class RFC3743Parser(LGRParser):
def unicode_version(self):
# No Unicode version defined in file
return ""
def validate_document(self, schema=None):
# No validation of document done for now
return ""
def parse_document(self):
if not self.filename and isinstance(self.source, str):
self.filename = os.path.basename(self.source)
self._lgr = LGR(name=self.filename)
logger.debug('Start parsing of file: %s', self.filename)
if hasattr(self.source, "read"):
self._parse_doc(self.source)
else:
with io.open(self.source, 'r', encoding='utf-8') as rule_file:
self._parse_doc(rule_file)
return self._lgr
def _parse_doc(self, rule_file):
"""
Actual parsing of document.
:param rule_file: Content of the rule, as a file-like object.
"""
line_num = 0
for line in rule_file:
line_num += 1
line = line.strip()
if len(line) == 0:
continue
if line[0] == '#':
continue
reference = REFERENCE_RE.match(line)
if reference is not None:
ref_id = reference.group('ref_id')
value = reference.group('value')
comment = reference.group('comment')
try:
self._lgr.add_reference(value,
ref_id=ref_id,
comment=comment)
except LGRException:
logger.error("Invalid reference '%s' on line %d", line,
line_num)
continue
version = VERSION_RE.match(line)
if version is not None:
version_no = version.group('version_no')
date = version.group('date')
comment = version.group('comment')
try:
self._lgr.metadata.version = Version(version_no,
comment=comment)
self._lgr.metadata.date = date
except LGRException:
logger.error("Invalid version '%s' on line %d", line,
line_num)
continue
if UNICODE_CODEPOINT_RE.match(line) is None:
logger.debug("Skipping non-parsable line %d:\n%s", line_num,
line)
# Line is not starting with a valid unicode code point, skip
continue
# Split base character from variant(s)
char_variant = line.split(';')
char = char_variant[0]
try:
[(codepoints, references)] = parse_char(char)
self._lgr.add_cp(codepoints, ref=references)
except ValueError:
logger.error("Invalid character '%s' at line %d", char,
line_num)
except LGRException as exc:
logger.error("Cannot add code point '%s' at line %d: %s",
format_cp(codepoints), line_num, exc)
if len(char_variant) > 1:
preferred_variants = char_variant[1].strip()
if len(preferred_variants
) > 0 and preferred_variants[0] != '#':
# From RFC7940, Section 7.3. Recommended Disposition Values:
# activated The resulting string should be activated for use. (This
# is the same as a Preferred Variant [RFC3743].)
var_type = "activated"
self.insert_variant(line_num, codepoints,
preferred_variants, var_type)
if len(char_variant) > 2:
variants = char_variant[2].strip()
if len(variants) > 0 and variants[0] != '#':
self.insert_variant(line_num, codepoints, variants)
def insert_variant(self, line_num, codepoints, var, var_type=None):
try:
variants = parse_char(var)
except ValueError:
logger.error("Invalid variant '%s' at line %d", var, line_num)
return
for (var_codepoints, references) in variants:
try:
self._lgr.add_variant(codepoints,
var_codepoints,
ref=references,
variant_type=var_type)
except LGRException as exc:
logger.error(
"Cannot add variant '%s' to code point '%s' at line %d: %s",
format_cp(var_codepoints), format_cp(codepoints), line_num,
exc)
class TestLGRCore(unittest.TestCase):
def setUp(self):
unidb = IDNADatabase('6.3.0')
self.lgr = LGR(unicode_database=unidb)
def test_add_single_cp_list(self):
self.lgr.add_cp([0x0061])
self.assertIn(0x0061, self.lgr.repertoire)
def test_add_single_cp_int(self):
self.lgr.add_cp(0x0061)
self.assertIn(0x0061, self.lgr.repertoire)
def test_add_cp_sequence(self):
self.lgr.add_cp([0x0061, 0x0062])
self.assertIn([0x0061, 0x0062], self.lgr.repertoire)
self.assertNotIn(0x0061, self.lgr.repertoire)
self.assertNotIn(0x0062, self.lgr.repertoire)
def test_add_multiple_cp_sequences(self):
self.lgr.add_cp([0x0061, 0x0062])
self.lgr.add_cp([0x0061, 0x0062, 0x0063])
self.assertIn([0x0061, 0x0062], self.lgr.repertoire)
self.assertIn([0x0061, 0x0062, 0x0063], self.lgr.repertoire)
self.assertNotIn(0x0061, self.lgr.repertoire)
self.assertNotIn(0x0062, self.lgr.repertoire)
self.assertNotIn(0x0063, self.lgr.repertoire)
def test_add_cp_in_repertoire(self):
self.lgr.add_cp([0x0061])
self.assertRaises(CharAlreadyExists, self.lgr.add_cp, [0x0061])
self.assertRaises(CharAlreadyExists, self.lgr.add_cp, 0x0061)
def test_add_cp_validation(self):
validation_lgr = LGR()
validation_lgr.add_cp([0x0061])
self.lgr.add_cp([0x0061],
validating_repertoire=validation_lgr,
override_repertoire=False)
self.assertRaises(NotInRepertoire,
self.lgr.add_cp, [0x0062],
validating_repertoire=validation_lgr,
override_repertoire=False)
def test_add_cp_validation_override(self):
validation_lgr = LGR()
validation_lgr.add_cp([0x0061])
self.lgr.add_cp([0x0061],
validating_repertoire=validation_lgr,
override_repertoire=False)
self.lgr.add_cp([0x0062],
validating_repertoire=validation_lgr,
override_repertoire=True)
self.assertIn(0x0062, self.lgr.repertoire)
def test_del_single_cp_list(self):
self.lgr.add_cp(0x0061)
self.lgr.del_cp([0x0061])
self.assertNotIn(0x0061, self.lgr.repertoire)
def test_del_single_cp_int(self):
self.lgr.add_cp([0x0061])
self.lgr.del_cp(0x0061)
self.assertNotIn(0x0061, self.lgr.repertoire)
def test_del_cp_sequence(self):
self.lgr.add_cp([0x0061, 0x0062])
self.lgr.del_cp([0x0061, 0x0062])
self.assertEqual(len(self.lgr.repertoire), 0)
def test_del_cp_sequence_with_cp(self):
self.lgr.add_cp([0x0061, 0x0062])
self.assertRaises(NotInLGR, self.lgr.del_cp, 0x0061)
self.assertRaises(NotInLGR, self.lgr.del_cp, 0x0062)
self.assertIn([0x0061, 0x0062], self.lgr.repertoire)
def test_add_cp_when_not_when(self):
self.lgr.add_cp([0x0061], when='w1')
with self.assertRaises(CharInvalidContextRule) as cm:
self.lgr.add_cp([0x0062], when='w2', not_when='nw1')
the_exception = cm.exception
self.assertEqual(the_exception.cp, [0x0062])
self.lgr.add_cp([0x0062], not_when='nw2')
with self.assertRaises(CharInvalidContextRule) as cm:
self.lgr.add_cp([0x0063], when='w3', not_when='nw3')
the_exception = cm.exception
self.assertEqual(the_exception.cp, [0x0063])
def test_add_range(self):
self.lgr.add_range(0x0061, 0x007A)
for cp in range(0x0061, 0x007A + 1):
self.assertIn(cp, self.lgr.repertoire)
def test_add_range_in_repertoire(self):
self.lgr.add_range(0x0061, 0x007A)
self.assertRaises(CharAlreadyExists, self.lgr.add_range, 0x0061,
0x007A)
def test_add_range_validation(self):
validation_lgr = LGR()
for cp in range(0x0061, 0x007A + 1):
validation_lgr.add_cp(cp)
self.lgr.add_range(0x0061,
0x007A,
validating_repertoire=validation_lgr,
override_repertoire=False)
self.assertRaises(NotInRepertoire,
self.lgr.add_range,
0x00F8,
0x00FF,
validating_repertoire=validation_lgr,
override_repertoire=False)
def test_add_range_validation_with_range(self):
validation_lgr = LGR()
validation_lgr.add_range(0x0061, 0x007A)
self.lgr.add_range(0x0061,
0x007A,
validating_repertoire=validation_lgr,
override_repertoire=False)
self.assertRaises(NotInRepertoire,
self.lgr.add_range,
0x00F8,
0x00FF,
validating_repertoire=validation_lgr,
override_repertoire=False)
def test_add_range_validation_override(self):
validation_lgr = LGR()
for cp in range(0x0061, 0x007A):
validation_lgr.add_cp(cp)
self.lgr.add_range(0x0031,
0x0032,
validating_repertoire=validation_lgr,
override_repertoire=True)
self.assertIn(0x0031, self.lgr.repertoire)
def test_add_range_when_not_when(self):
self.lgr.add_range(0x0061, 0x0065, when='w1')
with self.assertRaises(RangeInvalidContextRule) as cm:
self.lgr.add_range(0x0066, 0x007A, when='w2', not_when='nw1')
the_exception = cm.exception
self.assertEqual(the_exception.first_cp, 0x0066)
self.assertEqual(the_exception.last_cp, 0x007A)
self.lgr.add_range(0x0066, 0x007A, not_when='nw2')
with self.assertRaises(RangeInvalidContextRule) as cm:
self.lgr.add_range(0x01BD, 0x01C3, when='w3', not_when='nw3')
the_exception = cm.exception
self.assertEqual(the_exception.first_cp, 0x01BD)
self.assertEqual(the_exception.last_cp, 0x01C3)
def test_expand_ranges(self):
self.lgr.add_range(0x0061, 0x007A)
for cp in range(0x0061, 0x007A + 1):
self.assertIsInstance(self.lgr.get_char(cp), RangeChar)
self.lgr.add_range(0x01BD, 0x01C3)
for cp in range(0x01BD, 0x01C3 + 1):
self.assertIsInstance(self.lgr.get_char(cp), RangeChar)
self.lgr.expand_ranges()
for cp in range(0x0061, 0x007A + 1):
char = self.lgr.get_char(cp)
self.assertIsInstance(char, Char)
self.assertNotIsInstance(char, RangeChar)
for cp in range(0x01BD, 0x01C3 + 1):
char = self.lgr.get_char(cp)
self.assertIsInstance(char, Char)
self.assertNotIsInstance(char, RangeChar)
def test_expand_range(self):
self.lgr.add_range(0x0061, 0x007A)
for cp in range(0x0061, 0x007A + 1):
self.assertIsInstance(self.lgr.get_char(cp), RangeChar)
self.lgr.expand_range(0x0061, 0x007A)
for cp in range(0x0061, 0x007A + 1):
char = self.lgr.get_char(cp)
self.assertIsInstance(char, Char)
self.assertNotIsInstance(char, RangeChar)
def test_add_variant_in_repertoire(self):
self.lgr.add_cp([0x0061])
self.lgr.add_variant([0x0061], [0x0030])
self.assertRaises(VariantAlreadyExists, self.lgr.add_variant, [0x0061],
[0x0030])
def test_add_variant_validation(self):
validation_lgr = LGR()
validation_lgr.add_cp([0x0061])
validation_lgr.add_cp([0x0030])
self.lgr.add_cp([0x0061])
self.lgr.add_variant([0x0061], [0x0030])
self.assertRaises(NotInRepertoire,
self.lgr.add_variant, [0x0061], [0x0062],
validating_repertoire=validation_lgr,
override_repertoire=False)
def test_add_variant_when_not_when(self):
self.lgr.add_cp([0x0061])
self.lgr.add_variant([0x0061], [0x0030], when='w1')
with self.assertRaises(VariantInvalidContextRule) as cm:
self.lgr.add_variant([0x0061], [0x0031], when='w2', not_when='nw1')
the_exception = cm.exception
self.assertEqual(the_exception.cp, [0x0061])
self.assertEqual(the_exception.variant, [0x0031])
self.lgr.add_variant([0x0061], [0x0030], not_when='nw2')
with self.assertRaises(VariantInvalidContextRule) as cm:
self.lgr.add_variant([0x0061], [0x0031], when='w3', not_when='nw3')
the_exception = cm.exception
self.assertEqual(the_exception.cp, [0x0061])
self.assertEqual(the_exception.variant, [0x0031])
def test_del_cp_validation_override(self):
validation_lgr = LGR()
validation_lgr.add_cp([0x0061])
validation_lgr.add_cp([0x0030])
self.lgr.add_cp([0x0061])
self.lgr.add_variant([0x0061], [0x0030])
self.lgr.add_variant([0x0061], [0x0062],
validating_repertoire=validation_lgr,
override_repertoire=True)
self.assertIn((0x0062, ), self.lgr.repertoire[0x0061]._variants)
def test_get_variants(self):
self.lgr.add_cp([0x0061])
self.lgr.add_variant([0x0061], [0x0030])
variants = self.lgr.get_variants([0x0061])
self.assertIsInstance(variants, types.GeneratorType)
variant_list = list(variants)
self.assertEqual(len(variant_list), 1)
def test_check_range_no_modification(self):
self.lgr.check_range(0x0060, 0x007F)
self.assertEqual(len(self.lgr.repertoire), 0)
def test_check_range(self):
self.lgr.add_cp([0x0061])
self.lgr.add_cp([0x007A])
codepoints = self.lgr.check_range(0x0060, 0x007F)
for result in codepoints:
cp = result[0]
prop = result[1]
if cp == 0x060 or cp >= 0x007B:
self.assertIsInstance(prop, CharInvalidIdnaProperty)
elif cp == 0x0061 or cp == 0x007A:
self.assertIsInstance(prop, CharAlreadyExists)
else:
self.assertIsNone(prop)
def test_add_codepoints(self):
self.lgr.add_codepoints([c for c in range(0x0061, 0x007A + 1)] +
[0x0107] + [0x0137, 0x0138])
expected_output = [
RangeChar(0x061, 0x0061, 0x007A),
Char(0x0107),
RangeChar(0x0137, 0x0137, 0x0138)
]
self.assertEqual(expected_output, list(self.lgr.repertoire))
def test_tags_on_codepoint(self):
self.lgr.add_cp([0x0061], tag=['t1', 't2'])
with self.assertRaises(LGRFormatException) as cm:
self.lgr.add_cp([0x0062], tag=['t1', 't1'])
the_exception = cm.exception
self.assertEqual(the_exception.reason,
LGRFormatException.LGRFormatReason.DUPLICATE_TAG)
def test_tags_on_codepoint_sequence(self):
with self.assertRaises(LGRFormatException) as cm:
self.lgr.add_cp([0x0061, 0x0062], tag=['t1'])
the_exception = cm.exception
self.assertEqual(the_exception.reason,
LGRFormatException.LGRFormatReason.SEQUENCE_NO_TAG)
def test_tags_on_range(self):
self.lgr.add_range(0x0061, 0x0062, tag=['t1', 't2'])
with self.assertRaises(LGRFormatException) as cm:
self.lgr.add_range(0x0063, 0x0064, tag=['t1', 't1'])
the_exception = cm.exception
self.assertEqual(the_exception.reason,
LGRFormatException.LGRFormatReason.DUPLICATE_TAG)
def test_list_types(self):
self.lgr.add_cp([0x0061])
self.lgr.add_variant([0x0061], [0x0030], variant_type='BLOCK')
self.lgr.add_variant([0x0061], [0x0031], variant_type='VALID')
self.lgr.add_variant([0x0061], [0x0032], variant_type='BLOCK')
self.assertEquals(self.lgr.types, set(['BLOCK', 'VALID']))
def test_del_reference(self):
ref_id_1 = self.lgr.add_reference("Test - 1")
ref_id_2 = self.lgr.add_reference("Test - 2")
self.lgr.add_cp([0x0061], ref=[ref_id_1])
self.lgr.add_cp([0x0062], ref=[ref_id_1, ref_id_2])
self.lgr.del_reference(ref_id_1)
self.assertNotIn(ref_id_1, self.lgr.reference_manager)
self.assertEquals(self.lgr.get_char([0x0061]).references, [])
self.assertEquals(self.lgr.get_char([0x0062]).references, [ref_id_2])
def test_add_cp_duplicate_reference(self):
ref_id = self.lgr.add_reference("Test - 1")
with self.assertRaises(DuplicateReference) as cm:
self.lgr.add_cp([0x0061], ref=[ref_id, ref_id])
the_exception = cm.exception
self.assertEqual(the_exception.cp, [0x0061])
def test_add_range_duplicate_reference(self):
ref_id = self.lgr.add_reference("Test - 1")
with self.assertRaises(DuplicateReference) as cm:
self.lgr.add_range(0x0061, 0x0062, ref=[ref_id, ref_id])
the_exception = cm.exception
self.assertEqual(the_exception.cp, 0x0061)
def test_add_variant_duplicate_reference(self):
self.lgr.add_cp([0x0061])
ref_id = self.lgr.add_reference("Test - 1")
with self.assertRaises(DuplicateReference) as cm:
self.lgr.add_variant([0x0061], [0x0062], ref=[ref_id, ref_id])
the_exception = cm.exception
self.assertEqual(the_exception.cp, [0x0061])
def test_generate_variants(self):
self.lgr.add_cp([0x0061])
self.lgr.add_cp([0x0062])
self.lgr.add_cp([0x0063])
self.lgr.add_cp([0x0064])
self.lgr.add_variant([0x0061], [0x0070], variant_type="type0")
self.lgr.add_variant([0x0062], [0x0071], variant_type="type1")
self.lgr.add_variant([0x0062], [0x0072], variant_type="type2")
self.assertEqual([], list(self.lgr._generate_label_variants([])))
self.assertEqual([], list(self.lgr._generate_label_variants([0x0063])))
self.assertEqual(
[], list(self.lgr._generate_label_variants([0x0063, 0x0064])))
self.assertEqual(
set([((0x0071, 0x0063), frozenset(['type1']), False),
((0x0072, 0x0063), frozenset(['type2']), False)]),
set(self.lgr._generate_label_variants([0x0062, 0x0063])))
self.assertEqual(
set([
((0x0061, 0x0062), frozenset(), False),
((0x0061, 0x0071), frozenset(['type1']), False),
((0x0061, 0x0072), frozenset(['type2']), False),
((0x0070, 0x0062), frozenset(['type0']), False),
((0x0070, 0x0071), frozenset(['type0', 'type1']), True),
((0x0070, 0x0072), frozenset(['type0', 'type2']), True),
]), set(self.lgr._generate_label_variants([0x0061, 0x0062])))
self.assertEqual(
set([
((0x0061, 0x0062, 0x0062), frozenset(), False),
((0x0061, 0x0062, 0x0071), frozenset(['type1']), False),
((0x0061, 0x0062, 0x0072), frozenset(['type2']), False),
((0x0061, 0x0071, 0x0062), frozenset(['type1']), False),
((0x0061, 0x0071, 0x0071), frozenset(['type1']), False),
((0x0061, 0x0071, 0x0072), frozenset(['type1',
'type2']), False),
((0x0061, 0x0072, 0x0062), frozenset(['type2']), False),
((0x0061, 0x0072, 0x0071), frozenset(['type1',
'type2']), False),
((0x0061, 0x0072, 0x0072), frozenset(['type2']), False),
((0x0070, 0x0062, 0x0062), frozenset(['type0']), False),
((0x0070, 0x0062, 0x0071), frozenset(['type0',
'type1']), False),
((0x0070, 0x0062, 0x0072), frozenset(['type0',
'type2']), False),
((0x0070, 0x0071, 0x0062), frozenset(['type0',
'type1']), False),
((0x0070, 0x0071, 0x0071), frozenset(['type0',
'type1']), True),
((0x0070, 0x0071, 0x0072),
frozenset(['type0', 'type1', 'type2']), True),
((0x0070, 0x0072, 0x0062), frozenset(['type0',
'type2']), False),
((0x0070, 0x0072, 0x0071),
frozenset(['type0', 'type1', 'type2']), True),
((0x0070, 0x0072, 0x0072), frozenset(['type0',
'type2']), True),
]), set(self.lgr._generate_label_variants([0x0061, 0x0062,
0x0062])))
def test_generate_variants_reflexive(self):
self.lgr.add_cp([0x0061])
self.lgr.add_cp([0x0062])
self.lgr.add_cp([0x0063])
self.lgr.add_variant([0x0062], [0x0062], variant_type="reflexive")
self.lgr.add_variant([0x0063], [0x0070], variant_type="type")
self.assertEqual([], list(self.lgr._generate_label_variants([])))
self.assertEqual([], list(self.lgr._generate_label_variants([0x0061])))
self.assertEqual([((0x0062, ), frozenset(['reflexive']), True)],
list(self.lgr._generate_label_variants([0x0062])))
self.assertEqual(
set([
((0x0062, 0x0063), frozenset(['reflexive']), False),
((0x0062, 0x0070), frozenset(['reflexive', 'type']), True),
]), set(self.lgr._generate_label_variants([0x0062, 0x0063])))
def test_label_simple(self):
self.lgr.add_cp([0x0061])
self.lgr.add_cp([0x0062, 0x0063])
self.lgr.add_range(0x0064, 0x0068)
valid_labels = ([0x0061], [0x0062, 0x0063], [0x0064], [0x0068],
[0x0061, 0x0064], [0x0061, 0x0062, 0x0063,
0x0064], [0x0062, 0x0063, 0x0068])
invalid_labels = (([0x0060], [], [0x0060]), ([0x0069], [], [0x0069]),
([0x0062], [], [0x0062]), ([0x0063], [], [0x0063]),
([0x0061, 0x0062], [0x0061], [0x0062]))
for label in valid_labels:
self.assertEqual((True, label, []),
self.lgr._test_preliminary_eligibility(label))
for (label, label_part, not_in_lgr) in invalid_labels:
self.assertEqual((False, label_part, not_in_lgr),
self.lgr._test_preliminary_eligibility(label))
def test_label_eligibility_multiple_choices(self):
self.lgr.add_cp([0x0061])
self.lgr.add_cp([0x0061, 0x0062, 0x0063])
self.lgr.add_cp([0x0064])
self.assertEqual(self.lgr._test_preliminary_eligibility([0x0062]),
(False, [], [0x0062]))
self.assertEqual(
self.lgr._test_preliminary_eligibility(
[0x0061, 0x0062, 0x0063, 0x0064]),
(True, [0x0061, 0x0062, 0x0063, 0x0064], []))
def test_label_delayed_eligibilty(self):
self.lgr.add_cp([0x0061])
self.lgr.add_variant([0x0061], [0x0061], 'block')
self.lgr.add_cp([0x0062])
self.lgr.add_variant([0x0062], [0x0062], 'invalid')
self.lgr.add_cp([0x0063, 0x0064])
self.lgr.add_variant([0x0063, 0x0064], [0x0063, 0x0064], 'invalid')
self.assertEqual(self.lgr._test_label_disposition([0x0062]),
('invalid', 0))
self.assertEqual(self.lgr._test_label_disposition([0x0063, 0x0064]),
('invalid', 0))
self.assertEqual(self.lgr._test_label_disposition([0x0061, 0x0062]),
('invalid', 0))
def test_label_length(self):
self.lgr.add_cp([0x0061])
self.lgr.add_variant([0x0061], [0x0061], 'disp')
self.lgr.add_cp([0x0062])
self.lgr.add_variant([0x0062], [0x0062], 'disp')
self.assertEqual(PROTOCOL_LABEL_MAX_LENGTH,
self.lgr.max_label_length())
for i in range(80):
self.lgr.add_variant([0x0062], [0x074D + i], 'disp')
# 41: mean number of variants per character
self.assertEqual(int(math.log(MAX_NUMBER_GENERATED_VARIANTS, 41)),
self.lgr.max_label_length())
