def test_properties(tmpdir):
p = tmpdir.join('a.smi')
p.write('properties:\n' ' bool.\n' ' + < bool.\n' ' - < bool.\n')
s = semi.load(str(p))
assert len(s.properties) == 3
assert all([x in s.properties for x in ('bool', '+', '-')])
assert len(s.properties['bool'].supertypes) == 0
assert len(s.properties['+'].supertypes) == 1
assert len(s.properties['-'].supertypes) == 1
def test_roles(tmpdir):
p = tmpdir.join('a.smi')
p.write('roles:\n' ' ARG0 : i.\n' ' CARG : string.')
s = semi.load(str(p))
assert len(s.roles) == 2
assert all([r in s.roles for r in ('ARG0', 'CARG')])
assert s.roles['ARG0'].rargname == 'ARG0'
assert s.roles['ARG0'].value == 'i'
assert s.roles['CARG'].rargname == 'CARG'
assert s.roles['CARG'].value == 'string'
def test_comments(tmpdir):
p = tmpdir.join('a.smi')
p.write('; comment\n'
'variables:\n'
' ; comment\n'
' u.\n'
' ; x < u.\n'
' i < u.\n'
' e < i : PERF bool, TENSE tense.')
s = semi.load(str(p))
assert len(s.variables) == 3
assert 'x' not in s.variables
def test_variables(tmpdir):
p = tmpdir.join('a.smi')
p.write('variables:\n'
' u.\n'
' i < u.\n'
' e < i : PERF bool, TENSE tense.')
s = semi.load(str(p))
assert len(s.variables) == 3
assert all([v in s.variables for v in 'uie'])
assert len(s.variables['u'].supertypes) == 0
assert len(s.variables['i'].supertypes) == 1
assert len(s.variables['e'].supertypes) == 1
assert s.variables['u'].properties == {}
assert s.variables['i'].properties == {}
assert 'PERF' in s.variables['e'].properties
assert 'TENSE' in s.variables['e'].properties
def test_predicates(tmpdir):
p = tmpdir.join('a.smi')
p.write('predicates:\n'
' _the_q < existential_q.\n'
' _predicate_n_1 : ARG0 x { IND + }.\n'
' _predicate_v_of : ARG0 e, ARG1 i, ARG2 p, [ ARG3 i ].\n'
' _predominant_a_1 : ARG0 e, ARG1 e.\n'
' _predominant_a_1 : ARG0 e, ARG1 p.')
s = semi.load(str(p))
assert len(s.predicates) == 4
assert all([
x in s.predicates for x in ('_the_q', '_predicate_n_1',
'_predicate_v_of', '_predominant_a_1')
])
assert len(s.predicates['_the_q'].supertypes) == 1
assert len(s.predicates['_predicate_n_1'].supertypes) == 0
assert len(s.predicates['_predicate_v_of'].supertypes) == 0
assert len(s.predicates['_predominant_a_1'].supertypes) == 0
assert len(s.predicates['_the_q'].synopses) == 0
assert len(s.predicates['_predicate_n_1'].synopses) == 1
assert len(s.predicates['_predicate_v_of'].synopses) == 1
assert len(s.predicates['_predominant_a_1'].synopses) == 2
def main():
args = docopt.docopt(USAGE)
if args['--sem-i']:
smi = semi.load(args['--sem-i'])
else:
smi = semi.SemI()
lexmap = {}
if args['--lexicon']:
for lexicon in args['--lexicon']:
lexmap.update(get_lexmap(lexicon))
typemap = json.load(open(args['TYPEMAP']))
entry_generators = []
for prof in args['PROFILE']:
table, cols = itsdb.get_data_specifier(args['--data-specifier'])
entry_generators.append(
prof_entries(prof, typemap, lexmap, table, cols))
for mtr in args['--mtr']:
entry_generators.append(mtr_entries(mtr, typemap))
entries = chain.from_iterable(entry_generators)
created = set()
for (lename, supertype, orth, pred, cv) in entries:
# don't create if pred exists or entry already created
if (pred.short_form() in smi.predicates
or (supertype, orth, pred.string) in created):
continue
# guess if phonological onset is vocalic or consonant
if cv is None:
cv = 'voc' if pred.lemma[0] in 'aeiou' else 'con'
print(format_lex_entry(lename, supertype, orth, pred, cv))
created.add((supertype, orth, pred.string))
def test_include(tmpdir):
a = tmpdir.join('a.smi')
b = tmpdir.join('b.smi')
tmpdir.mkdir('sub')
c = tmpdir.join('sub', 'c.smi')
d = tmpdir.join('sub', 'd.smi')
a.write('predicates:\n'
' abstract_q : ARG0 x, RSTR h, BODY h.\n'
' can_able.\n'
' _able_a_1 < can_able.\n'
'include: b.smi\n'
'include: sub/c.smi')
b.write('predicates:\n'
' existential_q < abstract_q.\n'
' _able_a_1 : ARG0 e, ARG1 p.')
c.write('predicates:\n'
' universal_q < abstract_q\n'
' _able_a_1 : ARG0 e, ARG1 i, ARG2 h.\n'
'include: d.smi')
d.write('variables:\n'
' u.\n'
' i < u.\n'
'properties:\n'
' tense.\n'
' pres < tense.\n'
'roles:\n'
' ARG0 : i.')
s = semi.load(str(a))
assert len(s.variables) == 2
assert len(s.properties) == 2
assert len(s.roles) == 1
assert 'abstract_q' in s.predicates
assert 'existential_q' in s.predicates
assert 'can_able' in s.predicates
assert '_able_a_1' in s.predicates
assert 'can_able' in s.predicates['_able_a_1'].supertypes
assert len(s.predicates['_able_a_1'].synopses) == 2
def test_from_dict(tmpdir):
p = tmpdir.join('a.smi')
p.write('variables:\n'
' e < i : TENSE tense.\n'
'properties:\n'
' tense.\n'
' pres < tense.\n'
'roles:\n'
' ARG0 : i.\n'
' ARG1 : u.\n'
'predicates:\n'
' _the_q < existential_q.\n'
' _predicate_n_1 : ARG0 x { IND + }.\n'
' _predicate_v_of : ARG0 e, ARG1 i, ARG2 p, [ ARG3 i ].\n'
' _predominant_a_1 : ARG0 e, ARG1 e.\n'
' _predominant_a_1 : ARG0 e, ARG1 p.')
s1 = semi.load(str(p))
s2 = semi.SemI.from_dict({
'variables': {
'e': {
'type': 'e',
'parents': ['i'],
'properties': {
'TENSE': 'tense'
}
}
},
'properties': {
'tense': {
'type': 'tense',
'parents': []
},
'pres': {
'type': 'pres',
'parents': ['tense']
}
},
'roles': {
'ARG0': {
'rargname': 'ARG0',
'value': 'i'
},
'ARG1': {
'rargname': 'ARG1',
'value': 'u'
}
},
'predicates': {
'_the_q': {
'predicate': '_the_q',
'parents': ['existential_q'],
'synopses': []
},
'_predicate_n_1': {
'predicate':
'_predicate_n_1',
'parents': [],
'synopses': [[{
'rargname': 'ARG0',
'value': 'x',
'properties': {
'IND': '+'
}
}]]
},
'_predicate_v_of': {
'predicate':
'_predicate_v_of',
'parents': [],
'synopses': [[{
'rargname': 'ARG0',
'value': 'e'
}, {
'rargname': 'ARG1',
'value': 'i'
}, {
'rargname': 'ARG2',
'value': 'p'
}, {
'rargname': 'ARG3',
'value': 'i',
'optional': True
}]]
},
'_predominant_a_1': {
'predicate':
'_predominant_a_1',
'parents': [],
'synopses': [[{
'rargname': 'ARG0',
'value': 'e'
}, {
'rargname': 'ARG1',
'value': 'e'
}],
[{
'rargname': 'ARG0',
'value': 'e'
}, {
'rargname': 'ARG1',
'value': 'p'
}]]
}
}
})
assert s1.variables == s2.variables
assert s1.properties == s2.properties
assert s1.roles == s2.roles
assert s1.predicates == s2.predicates