def test_pattern_graph_structure1(self):
pattern = pattern_pfx + '''
<pattern name="ex" class="ex-patterns">
<word pos="JJ" min="1"/>
<word pos="IN" min="0"/>
<word pos="NN" min="1" max="2"/>
</pattern>'''
tree = etree.fromstring(pattern)
# print('\n') # print debug information
# pattern = Pattern(tree, True)
pattern = Pattern(tree)
self.assertTrue(pattern.graph.is_directed())
starting = pattern.graph_entry_points[0]
successor = None
for x in pattern.graph.successors(starting):
successor = x
break
successors2 = []
for node in pattern.graph.successors(successor):
successors2.append(node._pos)
self.assertTrue(successors2[0] == 'NN')
self.assertTrue(successors2[1] == 'NN')
def test_pattern_graph_exit_points2(self):
pattern = pattern_pfx + '''
<pattern name="ex" class="ex-patterns">
<word pos="JJ" min="1"/>
<word pos="NN" min="1" max="2"/>
</pattern>'''
tree = etree.fromstring(pattern)
pattern = Pattern(tree)
is_correct_length = len(pattern.graph_exit_points) == 1
contains_node1 = pattern.graph_exit_points[0]._pos == 'NN'
self.assertTrue(is_correct_length)
self.assertTrue(contains_node1)
def test_pattern_graph_matches5(self):
pattern = pattern_pfx + '''
<pattern name="ex" class="ex-patterns">
<word lemma="word*" />
<word word="abcd" />
</pattern>'''
tree = etree.fromstring(pattern)
pattern = Pattern(tree)
sentence = "He is running the race by eating the the mungo."
annotator = BasicNltkAnnotator()
annotated_sentence = annotator.annotate(sentence)
bfs_search = BreadthFirstWithQueue()
matches = MatchBuilder.find_all_matches(annotated_sentence, pattern,
bfs_search)
self.assertEqual(len(matches), 0)
def test_pattern_graph_matches4(self):
pattern = pattern_pfx + '''
<pattern name="ex" class="ex-patterns">
<word pos="VB*" />
<word pos="DT" min="2"/>
</pattern>'''
tree = etree.fromstring(pattern)
pattern = Pattern(tree)
sentence = "He is running the race by eating the the mungo."
annotator = BasicNltkAnnotator()
annotated_sentence = annotator.annotate(sentence)
bfs_search = BreadthFirstWithQueue()
matches = MatchBuilder.find_all_matches(annotated_sentence, pattern,
bfs_search)
actual = ' '.join([x.word for x in matches[0]])
expected = 'eating the the'
self.assertEqual(actual, expected)
def test_pattern_graph_successors1(self):
pattern = pattern_pfx + '''
<pattern name="ex" class="ex-patterns">
<word pos="JJ" min="1"/>
<word pos="IN" min="0"/>
<word pos="NN" min="1" max="2"/>
</pattern>'''
tree = etree.fromstring(pattern)
pattern = Pattern(tree)
self.assertTrue(pattern.graph.is_directed())
# starting_node = pattern.graph.nodes(data=pattern.graph_entry_points[0])
starting = pattern.graph_entry_points[0]
successors = []
for succ in pattern.graph.successors(starting):
successors.append(succ._pos)
self.assertTrue(starting._pos == 'JJ')
self.assertTrue(successors[0] == 'IN')
self.assertTrue(successors[1] == 'NN')
def test_pattern_graph_matches3(self):
pattern = pattern_pfx + '''
<pattern name="ex" class="ex-patterns">
<word pos="DT" min="0"/>
<word pos="NN*" />
</pattern>'''
tree = etree.fromstring(pattern)
pattern = Pattern(tree)
sentence = "This is the dog who barks at moons."
annotator = BasicNltkAnnotator()
annotated_sentence = annotator.annotate(sentence)
bfs_search = BreadthFirstWithQueue()
matches = MatchBuilder.find_all_matches(annotated_sentence, pattern,
bfs_search)
actual1 = ' '.join([x.word for x in matches[0]])
self.assertEqual(actual1, 'the dog')
actual2 = ' '.join([x.word for x in matches[1]])
self.assertEqual(actual2, 'dog')
actual3 = ' '.join([x.word for x in matches[2]])
self.assertEqual(actual3, 'moons')
def test_pattern_graph_matches2(self):
pattern = pattern_pfx + '''
<pattern name="ex" class="ex-patterns">
<word pos="JJ*" />
<word pos="NN*" />
</pattern>'''
tree = etree.fromstring(pattern)
pattern = Pattern(tree)
sentence = "He has a big dog that lives in the green house."
annotator = BasicNltkAnnotator()
annotated_sentence = annotator.annotate(sentence)
bfs_search = BreadthFirstWithQueue()
matches = MatchBuilder.find_all_matches(annotated_sentence, pattern,
bfs_search)
actual1 = ' '.join([x.word for x in matches[0]])
expected1 = 'big dog'
actual2 = ' '.join([x.word for x in matches[1]])
expected2 = 'green house'
self.assertEqual(actual1, expected1)
self.assertEqual(actual2, expected2)