Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Feature/ted 1312 #471

Merged
merged 2 commits into from
Apr 11, 2023
Merged

Feature/ted 1312 #471

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
1adc078
Updates for XPATH Queries
kaleanych Apr 6, 2023
5994de7
Tests coverage
kaleanych Apr 7, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 2 additions & 0 deletions ted_sws/core/model/manifestation.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -35,6 +35,7 @@ class SPARQLQueryRefinedResultType(Enum):
INVALID = "invalid"
ERROR = "error"
WARNING = "warning"
UNKNOWN = "unknown"


class Manifestation(PropertyBaseModel):
Expand Down Expand Up @@ -224,6 +225,7 @@ class SPARQLQueryResult(PropertyBaseModel):
fields_covered: Optional[bool] = True
missing_fields: Optional[List[str]] = []
error: Optional[str]
message: Optional[str]
identifier: Optional[str]

class Config:
Expand Down
1 change: 1 addition & 0 deletions ted_sws/core/model/validation_report.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,6 +30,7 @@ class SPARQLValidationSummaryCountReport(PropertyBaseModel):
warning: Optional[QueryValidationSummaryCountReportStatus] = QueryValidationSummaryCountReportStatus()
invalid: Optional[QueryValidationSummaryCountReportStatus] = QueryValidationSummaryCountReportStatus()
error: Optional[QueryValidationSummaryCountReportStatus] = QueryValidationSummaryCountReportStatus()
unknown: Optional[QueryValidationSummaryCountReportStatus] = QueryValidationSummaryCountReportStatus()


class SPARQLSummaryQuery(PropertyBaseModel):
Expand Down
6 changes: 4 additions & 2 deletions ted_sws/mapping_suite_processor/services/conceptual_mapping_generate_sparql_queries.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,13 +26,15 @@
SPARQL_PREFIX_LINE = 'PREFIX {prefix}: <{value}>'
SPARQL_LOGGER_NAME = "SPARQL"

SPARQL_XPATH_SEPARATOR = " ;; "


def get_sparql_prefixes(sparql_q: str) -> list:
finds: list = re.findall(SPARQL_PREFIX_PATTERN, sparql_q)
return sorted(set(finds))


def concat_field_xpath(base_xpath: str, field_xpath: str, separator: str = ", ") -> str:
def concat_field_xpath(base_xpath: str, field_xpath: str, separator: str = SPARQL_XPATH_SEPARATOR) -> str:
base_xpath = base_xpath if not pd.isna(base_xpath) else ''
field_xpath = field_xpath if not pd.isna(field_xpath) else ''
return separator.join(
Expand Down Expand Up @@ -137,7 +139,7 @@ def sparql_validation_generator(data: pd.DataFrame, base_xpath: str, controlled_
f"{e_form_bt_name}” in eForms. The corresponding XML element is " \
f"{concat_field_xpath(base_xpath, field_xpath)}. " \
f"The expected ontology instances are epo: {class_path} .\n" \
f"#xpath: {concat_field_xpath(base_xpath, field_xpath, separator=',')}" \
f"#xpath: {concat_field_xpath(base_xpath, field_xpath, separator=SPARQL_XPATH_SEPARATOR)}" \
"\n" + "\n" + "\n".join(prefixes) + "\n\n" \
f"ASK WHERE {{ " \
f"{subject_type_display}" \
Expand Down
10 changes: 9 additions & 1 deletion ted_sws/notice_validator/resources/templates/sparql_query_results_report.jinja2
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -124,7 +124,7 @@
</tr>
</thead>
<tbody>
{% for _value in ['valid', 'unverifiable', 'warning', 'invalid', 'error'] %}
{% for _value in ['valid', 'unverifiable', 'warning', 'invalid', 'error', 'unknown'] %}
<tr>
{% set _results = validation_results | selectattr("result", "equalto", _value) | list | count %}
<td class="strong {{ _value }}">{{ _value }}</td>
Expand Down Expand Up @@ -187,8 +187,16 @@
</ol>
</li>
{% endif %}
{% if result.message and (result.message|length > 0) %}
<li>
<hr>
<h4 class="info strong">MESSAGE:</h4>
{{ result.message }}
</li>
{% endif %}
{% if result.error and (result.error|length > 0) %}
<li>
<hr>
<h4 class="danger strong">ERROR:</h4>
{{ result.error }}
</li>
Expand Down
2 changes: 1 addition & 1 deletion ted_sws/notice_validator/resources/templates/sparql_summary_report.jinja2
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
{% set result_values = ['valid', 'unverifiable', 'warning', 'invalid', 'error'] %}
{% set result_values = ['valid', 'unverifiable', 'warning', 'invalid', 'error', 'unknown'] %}

<!DOCTYPE html>
<html lang="en">
Expand Down
8 changes: 3 additions & 5 deletions ted_sws/notice_validator/services/shacl_test_suite_runner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -185,7 +185,7 @@ def validate_notice_with_shacl_suite(notice: Notice, mapping_suite_package: Mapp
:return:
"""

def shacl_validation(notice_item: Notice, rdf_manifestation: RDFManifestation, with_html: bool = False) \
def shacl_validation(notice_item: Notice, rdf_manifestation: RDFManifestation) \
-> List[SHACLTestSuiteValidationReport]:
reports = []
shacl_test_suites = mapping_suite_package.shacl_test_suites
Expand All @@ -199,12 +199,10 @@ def shacl_validation(notice_item: Notice, rdf_manifestation: RDFManifestation, w
return sorted(reports, key=lambda x: x.test_suite_identifier)

if execute_full_validation:
for report in shacl_validation(notice_item=notice, rdf_manifestation=notice.rdf_manifestation,
with_html=with_html):
for report in shacl_validation(notice_item=notice, rdf_manifestation=notice.rdf_manifestation):
notice.set_rdf_validation(rdf_validation=report)

for report in shacl_validation(notice_item=notice, rdf_manifestation=notice.distilled_rdf_manifestation,
with_html=with_html):
for report in shacl_validation(notice_item=notice, rdf_manifestation=notice.distilled_rdf_manifestation):
notice.set_distilled_rdf_validation(rdf_validation=report)

return notice
Expand Down
204 changes: 127 additions & 77 deletions ted_sws/notice_validator/services/sparql_test_suite_runner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,6 +12,7 @@
ReportNotice
from ted_sws.core.model.validation_report_data import ReportPackageNoticeData
from ted_sws.data_manager.adapters.repository_abc import NoticeRepositoryABC, MappingSuiteRepositoryABC
from ted_sws.mapping_suite_processor.services.conceptual_mapping_generate_sparql_queries import SPARQL_XPATH_SEPARATOR
from ted_sws.notice_transformer.adapters.notice_transformer import NoticeTransformer
from ted_sws.notice_validator.adapters.sparql_runner import SPARQLRunner
from ted_sws.notice_validator.resources.templates import TEMPLATE_METADATA_KEY
Expand Down Expand Up @@ -58,11 +59,25 @@ def _sparql_query_from_file_resource(cls, file_resource: FileResource) -> SPARQL
if QUERY_METADATA_TITLE in metadata else DEFAULT_QUERY_TITLE
description = metadata[QUERY_METADATA_DESCRIPTION] \
if QUERY_METADATA_DESCRIPTION in metadata else DEFAULT_QUERY_DESCRIPTION
xpath = metadata[QUERY_METADATA_XPATH].split(",") if QUERY_METADATA_XPATH in metadata and metadata[
QUERY_METADATA_XPATH] else DEFAULT_QUERY_XPATH
xpath = metadata[QUERY_METADATA_XPATH].split(
SPARQL_XPATH_SEPARATOR
) if QUERY_METADATA_XPATH in metadata and metadata[QUERY_METADATA_XPATH] else DEFAULT_QUERY_XPATH
query = cls._sanitize_query(file_resource.file_content)
return SPARQLQuery(title=title, description=description, xpath=xpath, query=query)

@classmethod
def _refined_result(cls, ask_answer, sparql_query_result,
result: SPARQLQueryRefinedResultType) -> SPARQLQueryRefinedResultType:
if ask_answer and sparql_query_result.fields_covered:
result = SPARQLQueryRefinedResultType.VALID.value
elif not ask_answer and not sparql_query_result.fields_covered:
result = SPARQLQueryRefinedResultType.UNVERIFIABLE.value
elif ask_answer and not sparql_query_result.fields_covered:
result = SPARQLQueryRefinedResultType.WARNING.value
elif not ask_answer and sparql_query_result.fields_covered:
result = SPARQLQueryRefinedResultType.INVALID.value
return result

def _process_sparql_ask_result(self, query_result, sparql_query: SPARQLQuery,
sparql_query_result: SPARQLQueryResult):
ask_answer = query_result.askAnswer
Expand All @@ -87,14 +102,7 @@ def _process_sparql_ask_result(self, query_result, sparql_query: SPARQLQuery,
sparql_query_result.missing_fields = list(sparql_query_xpath - xpaths_in_notice)

# Refined result
if ask_answer and sparql_query_result.fields_covered:
result = SPARQLQueryRefinedResultType.VALID.value
elif not ask_answer and not sparql_query_result.fields_covered:
result = SPARQLQueryRefinedResultType.UNVERIFIABLE.value
elif ask_answer and not sparql_query_result.fields_covered:
result = SPARQLQueryRefinedResultType.WARNING.value
elif not ask_answer and sparql_query_result.fields_covered:
result = SPARQLQueryRefinedResultType.INVALID.value
result = self._refined_result(ask_answer, sparql_query_result, result)

sparql_query_result.result = result

Expand Down Expand Up @@ -153,11 +161,75 @@ def generate_report(self) -> SPARQLTestSuiteValidationReport:
return self.sparql_test_suite_execution


def generate_sparql_validation_summary_report(report_notices: List[ReportNotice], mapping_suite_package: MappingSuite,
execute_full_validation: bool = True,
with_html: bool = False,
report: SPARQLValidationSummaryReport = None,
metadata: dict = None) -> SPARQLValidationSummaryReport:
def process_sparql_validation_summary_report_data_with_notice(
notice: Notice,
mapping_suite_package: MappingSuite,
report_notice_path: Path,
report: SPARQLValidationSummaryReport
):
for sparql_validation in notice.rdf_manifestation.sparql_validations:
test_suite_id = sparql_validation.test_suite_identifier
report.test_suite_ids.append(test_suite_id)
mapping_suite_versioned_id = sparql_validation.mapping_suite_identifier
report.mapping_suite_ids.append(mapping_suite_versioned_id)
validation: SPARQLQueryResult
for validation in sparql_validation.validation_results:
validation_query_result: SPARQLValidationSummaryQueryResult
found_validation_query_result = list(filter(
lambda record:
(record.query.query == validation.query.query)
and (record.query.title == validation.query.title)
and (record.test_suite_identifier == test_suite_id),
report.validation_results
))
if found_validation_query_result:
validation_query_result = found_validation_query_result[0]
else:
validation_query_result = SPARQLValidationSummaryQueryResult(
test_suite_identifier=test_suite_id,
**validation.dict()
)
notice_data: ReportPackageNoticeData = ReportPackageNoticeData(
notice_id=notice.ted_id,
path=str(report_notice_path),
mapping_suite_versioned_id=mapping_suite_versioned_id,
mapping_suite_identifier=mapping_suite_package.identifier
)
if validation.result == SPARQLQueryRefinedResultType.VALID.value:
validation_query_result.aggregate.valid.count += 1
validation_query_result.aggregate.valid.notices.append(notice_data)
elif validation.result == SPARQLQueryRefinedResultType.UNVERIFIABLE.value:
validation_query_result.aggregate.unverifiable.count += 1
validation_query_result.aggregate.unverifiable.notices.append(notice_data)
elif validation.result == SPARQLQueryRefinedResultType.INVALID.value:
validation_query_result.aggregate.invalid.count += 1
validation_query_result.aggregate.invalid.notices.append(notice_data)
elif validation.result == SPARQLQueryRefinedResultType.WARNING.value:
validation_query_result.aggregate.warning.count += 1
validation_query_result.aggregate.warning.notices.append(notice_data)
elif validation.result == SPARQLQueryRefinedResultType.ERROR.value:
validation_query_result.aggregate.error.count += 1
validation_query_result.aggregate.error.notices.append(notice_data)
elif validation.result == SPARQLQueryRefinedResultType.UNKNOWN.value:
validation_query_result.aggregate.unknown.count += 1
validation_query_result.aggregate.unknown.notices.append(notice_data)
if not found_validation_query_result:
report.validation_results.append(validation_query_result)


def add_sparql_validation_summary_html_report(
report: SPARQLValidationSummaryReport,
metadata: dict = None
):
template_data: dict = report.dict()
template_data[TEMPLATE_METADATA_KEY] = metadata
html_report = TEMPLATES.get_template(SPARQL_SUMMARY_HTML_REPORT_TEMPLATE).render(template_data)
report.object_data = html_report


def init_sparql_validation_summary_report(
report: SPARQLValidationSummaryReport,
report_notices: List[ReportNotice]) -> SPARQLValidationSummaryReport:
if report is None:
report: SPARQLValidationSummaryReport = SPARQLValidationSummaryReport(
object_data="SPARQLValidationSummaryReport",
Expand All @@ -168,6 +240,32 @@ def generate_sparql_validation_summary_report(report_notices: List[ReportNotice]
report.notices = sorted(NoticeTransformer.transform_validation_report_notices(report_notices, group_depth=1) + (
report.notices or []), key=lambda report_data: report_data.notice_id)

return report


def finalize_sparql_validation_summary_report(report: SPARQLValidationSummaryReport, metadata: dict = None,
with_html: bool = False):
report.test_suite_ids = list(set(report.test_suite_ids))
report.mapping_suite_ids = list(set(report.mapping_suite_ids))

if with_html:
add_sparql_validation_summary_html_report(
report=report,
metadata=metadata
)


def generate_sparql_validation_summary_report(report_notices: List[ReportNotice],
mapping_suite_package: MappingSuite,
execute_full_validation: bool = True,
with_html: bool = False,
report: SPARQLValidationSummaryReport = None,
metadata: dict = None) -> SPARQLValidationSummaryReport:
report = init_sparql_validation_summary_report(
report=report,
report_notices=report_notices
)

for report_notice in report_notices:
notice = report_notice.notice
validate_notice_with_sparql_suite(
Expand All @@ -176,65 +274,19 @@ def generate_sparql_validation_summary_report(report_notices: List[ReportNotice]
execute_full_validation=execute_full_validation,
with_html=False
)
for sparql_validation in notice.rdf_manifestation.sparql_validations:
test_suite_id = sparql_validation.test_suite_identifier
report.test_suite_ids.append(test_suite_id)
mapping_suite_versioned_id = sparql_validation.mapping_suite_identifier
report.mapping_suite_ids.append(mapping_suite_versioned_id)

validation: SPARQLQueryResult
for validation in sparql_validation.validation_results:
validation_query_result: SPARQLValidationSummaryQueryResult
found_validation_query_result = list(filter(
lambda record:
(record.query.query == validation.query.query)
and (record.query.title == validation.query.title)
and (record.test_suite_identifier == test_suite_id),
report.validation_results
))

if found_validation_query_result:
validation_query_result = found_validation_query_result[0]
else:
validation_query_result = SPARQLValidationSummaryQueryResult(
test_suite_identifier=test_suite_id,
**validation.dict()
)

notice_data: ReportPackageNoticeData = ReportPackageNoticeData(
notice_id=notice.ted_id,
path=str(report_notice.metadata.path),
mapping_suite_versioned_id=mapping_suite_versioned_id,
mapping_suite_identifier=mapping_suite_package.identifier
)

if validation.result == SPARQLQueryRefinedResultType.VALID.value:
validation_query_result.aggregate.valid.count += 1
validation_query_result.aggregate.valid.notices.append(notice_data)
elif validation.result == SPARQLQueryRefinedResultType.UNVERIFIABLE.value:
validation_query_result.aggregate.unverifiable.count += 1
validation_query_result.aggregate.unverifiable.notices.append(notice_data)
elif validation.result == SPARQLQueryRefinedResultType.INVALID.value:
validation_query_result.aggregate.invalid.count += 1
validation_query_result.aggregate.invalid.notices.append(notice_data)
elif validation.result == SPARQLQueryRefinedResultType.WARNING.value:
validation_query_result.aggregate.warning.count += 1
validation_query_result.aggregate.warning.notices.append(notice_data)
elif validation.result == SPARQLQueryRefinedResultType.ERROR.value:
validation_query_result.aggregate.error.count += 1
validation_query_result.aggregate.error.notices.append(notice_data)

if not found_validation_query_result:
report.validation_results.append(validation_query_result)

report.test_suite_ids = list(set(report.test_suite_ids))
report.mapping_suite_ids = list(set(report.mapping_suite_ids))
process_sparql_validation_summary_report_data_with_notice(
notice=notice,
mapping_suite_package=mapping_suite_package,
report_notice_path=report_notice.metadata.path,
report=report
)

if with_html:
template_data: dict = report.dict()
template_data[TEMPLATE_METADATA_KEY] = metadata
html_report = TEMPLATES.get_template(SPARQL_SUMMARY_HTML_REPORT_TEMPLATE).render(template_data)
report.object_data = html_report
finalize_sparql_validation_summary_report(
report=report,
metadata=metadata,
with_html=with_html
)

return report

Expand All @@ -250,7 +302,7 @@ def validate_notice_with_sparql_suite(notice: Notice, mapping_suite_package: Map
:return:
"""

def sparql_validation(notice_item: Notice, rdf_manifestation: RDFManifestation, with_html: bool = False) \
def sparql_validation(notice_item: Notice, rdf_manifestation: RDFManifestation) \
-> List[SPARQLTestSuiteValidationReport]:
reports = []
sparql_test_suites = mapping_suite_package.sparql_test_suites
Expand All @@ -266,12 +318,10 @@ def sparql_validation(notice_item: Notice, rdf_manifestation: RDFManifestation,
return sorted(reports, key=lambda x: x.test_suite_identifier)

if execute_full_validation:
for report in sparql_validation(notice_item=notice, rdf_manifestation=notice.rdf_manifestation,
with_html=with_html):
for report in sparql_validation(notice_item=notice, rdf_manifestation=notice.rdf_manifestation):
notice.set_rdf_validation(rdf_validation=report)

for report in sparql_validation(notice_item=notice, rdf_manifestation=notice.distilled_rdf_manifestation,
with_html=with_html):
for report in sparql_validation(notice_item=notice, rdf_manifestation=notice.distilled_rdf_manifestation):
notice.set_distilled_rdf_validation(rdf_validation=report)

return notice
Expand Down
Loading