def test_multiple_keyterms(self):
self.assertEqual(
list(
tokenize('Decreto-Barros',
{'Decreto', 'Decreto-Lei', '-Barro'})),
[Token('Decreto'), Token('-Barro'),
Token('s')])
def test_keyterm_in_begin(self):
self.assertEqual(list(tokenize('pre-foo-suf', ('pre', 'pre-foo-suf'))),
[Token('pre-foo-suf')])
self.assertEqual(
list(tokenize('d-pre-foo-suf', ('pre', 'pre-foo-suf'))),
[Token('d-'), Token('pre-foo-suf')])
def test_shifted_keyterms(self):
self.assertEqual(
list(
tokenize('the foo is bad',
(' ', 'the foo stay', 'foo is bad'))),
[Token('the'), Token(' '),
Token('foo is bad')])
def observe(self, index, token, caught):
if not caught and re.match(DOCUMENT_NUMBER_REGEX, token.as_str()):
self._numbers[index] = token
return True
if token in (Token('.'), Token('\n')):
self.finish()
return False
def test_keyterm_in_end(self):
self.assertEqual(list(tokenize('pre-foo-suf', ('pre-foo-suf', 'suf'))),
[Token('pre-foo-suf')])
self.assertEqual(
list(tokenize('n.º 2', (' ', 'n.º', '.º'))),
[Token('n.º'), Token(' '), Token('2')])
self.assertEqual(list(tokenize('foo-bar', ('foo-bar', 'bar tal'))),
[Token('foo-bar')])
def test_single(self):
managers = [ObserverManager({'Decreto-Lei': DocumentRefObserver})]
self._test('Decreto-Lei nº 2/2013.', managers,
[(DocumentReference('2/2013', Token('Decreto-Lei')), 4)])
self._test('Decreto-Lei nº 2/2013/A,', managers,
[(DocumentReference('2/2013/A', Token('Decreto-Lei')), 4)])
self._test('Decreto-Lei nº 2-A/2013,', managers,
[(DocumentReference('2-A/2013', Token('Decreto-Lei')), 4)])
def test_single(self):
managers = [ObserverManager({'Diretiva': EULawRefObserver,
'Decisão de Execução': EULawReference})]
self._test('Diretiva nº 2011/778/UE ', managers,
[(EULawReference('2011/778/UE', Token('Diretiva')), 4)])
self._test('Diretiva nº 2000/29/CE,', managers,
[(EULawReference('2000/29/CE', Token('Diretiva')), 4)])
self._test('Diretiva nº 2000/778/UE.', managers,
[(EULawReference('2000/778/UE', Token('Diretiva')), 4)])
self._test('Diretiva False.', managers, ())
def test_keyterm_subset_of_keyterm(self):
"""
When keyterm is a subset of the other, return other.
"""
self.assertEqual(list(tokenize('Decreto-Lei', {'Decreto'})),
[Token('Decreto'), Token('-Lei')])
self.assertEqual(
list(tokenize('Decreto-Lei', {'Decreto', 'Decreto-Lei'})),
[Token('Decreto-Lei')])
self.assertEqual(
list(tokenize('Decreto-Barro', {'Decreto', 'Decreto-Lei'})),
[Token('Decreto'), Token('-Barro')])
def test_many_separated(self):
managers = [ObserverManager({'foo': DocumentRefObserver,
'bar': DocumentRefObserver})]
string = 'foo 1/2000. bar 2/2000'
expected = [(DocumentReference('1/2000', Token('foo')), 2),
(DocumentReference('2/2000', Token('bar')), 7)]
self._test(string, managers, expected)
string = 'foo 1/2000 bar 2/2000'
expected = [(DocumentReference('1/2000', Token('foo')), 2),
(DocumentReference('2/2000', Token('bar')), 6)]
self._test(string, managers, expected)
def analyse(tokens):
root = Document()
root_parser = HierarchyParser(root)
paragraph = Paragraph()
block_mode = False
for token in tokens:
if token.as_str() == '':
continue
# start of quote
if token == Token('«') and len(paragraph) == 0:
block_mode = True
block_parser = HierarchyParser(QuotationSection(), add_links=False)
# end of quote
elif token == Token('»') and len(paragraph) == 0:
block_mode = False
root_parser.add(block_parser.root)
paragraph = Paragraph()
# construct the paragraphs
# paragraph can end by '\n' or by starting a new section.
elif isinstance(token, Anchor) or token.string == '\n':
# it is end of paragraph; complete it if it ends by a normal \n.
if token.string == '\n':
paragraph.append(token)
# select current parser
p = root_parser
if block_mode:
p = block_parser
# add paragraph to the current parser if it is not empty
if len(paragraph):
p.add(paragraph)
# start a new paragraph
paragraph = Paragraph()
if isinstance(token, Anchor):
# create a new section
section = p.new_section(token)
# if new new section is inline, change to inline paragraph
if isinstance(section, InlineDocumentSection):
paragraph = InlineParagraph()
else:
paragraph.append(token)
return root
def test_keyterm_in_word(self):
"""
262.º with keyterm `.º` must return 262 and .º
"""
self.assertEqual(list(tokenize('262.º', (
' ',
',',
'.º',
))), [Token('262'), Token('.º')])
self.assertEqual(
list(tokenize('262.º-A', (
' ',
',',
'.º',
))),
[Token('262'), Token('.º'), Token('-A')])
def test_keyterm_not_found(self):
self.assertEqual(
list(tokenize('this is the foo of', (
' ',
'the end',
))), [
Token('this'),
Token(' '),
Token('is'),
Token(' '),
Token('the'),
Token(' '),
Token('foo'),
Token(' '),
Token('of')
])
def test_with_document(self):
managers = [ObserverManager({'Decreto-Lei': DocumentRefObserver}),
ObserverManager({'artigo': ArticleRefObserver}),
ObserverManager({'nº': NumberRefObserver,
'n.os': NumberRefObserver})]
doc = DocumentReference('2/2013', Token('Decreto-Lei'))
art = ArticleReference('26º', doc)
self._test('no nº 2 do artigo 26º do Decreto-Lei 2/2013,',
managers, [(doc, 16), (art, 10),
(NumberReference('2', art), 4)])
def observe(self, index, token, caught):
if token in (Token('.'), Token('\n')):
self.finish()
return False
if self._parent:
return False
if not caught and re.match(LINE_REGEX, token.as_str()):
self._numbers[index] = token
return True
elif re.match(NUMBER_REGEX, token.as_str()):
self._parent = index
elif re.match(ARTICLE_NUMBER_REGEX, token.as_str()):
self._parent = index
# never found such case
# elif re.match(DOCUMENT_NUMBER_REGEX, token.as_str()):
# self._parent = index
return False
def test_with_document(self):
managers = [ObserverManager({'Decreto-Lei': DocumentRefObserver}),
ObserverManager({'artigos': ArticleRefObserver})]
doc = DocumentReference('2/2013', Token('Decreto-Lei'))
self._test('Os artigos 3º, 4º-A, 7º e 25º do Decreto-Lei 2/2013,',
managers,
[(ArticleReference('3º', doc), 4),
(ArticleReference('4º-A', doc), 7),
(ArticleReference('7º', doc), 10),
(ArticleReference('25º', doc), 14),
(doc, 20)])
def test_with_document(self):
managers = [ObserverManager({'Decreto-Lei': DocumentRefObserver}),
ObserverManager({'artigo': ArticleRefObserver}),
ObserverManager({'nº': NumberRefObserver}),
ObserverManager({'alíneas': LineRefObserver})]
document = DocumentReference('2/2013', Token('Decreto-Lei'))
article = ArticleReference('26º', document)
number = NumberReference('4', article)
self._test('nas alíneas f) e g) do nº 4 do artigo 26º '
'do Decreto-Lei nº 2/2013', managers,
[(LineReference('f)', number), 4),
(LineReference('g)', number), 8), (number, 14),
(article, 20), (document, 28)])
def test_tokenize(self):
self.assertEqual(
list(tokenize('the end', ' ')),
[Token('the'), Token(' '), Token('end')])
self.assertEqual(list(tokenize('the end is', (
' ',
'the end',
))), [Token('the end'), Token(' '),
Token('is')])
def test_similar_keyterms(self):
expected = [
Token('this'),
Token(' '),
Token('is'),
Token(' '),
Token('the'),
Token(' '),
Token('foo'),
Token(' '),
Token('of')
]
self.assertEqual(
list(tokenize('this is the foo of', (' ', 'the end', 'the bar'))),
expected)
self.assertEqual(
list(
tokenize('this is the foo of',
(' ', 'the foo is', 'foo is bad'))), expected)
def test_many(self):
managers = [ObserverManager({'Decretos-Leis': DocumentRefObserver})]
self._test('Decretos-Leis n.os 1/2006, 2/2006, e 3/2006', managers,
[(DocumentReference('1/2006', Token('Decretos-Leis')), 4),
(DocumentReference('2/2006', Token('Decretos-Leis')), 7),
(DocumentReference('3/2006', Token('Decretos-Leis')), 12)])
self._test('Decretos-Leis n.os 1/2006, e 2/2006', managers,
[(DocumentReference('1/2006', Token('Decretos-Leis')), 4),
(DocumentReference('2/2006', Token('Decretos-Leis')), 9)])
self._test('Decretos-Leis n.os 64/2006, de 21 de março, '
'88/2006, de 23 de maio, e '
'196/2006, de 10 de outubro', managers,
[(DocumentReference('64/2006', Token('Decretos-Leis')), 4),
(DocumentReference('88/2006', Token('Decretos-Leis')), 16),
(DocumentReference('196/2006', Token('Decretos-Leis')), 30)])
def test_fail(self):
managers = [ObserverManager({'\n': ClauseObserver})]
result = parser.parse('\nImagina\n', managers, {'\n'})
self.assertEqual([Token('\n'), Token('Imagina'), Token('\n')], result)
def tokenize(string, keyterms=()):
return [Token(token) for token in _tokenizer.tokenize(string, keyterms)]
def test_old_notation(self):
managers = [ObserverManager({'Decreto-Lei': DocumentRefObserver})]
self._test('Decreto-Lei nº 2-A/90,', managers,
[(DocumentReference('2-A/90', Token('Decreto-Lei')), 4)])
def replace_in(self, result):
result[self._index + 2] = Token('')
result[self._index + 1] = Annex('')
def _replace_in(self, result):
result[self._index] = EULawReference(self._number, Token(self._string))
def test_simple(self):
managers = [ObserverManager({'\n': ClauseObserver})]
result = parser.parse('\nIV\n', managers, {'\n'})
self.assertEqual([Token('\n'), Clause('IV'), Token('')], result)
def test_real(self):
self.assertEqual(
list(
tokenize('no n.º 2 do artigo 26.º do Decreto-Lei 2/2013,',
(' ', '.', ',', 'Decreto-Lei', 'Decretos-Leis', 'n.º',
'.º', 'n.os'))), [
Token('no'),
Token(' '),
Token('n.º'),
Token(' '),
Token('2'),
Token(' '),
Token('do'),
Token(' '),
Token('artigo'),
Token(' '),
Token('26'),
Token('.º'),
Token(' '),
Token('do'),
Token(' '),
Token('Decreto-Lei'),
Token(' '),
Token('2/2013'),
Token(',')
])
def replace_in(self, result):
assert (self._number_index == self._index + self.number_at)
for i in reversed(range(2, self.take_up_to + 1)):
result[self._index + i] = Token('') # '\n'
result[self._index + 1] = self.anchor_klass(self._number)
def test_simple(self):
# document 455149 contains this example
managers = parser.common_managers
result = parser.parse('\nAnexo\n', managers, {'\n'})
self.assertEqual([Token('\n'), Annex(''), Token('')], result)
def test_fails(self):
managers = [ObserverManager({'\n': UnnumberedAnnexObserver})]
result = parser.parse('\nTítulo\n', managers, {'\n'})
self.assertEqual([Token('\n'), Token('Título'), Token('\n')], result)
def test_basic(self):
self.assertEqual(
list(tokenize('the the', ' ')),
[Token('the'), Token(' '), Token('the')])