def _get_matched_and_unmatched_contents(self):
if len(self.text_contents) == 0 or len(self.media_contents) == 0:
return {
"matched_contents": [],
"unused_contents":
self.text_contents
if len(self.text_contents) != 0 else self.media_contents,
"unused_content_type":
"text" if len(self.text_contents) != 0 else "media"
}
preprocessor = TextMediaMatchingPreprocessor(self.text_contents,
self.media_contents)
preprocessed_contents_dict = preprocessor.get_formatted_content()
text_for_matching = preprocessed_contents_dict["sentences"]
media_for_matching = preprocessed_contents_dict["media"]
unused_contents \
= preprocessed_contents_dict["content_unused_for_matching"]
unused_content_type = preprocessed_contents_dict["unused_content_type"]
matcher = TextMediaMatchingHelper(text_for_matching,
media_for_matching,
self.distance_metric_type)
matched_contents = matcher.get_text_media_matching()
return {
"matched_contents": matched_contents,
"unused_contents": unused_contents,
"unused_content_type": unused_content_type
}
def test_text_media_preprocessor_return_format():
''' Tests if the return format of the preprocessor is correct'''
preprocessor = TextMediaMatchingPreprocessor([sentence_1],
[media_related_to_sentence_1])
preprocessed_contents_dict = preprocessor.get_formatted_content()
assert isinstance(preprocessed_contents_dict, dict)
assert "sentences" in preprocessed_contents_dict
assert "media" in preprocessed_contents_dict
assert "content_unused_for_matching" in preprocessed_contents_dict
def test_text_media_preprocessor_eliminates_sentences():
''' Tests if pre-processor eliminates extra sentences'''
preprocessor = TextMediaMatchingPreprocessor([sentence_1, sentence_2],
[media_related_to_sentence_2])
preprocessed_contents_dict = preprocessor.get_formatted_content()
assert preprocessed_contents_dict["sentences"] == [sentence_2]
assert preprocessed_contents_dict["media"] == [media_related_to_sentence_2]
assert preprocessed_contents_dict["content_unused_for_matching"] == [
sentence_1
]
def test_text_media_preprocessor_prunes_list():
''' Tests the pruning functionality'''
preprocessor = TextMediaMatchingPreprocessor(
[sentence_1, sentence_2],
[media_related_to_sentence_1, media_related_to_sentence_2])
# the first 'count_content_to_eliminate' contents will be eliminated
preprocessor.count_of_content_to_eliminate = 1
filtered_list = preprocessor._get_pruned_list(
[1, 0, 2], # index 1 should be eliminated
['list_item1', 'list_item2', 'list_item3'])
assert filtered_list == ['list_item1', 'list_item3']
assert preprocessor.content_unused_for_matching == ['list_item2']
def test_text_media_preprocessor_returns_sentence_list_when_media_is_empty():
''' Sentence list is content unused
when media list is empty
'''
preprocessor = TextMediaMatchingPreprocessor([sentence_1], [])
preprocessed_contents_dict = preprocessor.get_formatted_content()
assert preprocessed_contents_dict["sentences"] == []
assert preprocessed_contents_dict["media"] == []
assert preprocessed_contents_dict["content_unused_for_matching"] == [
sentence_1
]
def test_text_media_preprocessor_returns_empty_unused_content_when_content_count_is_equal(
): # noqa
''' Unused content is empty when sentence and media
count are equal
'''
preprocessor = TextMediaMatchingPreprocessor(
[sentence_1, sentence_2],
[media_related_to_sentence_1, media_related_to_sentence_2])
preprocessed_contents_dict = preprocessor.get_formatted_content()
assert preprocessed_contents_dict["sentences"] == [sentence_2, sentence_1]
assert preprocessed_contents_dict["media"] == [
media_related_to_sentence_1, media_related_to_sentence_2
]
assert preprocessed_contents_dict["content_unused_for_matching"] == []