def post(self, request):
serializer = ParseSerializer(data=request.data)
serializer.is_valid(raise_exception=True)
fragment = serializer.validated_data.get('fragment')
frbr_uri = FrbrUri.parse(serializer.validated_data.get('frbr_uri'))
importer = plugins.for_locale('importer', frbr_uri.country,
frbr_uri.language, frbr_uri.locality)
importer.fragment = fragment
importer.fragment_id_prefix = serializer.validated_data.get(
'id_prefix')
try:
text = serializer.validated_data.get('content')
xml = importer.import_from_text(text, frbr_uri.work_uri(), '.txt')
except ValueError as e:
log.error("Error during import: %s" % str(e), exc_info=e)
raise ValidationError({'content': str(e) or "error during import"})
# parse and re-serialize the XML to ensure it's clean, and sort out encodings
xml = Base(xml).to_xml()
# output
return Response({'output': xml})
def get(self, request, **kwargs):
# try parse it as an FRBR URI, if that succeeds, we'll lookup the document
# that document matches, otherwise we'll assume they're trying to
# list documents with a prefix URI match.
try:
self.frbr_uri = FrbrUri.parse(self.kwargs['frbr_uri'])
# ensure we haven't mistaken '/za-cpt/act/by-law/2011/full.atom' for a URI
if self.frbr_uri.number in ['full', 'summary'] and self.format_kwarg == 'atom':
raise ValueError()
# in a URL like
#
# /act/1980/1/toc
#
# don't mistake 'toc' for a language, it's really equivalent to
#
# /act/1980/1/eng/toc
#
# if eng is the default language.
if self.frbr_uri.language == 'toc':
self.frbr_uri.language = self.frbr_uri.default_language
self.frbr_uri.expression_component = 'toc'
return self.retrieve(request)
except ValueError:
return self.list(request)
def parse_frbr_uri(self, frbr_uri):
FrbrUri.default_language = None
try:
frbr_uri = FrbrUri.parse(frbr_uri)
except ValueError:
return None
# ensure we haven't mistaken '/za-cpt/act/by-law/2011/full.atom' for a URI
if frbr_uri.number in ['full', 'summary'
] and self.format_kwarg == 'atom':
return None
frbr_uri.default_language = self.country.primary_language.code
if not frbr_uri.language:
frbr_uri.language = frbr_uri.default_language
# in a URL like
#
# /act/1980/1/toc
#
# don't mistake 'toc' for a language, it's really equivalent to
#
# /act/1980/1/eng/toc
#
# if eng is the default language.
if frbr_uri.language == 'toc':
frbr_uri.language = frbr_uri.default_language
frbr_uri.expression_component = 'toc'
return frbr_uri
def test_dont_link_constitution_in_constitution(self):
constitution = Work(frbr_uri='/akn/za/act/1996/constitution')
doc = Document(work=constitution, content=document_fixture(xml="""
<section id="section-1">
<num>1.</num>
<heading>Tester</heading>
<paragraph id="section-1.paragraph-0">
<content>
<p>the Constitution</p>
<p>the Constitution, 1996</p>
<p>the Constitution of South Africa</p>
<p>the Constitution of the Republic of South Africa</p>
<p>the Constitution of the Republic of South Africa, 1996</p>
<p>the Constitution of the Republic of South Africa 1996</p>
<p>the Constitution of the Republic of South Africa Act, 1996</p>
<p>the Constitution of the Republic of South Africa, 1996 ( Act 108 of 1996 )</p>
<p>the Constitution of the Republic of South Africa, 1996 ( Act No 108 of 1996 )</p>
<p>the Constitution of the Republic of South Africa, 1996 ( Act No. 108 of 1996 )</p>
<p>the Constitution of the Republic of South Africa Act, 1996 ( Act No. 108 of 1996 )</p>
<p>the Constitution of the Republic of South Africa Act 108 of 1996</p>
<p>the Constitution of the Republic of South Africa (Act 108 of 1996)</p>
</content>
</paragraph>
</section>"""))
unchanged = doc.document_xml
doc.doc.frbr_uri = FrbrUri.parse(constitution.frbr_uri)
self.finder.find_references_in_document(doc)
self.assertMultiLineEqual(unchanged, doc.document_xml)
def validate_frbr_uri(self, value):
try:
if not value:
raise ValueError()
return FrbrUri.parse(value.lower()).work_uri()
except ValueError:
raise ValidationError("Invalid FRBR URI: %s" % value)
def find_work(self, given_string):
""" The string we get from the spreadsheet could be e.g.
`/ug/act/1933/14 - Administrator-General’s Act` (new and preferred style)
`Administrator-General’s Act` (old style)
First see if the string before the first space can be parsed as an FRBR URI, and find a work based on that.
If not, assume a title has been given and try to match on the whole string.
"""
first = given_string.split()[0]
try:
FrbrUri.parse(first)
return Work.objects.filter(frbr_uri=first).first()
except ValueError:
potential_matches = Work.objects.filter(title=given_string,
country=self.country,
locality=self.locality)
if len(potential_matches) == 1:
return potential_matches.first()
def get_for_frbr_uri(self, frbr_uri):
""" Find a single document matching the FRBR URI.
Raises ValueError if any part of the URI isn't valid.
See http://docs.oasis-open.org/legaldocml/akn-nc/v1.0/cs01/akn-nc-v1.0-cs01.html#_Toc492651893
"""
if not isinstance(frbr_uri, FrbrUri):
frbr_uri = FrbrUri.parse(frbr_uri)
query = self.filter(frbr_uri=frbr_uri.work_uri())
# filter on language
if frbr_uri.language:
query = query.filter(
language__language__iso_639_2B=frbr_uri.language)
# filter on expression date
expr_date = frbr_uri.expression_date
if not expr_date:
# no expression date is equivalent to the "current" version, at time of retrieval
expr_date = ':' + datetime.date.today().strftime("%Y-%m-%d")
try:
if expr_date == '@':
# earliest document
query = query.order_by('expression_date')
elif expr_date[0] == '@':
# document at this date
query = query.filter(
expression_date=parse_date(expr_date[1:]).date())
elif expr_date[0] == ':':
# latest document at or before this date
query = query \
.filter(expression_date__lte=parse_date(expr_date[1:]).date()) \
.order_by('-expression_date')
else:
raise ValueError("The expression date %s is not valid" %
expr_date)
except ParseError:
raise ValueError("The expression date %s is not valid" % expr_date)
obj = query.first()
if obj is None:
raise ValueError("Document doesn't exist")
if obj and frbr_uri.language and obj.language.code != frbr_uri.language:
raise ValueError(
"The document %s exists but is not available in the language '%s'"
% (frbr_uri.work_uri(), frbr_uri.language))
return obj
def new_frbr_uri(self, uri, forward, for_work=True):
""" Sets prefix on uri:
'akn' if forward is True, None if it's False.
"""
if not isinstance(uri, FrbrUri):
uri = FrbrUri.parse(uri)
uri.prefix = 'akn' if forward else None
if for_work:
return uri.work_uri()
else:
return uri.expression_uri()
def parse_frbr_uri(self, frbr_uri):
FrbrUri.default_language = None
try:
frbr_uri = FrbrUri.parse(frbr_uri)
except ValueError:
return None
frbr_uri.default_language = self.country.primary_language.code
if not frbr_uri.language:
frbr_uri.language = frbr_uri.default_language
return frbr_uri
def validate_frbr_uri(self, value):
try:
if not value:
raise ValueError()
value = FrbrUri.parse(value.lower()).work_uri()
except ValueError:
raise ValidationError("Invalid FRBR URI: %s" % value)
# does a work exist for this frbr_uri?
# raises ValueError if it doesn't
Work.objects.get_for_frbr_uri(value)
return value
def parse_frbr_uri(self, frbr_uri):
FrbrUri.default_language = None
self.frbr_uri = FrbrUri.parse(frbr_uri)
# validate the country and set the default language
try:
country = Country.for_frbr_uri(self.frbr_uri)
self.frbr_uri.default_language = country.primary_language.code
except Country.DoesNotExist:
raise Http404("Country %s from FRBR URI not found" %
self.frbr_uri.country)
if not self.frbr_uri.language:
self.frbr_uri.language = self.frbr_uri.default_language
def randomized(cls, frbr_uri, **kwargs):
""" Helper to return a new document with a random FRBR URI
"""
from .works import Work
from .places import Country
frbr_uri = FrbrUri.parse(frbr_uri)
kwargs['work'] = Work.objects.get_for_frbr_uri(frbr_uri.work_uri())
kwargs['language'] = Country.for_frbr_uri(frbr_uri).primary_language
doc = cls(frbr_uri=frbr_uri.work_uri(False),
expression_date=frbr_uri.expression_date,
**kwargs)
doc.copy_attributes()
return doc
def clean(self):
# validate and clean the frbr_uri
try:
frbr_uri = FrbrUri.parse(
self.frbr_uri).work_uri(work_component=False)
except ValueError:
raise ValidationError("Invalid FRBR URI")
# Assume frbr_uri starts with /akn; `rest` is everything after the country/locality, e.g.
# in `/akn/za-wc/act/2000/12`, `rest` is `act/2000/12`.
rest = frbr_uri.split('/', 3)[3]
# force akn prefix, country and locality codes in frbr uri
prefix = '/akn/' + self.country.code
if self.locality:
prefix = prefix + '-' + self.locality.code
self.frbr_uri = f'{prefix}/{rest}'.lower()
def test_dont_find_self(self):
document = Document(work=self.work,
frbr_uri=self.work.frbr_uri,
document_xml=document_fixture(xml="""
<section eId="sec_1">
<num>1.</num>
<heading>Tester</heading>
<paragraph eId="sec_1.paragraph-0">
<content>
<p>Something to do with Act 1 of 1991.</p>
<p>Something to do with Act no 22 of 2012.</p>
<p>And another thing about Act 4 of 1998.</p>
</content>
</paragraph>
</section>"""),
language=self.eng)
expected = Document(work=self.work,
document_xml=document_fixture(xml="""
<section eId="sec_1">
<num>1.</num>
<heading>Tester</heading>
<paragraph eId="sec_1.paragraph-0">
<content>
<p>Something to do with Act 1 of 1991.</p>
<p>Something to do with Act <ref href="/akn/za/act/2012/22">no 22 of 2012</ref>.</p>
<p>And another thing about Act <ref href="/akn/za/act/1998/4">4 of 1998</ref>.</p>
</content>
</paragraph>
</section>"""),
language=self.eng)
document.doc.frbr_uri = FrbrUri.parse(self.work.frbr_uri)
self.finder.find_references_in_document(document)
root = etree.fromstring(expected.content)
expected.content = etree.tostring(root,
encoding='utf-8').decode('utf-8')
self.assertEqual(expected.content, document.content)
def post(self, request):
serializer = ParseSerializer(data=request.data)
serializer.is_valid(raise_exception=True)
fragment = serializer.validated_data.get('fragment')
frbr_uri = FrbrUri.parse(serializer.validated_data.get('frbr_uri'))
importer = plugins.for_locale('importer', frbr_uri.country, frbr_uri.language, frbr_uri.locality)
importer.fragment = fragment
importer.fragment_id_prefix = serializer.validated_data.get('id_prefix')
upload = self.request.data.get('file')
if upload:
# we got a file
try:
document = importer.import_from_upload(upload, frbr_uri.work_uri(), self.request)
except ValueError as e:
log.error("Error during import: %s" % e.message, exc_info=e)
raise ValidationError({'file': e.message or "error during import"})
else:
# plain text
try:
text = serializer.validated_data.get('content')
document = importer.import_from_text(text, frbr_uri.work_uri())
except ValueError as e:
log.error("Error during import: %s" % e.message, exc_info=e)
raise ValidationError({'content': e.message or "error during import"})
if not document:
raise ValidationError("Nothing to parse! Either 'file' or 'content' must be provided.")
# output
if fragment:
return Response({'output': document.to_xml()})
else:
return Response({'output': document.document_xml})
def work_uri(self):
""" The FRBR Work URI as a :class:`FrbrUri` instance that uniquely identifies this work universally. """
if self._work_uri is None:
self._work_uri = FrbrUri.parse(self.frbr_uri)
return self._work_uri
