def test_citation__empty_dataset(self):
citation = self.citation
self.citation_metadata.title = "Obama occurrences as Republican"
# Actual results
citation.cite(select_statement=self.select_statement,
citation_metadata=self.citation_metadata,
citation_timestamp=self.citation_timestamp)
actual_result = citation.result_set_utils.description + "\n" + citation.citation_metadata.citation_snippet
self.citation = citation
# Expected results
query_utils = QueryUtils(query=self.select_statement,
citation_timestamp=self.citation_timestamp)
citation_snippet = generate_citation_snippet(
identifier=query_utils.pid,
creator=self.citation_metadata.creator,
title=self.citation_metadata.title,
publisher=self.citation_metadata.publisher,
pulication_year=self.citation_metadata.publication_year,
resource_type=self.citation_metadata.resource_type)
expected_result = "This is an empty dataset. We cannot infer any description from it.\n" + citation_snippet
# Test object
test = Test(
test_number=1,
tc_desc=
'Test if an empty dataset can be cited and a citation snippet is returned.',
expected_result=expected_result,
actual_result=actual_result)
return test
def test_citation__recite_semantically_identical(self):
citation1 = self.citation
citation2 = ct.Citation(self.test_config.get('RDFSTORE', 'get'),
self.test_config.get('RDFSTORE', 'post'))
self.citation_metadata.title = "Obama occurrences, new mention"
# Actual results
# Citation1
citation1.cite(select_statement=self.select_statement,
citation_metadata=self.citation_metadata,
citation_timestamp=self.citation_timestamp)
# Citation2
vie_tz = timezone(timedelta(hours=2))
citation_timestamp2 = datetime.now(vie_tz)
select_statement2 = open(
"test_data/test_citation__recite_semantically_identical2.txt",
"r").read()
citation2.cite(select_statement=select_statement2,
citation_metadata=self.citation_metadata,
citation_timestamp=citation_timestamp2)
# Concat Citation1 + Citation2 into actual results
actual_result = citation1.citation_metadata.citation_snippet \
+ "\n" + citation2.citation_metadata.citation_snippet
self.citation = citation1
# Expected results
# Citation1
query_utils = QueryUtils(query=self.select_statement,
citation_timestamp=self.citation_timestamp)
citation_snippet1 = generate_citation_snippet(
identifier=query_utils.pid,
creator=self.citation_metadata.creator,
title=self.citation_metadata.title,
publisher=self.citation_metadata.publisher,
pulication_year=self.citation_metadata.publication_year,
resource_type=self.citation_metadata.resource_type)
expected_result = citation_snippet1 + "\n" + citation_snippet1
test = Test(
test_number=5,
tc_desc=
'Test if citing a query which is semantically equivalent to an existing query in the '
'query store will return the existing citation snippet. The test needs to pass only '
'for covered normalization measures where two queries would be transformed to the same '
'normal query and thus recognized as semantically identical.',
expected_result=expected_result,
actual_result=actual_result)
return test
def test_citation__recite_unchanged_dataset(self):
citation1 = self.citation
citation2 = ct.Citation(self.test_config.get('RDFSTORE', 'get'),
self.test_config.get('RDFSTORE', 'post'))
self.citation_metadata.title = "Obama occurrences, new mention"
# Actual results
# Citation1
citation1.cite(select_statement=self.select_statement,
citation_metadata=self.citation_metadata,
citation_timestamp=self.citation_timestamp)
# Citation2
vie_tz = timezone(timedelta(hours=2))
citation_timestamp2 = datetime.now(vie_tz)
citation2.cite(select_statement=self.select_statement,
citation_metadata=self.citation_metadata,
citation_timestamp=citation_timestamp2)
# Concat Citation1 + Citation2 into actual results
actual_result = citation1.citation_metadata.citation_snippet \
+ "\n" + citation2.citation_metadata.citation_snippet
self.citation = citation1
# Expected results
# Citation1
query_utils = QueryUtils(query=self.select_statement,
citation_timestamp=self.citation_timestamp)
citation_snippet1 = generate_citation_snippet(
identifier=query_utils.pid,
creator=self.citation_metadata.creator,
title=self.citation_metadata.title,
publisher=self.citation_metadata.publisher,
pulication_year=self.citation_metadata.publication_year,
resource_type=self.citation_metadata.resource_type)
expected_result = citation_snippet1 + "\n" + citation_snippet1
test = Test(
test_number=4,
tc_desc=
'Test if reciting a query where the dataset has not changed since the last citation '
'returns the citation snippet as of last citation where the query was either new '
'or the dataset changed.',
expected_result=expected_result,
actual_result=actual_result)
return test
def test_citation__non_empty_dataset(self):
citation = self.citation
self.citation_metadata.title = "Obama occurrences"
# Actual results
citation.cite(select_statement=self.select_statement,
citation_metadata=self.citation_metadata,
citation_timestamp=self.citation_timestamp)
actual_result = citation.result_set_utils.description + "\n" + citation.citation_metadata.citation_snippet
self.citation = citation
# Expected results
query_utils = QueryUtils(query=self.select_statement,
citation_timestamp=self.citation_timestamp)
citation_snippet = generate_citation_snippet(
identifier=query_utils.pid,
creator=self.citation_metadata.creator,
title=self.citation_metadata.title,
publisher=self.citation_metadata.publisher,
pulication_year=self.citation_metadata.publication_year,
resource_type=self.citation_metadata.resource_type)
result_set = self.rdf_engine.get_data(self.select_statement,
self.citation_timestamp)
dataset_utils = RDFDataSetUtils(dataset=result_set)
dataset_description = dataset_utils.describe()
expected_result = dataset_description + "\n" + citation_snippet
# Test object
test = Test(
test_number=2,
tc_desc=
'Test if a non-empty dataset can be cited and a citation snippet is returned.',
expected_result=expected_result,
actual_result=actual_result)
return test
def test_citation__aggregated_dataset(self):
citation1 = self.citation
citation2 = ct.Citation(self.test_config.get('RDFSTORE', 'get'),
self.test_config.get('RDFSTORE', 'post'))
self.citation_metadata.title = "Count Obama mentions with Democratic party."
# Actual results
# Citation1
citation1.cite(select_statement=self.select_statement,
citation_metadata=self.citation_metadata,
citation_timestamp=self.citation_timestamp)
# Insert two new triple
prefixes = {
'pub': 'http://ontology.ontotext.com/taxonomy/',
'publishing': 'http://ontology.ontotext.com/publishing#'
}
mention = "<http://data.ontotext.com/publishing#newMention>"
hasInstance = "publishing:hasInstance"
person = "<http://ontology.ontotext.com/resource/tsk4wye1ftog>"
self.rdf_engine.insert_triples([mention, hasInstance, person],
prefixes)
document = "<http://www.reuters.com/article/2021/01/01/newDocument>"
containsMention = "publishing:containsMention"
self.rdf_engine.insert_triples([document, containsMention, mention],
prefixes)
# Citation2
vie_tz = timezone(timedelta(hours=2))
citation_timestamp2 = datetime.now(vie_tz)
citation2.cite(select_statement=self.select_statement,
citation_metadata=self.citation_metadata,
citation_timestamp=citation_timestamp2)
# Concat Citation1 + Citation2 into actual results
actual_result = citation1.citation_metadata.citation_snippet \
+ "\n" + citation2.citation_metadata.citation_snippet
self.citation = citation1
# Expected results
# Citation1
query_utils = QueryUtils(query=self.select_statement,
citation_timestamp=self.citation_timestamp)
citation_snippet1 = generate_citation_snippet(
identifier=query_utils.pid,
creator=self.citation_metadata.creator,
title=self.citation_metadata.title,
publisher=self.citation_metadata.publisher,
pulication_year=self.citation_metadata.publication_year,
resource_type=self.citation_metadata.resource_type)
# Citation2
query_utils = QueryUtils(query=self.select_statement,
citation_timestamp=citation_timestamp2)
citation_snippet2 = generate_citation_snippet(
identifier=query_utils.pid,
creator=self.citation_metadata.creator,
title=self.citation_metadata.title,
publisher=self.citation_metadata.publisher,
pulication_year=self.citation_metadata.publication_year,
resource_type=self.citation_metadata.resource_type)
expected_result = citation_snippet1 + "\n" + citation_snippet2
# Clean up
self.rdf_engine._delete_triples([mention, hasInstance, person],
prefixes)
self.rdf_engine._delete_triples([document, containsMention, mention],
prefixes)
test = Test(
test_number=9,
tc_desc=
'Test if a query that uses aggregation functions yields the right result before '
'and after an insert.',
expected_result=expected_result,
actual_result=actual_result)
return test
def test_citation__changed_dataset(self):
citation1 = self.citation
citation2 = ct.Citation(self.test_config.get('RDFSTORE', 'get'),
self.test_config.get('RDFSTORE', 'post'))
self.citation_metadata.title = "Obama occurrences, new mention"
# Actual results
# Citation1
citation1.cite(select_statement=self.select_statement,
citation_metadata=self.citation_metadata,
citation_timestamp=self.citation_timestamp)
# Insert two new triple
prefixes = {
'pub': 'http://ontology.ontotext.com/taxonomy/',
'publishing': 'http://ontology.ontotext.com/publishing#'
}
mention = "<http://data.ontotext.com/publishing#newMention>"
hasInstance = "publishing:hasInstance"
person = "<http://ontology.ontotext.com/resource/tsk4wye1ftog>"
self.rdf_engine.insert_triples([mention, hasInstance, person],
prefixes)
document = "<http://www.reuters.com/article/2021/01/01/newDocument>"
containsMention = "publishing:containsMention"
self.rdf_engine.insert_triples([document, containsMention, mention],
prefixes)
# Citation2
vie_tz = timezone(timedelta(hours=2))
citation_timestamp2 = datetime.now(vie_tz)
citation2.cite(select_statement=self.select_statement,
citation_metadata=self.citation_metadata,
citation_timestamp=citation_timestamp2)
# Concat Citation1 + Citation2 into actual results
actual_result = citation1.citation_metadata.citation_snippet \
+ "\n" + citation2.citation_metadata.citation_snippet
self.citation = citation1
# Expected results
# Citation1
query_utils = QueryUtils(query=self.select_statement,
citation_timestamp=self.citation_timestamp)
citation_snippet1 = generate_citation_snippet(
identifier=query_utils.pid,
creator=self.citation_metadata.creator,
title=self.citation_metadata.title,
publisher=self.citation_metadata.publisher,
pulication_year=self.citation_metadata.publication_year,
resource_type=self.citation_metadata.resource_type)
# Citation2
query_utils = QueryUtils(query=self.select_statement,
citation_timestamp=citation_timestamp2)
citation_snippet2 = generate_citation_snippet(
identifier=query_utils.pid,
creator=self.citation_metadata.creator,
title=self.citation_metadata.title,
publisher=self.citation_metadata.publisher,
pulication_year=self.citation_metadata.publication_year,
resource_type=self.citation_metadata.resource_type)
expected_result = citation_snippet1 + "\n" + citation_snippet2
# Clean up
self.rdf_engine._delete_triples([mention, hasInstance, person],
prefixes)
self.rdf_engine._delete_triples([document, containsMention, mention],
prefixes)
test = Test(
test_number=3,
tc_desc=
'Test if a new query PID is created if the dataset has changed since the last citation and'
'the query stayed the same. Check also if the query checksum must stayed the same.',
expected_result=expected_result,
actual_result=actual_result)
return test