def find_places_in_lyrics(lyrics, song_title, song_artist):
st = StanfordNERTagger(
'/Users/yuvalhering/Desktop/stanford-ner-2018-10-16/classifiers/english.all.3class.distsim.crf.ser.gz',
'/Users/yuvalhering/Desktop/stanford-ner-2018-10-16/stanford-ner.jar',
encoding='utf-8')
tokenized_text = word_tokenize(lyrics)
classified_text = st.tag(tokenized_text)
ner_places = []
for classification in classified_text:
if classification[1] == 'LOCATION':
ner_places.append(classification[0])
extracted_places = list(set(ner_places))
places = []
for new_place_name in extracted_places:
exists = False
for place in places:
if place.name == new_place_name:
place.add_song(song_title, song_artist)
exists = True
break
if not exists:
p = Place(new_place_name)
p.add_song(song_title, song_artist)
places.append(p)
json_places = []
for p in places:
place_dict = p.to_json()
json_places.append(place_dict)
with open('places.json', 'w', encoding='utf-8') as places_file:
json.dump(json_places, places_file, indent=2)
