit()
# Create the corpus
name = u'Sentences with 6 or more words.'
params = self.corpus_create_params.copy()
params.update({'name': name, 'content': long_sentences})
params = json.dumps(params)
original_corpus_count = Session.query(Corpus).count()
response = self.app.post(url('corpora'), params, self.json_headers,
self.extra_environ_admin)
resp = json.loads(response.body)
corpus_id = resp['id']
new_corpus_count = Session.query(Corpus).count()
corpus = Session.query(Corpus).get(corpus_id)
corpus_dir = os.path.join(self.corpora_path, 'corpus_%d' % corpus_id)
corpus_dir_contents = os.listdir(corpus_dir)
assert new_corpus_count == original_corpus_count + 1
assert resp['name'] == name
assert corpus_dir_contents == []
assert response.content_type == 'application/json'
assert resp['content'] == long_sentences
# The ``forms`` attribute is a collection, no repeats, that's why the following is true:
assert len(corpus.forms) == len_long_sentences
# Search the corpus for forms beginning in vowels.
query = json.dumps({
"query": {
"filter": ['Form', 'transcription', 'regex', '^[AEIOUaeiou]']
},
"paginator": {
'page': 1,
'items_per_page': 10
}
})
response = self.app.post(url('/corpora/%d/search' % corpus_id), query,
self.json_headers, self.extra_environ_admin)
resp = json.loads(response.body)
matches = resp['items']
assert not set([f['id'] for f in matches]) - set(long_sentence_ids)
assert not filter(
lambda f: f['transcription'][0].lower() not in
['a', 'e', 'i', 'o', 'u'], matches)
assert not filter(lambda f: len(f['transcription'].split(' ')) < 6,
matches)
# Vacuous search of the corpus returns everything.
query = json.dumps(
{"query": {
"filter": ['Form', 'transcription', 'like', '%']
}})
response = self.app.post(url('/corpora/%d/search' % corpus_id), query,
self.json_headers, self.extra_environ_admin)
resp = json.loads(response.body)
assert set([f['id'] for f in resp]) == set(long_sentence_ids)
# Vacuous search as the viewer returns everything that is not restricted.
query = json.dumps(
{"query": {
"filter": ['Form', 'transcription', 'like', '%']
}})
response = self.app.post(url('/corpora/%d/search' % corpus_id), query,
self.json_headers, self.extra_environ_view)
resp2 = json.loads(response.body)
# Viewer will get 1 or 2 forms fewer (2 are restricted, 1 assuredly a long sentence.)
assert len(resp) > len(resp2)
# Failed search with an invalid corpus id
query = json.dumps(
{"query": {
"filter": ['Form', 'transcription', 'like', '%']
}})
response = self.app.post(url('/corpora/123456789/search'),
query,
self.json_headers,
self.extra_environ_admin,
status=404)
resp = json.loads(response.body)
assert resp['error'] == u'There is no corpus with id 123456789'
# Failed search with an invalid query
query = json.dumps(
{"query": {
"filter": ['Form', 'thingamafracasicle', 'like', '%']
}})
response = self.app.post(url('/corpora/%d/search' % corpus_id),
query,
self.json_headers,
self.extra_environ_admin,
status=400)
resp = json.loads(response.body)
assert resp['errors'][
'Form.thingamafracasicle'] == 'There is no attribute thingamafracasicle of Form'
# Request GET /corpora/new_search
response = self.app.get(url(controller='corpora', action='new_search'),
headers=self.json_headers,
extra_environ=self.extra_environ_admin)
resp = json.loads(response.body)
assert resp == {
'search_parameters':
h.get_search_parameters(SQLAQueryBuilder('Form'))
}