def tag_files():
path = "untagged/"
tokenizer = nltk.data.load('tokenizers/punkt/PY3/english.pickle')
files = FileReadingFuncts.get_files(path)
for file in files:
data = FileReadingFuncts.read_file(path, file)
#tags the times
entities_to_tag = TimeFunct.get_end_time_examples(data)
data = tag_entities(data, entities_to_tag)
#tags the named entitites
data = tag_named_entities(data)
#tags the sentences
sentences = SentenceTaggingFuncts.get_sentences(data)
entities_to_tag = get_in_dict(sentences, "sentence")
data = tag_entities(data, entities_to_tag)
#tags the paragraphs
paragraphs = SentenceTaggingFuncts.get_paragraphs(data)
entities_to_tag = get_in_dict(paragraphs, "paragraph")
data = tag_entities(data, entities_to_tag)
FileWritingFuncts.writeTaggedFile(data, file)
def read_in_tags():
tag_names = [
"sentence", "paragraph", "speaker", "location", "etime", "stime"
]
#creates a dictionary in which to store all the tag scores
tag_scores = TagScoreCalculator.create_tag_dict(tag_names)
#(I only have to read in the file nams from one directory because
#they're the same for both)
for tagFile in FileReadingFuncts.get_files("test_tagged/"):
for tag_name in tag_names:
#reads in all the tagged stuff from the tagged files
test_tags = get_all_tags("test_tagged/", tagFile, tag_name)
my_tags = get_all_tags("my_tagged/", tagFile, tag_name)
#adds the tags I read in into my tag_score dictionary
tag_scores[tag_name].add_tags(test_tags, my_tags)
return tag_scores
def createsFiles(tag_names):
#reads in the tags
path = "training/"
files = FileReadingFuncts.get_files(path)
matches = getblank2d(len(tag_names))
for file in files:
data = FileReadingFuncts.read_file(path, file)
for i in range(0, len(tag_names)):
newMatches = TagExtractingFuncts.find_tag_matches(tag_names[i], data)
matches[i] = matches[i] + (newMatches)
#writes the tags to the files
for i in range(0, len(tag_names)):
outputFile = "tagFiles/{}.txt".format(tag_names[i])
writeFile(set(matches[i]), outputFile)