def test_strip_xml_nodes_by_xpaths_similar_names(self):
expected_subgroup = ('<root>'
' <group1>'
' <subgroup1>'
' <question_1>Answer 1</question_1>'
' <question_2>Answer 2</question_2>'
' </subgroup1>'
' </group1>'
'</root>')
expected_group = ('<root>'
' <group1>'
' <subgroup1>'
' <question_1>Answer 1</question_1>'
' <question_2>Answer 2</question_2>'
' </subgroup1>'
' <subgroup11>'
' <question_3>Answer 3</question_3>'
' <question_4>Answer 4</question_4>'
' </subgroup11>'
' <question_5>Answer 5</question_5>'
' </group1>'
'</root>')
self.__compare_xml(
strip_nodes(self.__submission, ['group1/subgroup1'],
use_xpath=True),
expected_subgroup,
)
self.__compare_xml(
strip_nodes(self.__submission, ['group1'], use_xpath=True),
expected_group,
)
def test_strip_xml_nodes_by_xpaths_with_slashes(self):
expected = ('<root>'
' <group1>'
' <question_5>Answer 5</question_5>'
' </group1>'
'</root>')
# With trailing slash
self.__compare_xml(
strip_nodes(self.__submission, ['group1/question_5/'],
use_xpath=True),
expected,
)
# With leading slash
self.__compare_xml(
strip_nodes(self.__submission, ['/group1/question_5'],
use_xpath=True),
expected,
)
# With both
self.__compare_xml(
strip_nodes(self.__submission, ['/group1/question_5/'],
use_xpath=True),
expected,
)
def test_strip_xml_nodes_and_rename_root_node(self):
source = '<abcdef><a><b><c>abcdef</c></b></a></abcdef>'
expected = '<root><a><b><c>abcdef</c></b></a></root>'
result = strip_nodes(
source=source,
nodes_to_keep=['a', 'b', 'c'],
rename_root_node_to='root',
)
self.__compare_xml(result, expected)
def test_strip_xml_nodes_by_fields(self):
expected = ('<root>'
' <group1>'
' <subgroup1>'
' <question_1>Answer 1</question_1>'
' </subgroup1>'
' <question_5>Answer 5</question_5>'
' </group1>'
'</root>')
self.__compare_xml(
strip_nodes(self.__submission, ['question_1', 'question_5']),
expected,
)
expected = ('<root>'
' <group1>'
' <question_5>Answer 5</question_5>'
' </group1>'
'</root>')
self.__compare_xml(
strip_nodes(self.__submission, ['question_5']),
expected,
)
def test_strip_xml_nodes_by_xpaths(self):
expected = ('<root>'
' <group1>'
' <subgroup1>'
' <question_1>Answer 1</question_1>'
' </subgroup1>'
' <question_5>Answer 5</question_5>'
' </group1>'
'</root>')
self.__compare_xml(
strip_nodes(
self.__submission,
['group1/subgroup1/question_1', 'group1/question_5'],
use_xpath=True,
),
expected,
)
def external(self, request, paired_data_uid, **kwargs):
"""
Returns an XML which contains data submitted to paired asset
Creates the endpoints
- /api/v2/assets/<parent_lookup_asset>/paired-data/<paired_data_uid>/external/
- /api/v2/assets/<parent_lookup_asset>/paired-data/<paired_data_uid>/external.xml/
"""
paired_data = self.get_object()
# Retrieve the source if it exists
source_asset = paired_data.get_source()
if not source_asset:
# We can enter this condition when source data sharing has been
# deactivated after it has been paired with current form.
# We don't want to keep zombie files on storage.
try:
asset_file = self.asset.asset_files.get(uid=paired_data_uid)
except AssetFile.DoesNotExist:
pass
else:
asset_file.delete()
raise Http404
if not source_asset.has_deployment or not self.asset.has_deployment:
raise Http404
old_hash = None
# Retrieve data from related asset file.
# If data has already been fetched once, an `AssetFile` should exist.
# Otherwise, we create one to store the generated XML.
try:
asset_file = self.asset.asset_files.get(uid=paired_data_uid)
except AssetFile.DoesNotExist:
asset_file = AssetFile(
uid=paired_data_uid,
asset=self.asset,
file_type=AssetFile.PAIRED_DATA,
user=self.asset.owner,
)
# When asset file is new, we consider its content as expired to
# force its creation below
has_expired = True
else:
if not asset_file.content:
# if `asset_file` exists but does not have any content, it means
# `paired_data` has changed since last time this endpoint has been
# called. E.g.: Project owner has changed the questions they want
# to include in the `xml-external` file
has_expired = True
else:
old_hash = asset_file.md5_hash
timedelta = timezone.now() - asset_file.date_modified
has_expired = (timedelta.total_seconds() >
settings.PAIRED_DATA_EXPIRATION)
# ToDo evaluate adding headers for caching and a HTTP 304 status code
if not has_expired:
return Response(asset_file.content.file.read().decode())
# If the content of `asset_file' has expired, let's regenerate the XML
submissions = source_asset.deployment.get_submissions(
self.asset.owner, format_type=SUBMISSION_FORMAT_TYPE_XML)
parsed_submissions = []
for submission in submissions:
# Use `rename_root_node_to='data'` to rename the root node of each
# submission to `data` so that form authors do not have to rewrite
# their `xml-external` formulas any time the asset UID changes,
# e.g. when cloning a form or creating a project from a template.
# Set `use_xpath=True` because `paired_data.fields` uses full group
# hierarchies, not just question names.
parsed_submissions.append(
strip_nodes(
submission,
paired_data.allowed_fields,
use_xpath=True,
rename_root_node_to='data',
))
filename = paired_data.filename
parsed_submissions_to_str = ''.join(parsed_submissions)
root_tag_name = SubmissionXMLRenderer.root_tag_name
xml_ = add_xml_declaration(f'<{root_tag_name}>'
f'{parsed_submissions_to_str}'
f'</{root_tag_name}>')
if not parsed_submissions:
# We do not want to cache an empty file
return Response(xml_)
# We need to delete the current file (if it exists) when filename
# has changed. Otherwise, it would leave an orphan file on storage
if asset_file.pk and asset_file.content.name != filename:
asset_file.content.delete()
asset_file.content = ContentFile(xml_.encode(), name=filename)
# `xml_` is already there in memory, let's use its content to get its
# hash and store it within `asset_file` metadata
asset_file.set_md5_hash(calculate_hash(xml_, prefix=True))
asset_file.save()
if old_hash != asset_file.md5_hash:
# resync paired data to the deployment backend
self.asset.deployment.sync_media_files(AssetFile.PAIRED_DATA)
return Response(xml_)
def _parse_data(self, submission, fields):
return strip_nodes(submission, fields, xml_declaration=True)