def file_iterator(benchmark_name: str = "correct",
split: str = "training") -> Iterator[str]:
"""Iterates over the paragraph files of a given benchmark and partition.
:param benchmark_name: name of the benchmark, equals the benchmark folder name
:param split: name of the partition, either training, development or test
:return: iterator over file paths, relative from the Wikipedia directory defined in src.settings.paths
"""
dir = split + "/" + benchmark_name + "/texts/"
subdirs = sorted(get_files(paths.WIKI_DIR + dir))
for subdir in subdirs:
files = sorted(get_files(paths.WIKI_DIR + dir + subdir))
for file in files:
yield dir + subdir + "/" + file
def get_files_depth_two(directory):
"""
Returns the paths to all files from the directories in the given directory.
:param directory: a directory with directories containing the wanted files
:return: list of full paths to all files at depth two
"""
subdirs = get_files(directory)
files = []
for subdir in sorted(subdirs):
path = directory + "/" + subdir + "/"
subdir_files = get_files(path)
for file in sorted(subdir_files):
files.append(path + file)
return files
def _remove_saved_models(self):
files = get_files(self.model_dir())
saved_models = [
f for f in files
if file_exists(self.model_dir() + "/" + f + "/saved_model.pb")
]
for model_name in saved_models:
remove_dir(self.model_dir() + "/" + model_name)
def latest_saved_model_dir(path):
"""
Finds the most recent timestamped subfolder at the given path.
Assumes folders are named by a timestamp.
:param path: path containing at least one timestamped subfolder
:return: string consisting of path/latest_timestamped_subfolder, or None if no saved model exists
"""
files = get_files(path)
saved_model_dirs = [
f for f in files
if "temp" not in f and file_exists(path + "/" + f + "/saved_model.pb")
]
if len(saved_model_dirs) == 0:
return None
latest = sorted([int(model_dir) for model_dir in saved_model_dirs])[-1]
return path + "/" + str(latest)
import project
from src.helper.files import get_files, read_lines
from select_acl_articles import get_year
def preprocess(line: str):
tokens = [token for token in line.split() if len(token) > 0]
line = " ".join(tokens)
return line
if __name__ == "__main__":
random.seed(42)
acl_dir = "/home/hertel/tokenization-repair-dumps/nastase/acl-201302_word-resegmented/raw/"
files = sorted(get_files(acl_dir))
files = [file for file in files if get_year(file) >= 2005]
examples = []
for file in files:
lines = read_lines(acl_dir + file)
lines = [preprocess(line) for line in lines]
lines = [line for line in lines if len(line) > 0]
examples.extend(lines)
random.shuffle(examples)
for line in examples:
print(line)
import project
from src.settings import paths
from src.helper.files import get_files
from src.helper.pickle import load_object
if __name__ == "__main__":
for file in sorted(get_files(paths.THRESHOLD_FITTER_DIR)):
fitter = load_object(paths.THRESHOLD_FITTER_DIR + file)
print(file, fitter.n_sequences)
def _remove_training_split_files():
for file in get_files(paths.WIKI_TRAINING_SPLIT_DIR):
remove_file(paths.WIKI_TRAINING_SPLIT_DIR + file)
tuning_ids = set(random.sample(training_ids, 10000))
training_ids = set(training_ids)
development_ids = set(development_ids)
test_ids = set(test_ids)
evaluation_ids = development_ids.union(test_ids).union(tuning_ids)
tuning_sentences, development_sentences, test_sentences = [], [], []
sentence_splitter = WikiPunktTokenizer()
if TRAINING:
training_file = open(paths.WIKI_TRAINING_SENTENCES,
'w',
encoding="utf8")
for sub_dir in sorted(get_files(base_dir)):
print(sub_dir)
for file in sorted(get_files(base_dir + sub_dir)):
path = base_dir + sub_dir + "/" + file
for line in read_sequences(path):
article = json.loads(line)
id = article["id"]
if not TRAINING and id in evaluation_ids:
sentences = split_article(article["text"],
sentence_splitter)
sentences = filter_sentences(sentences)
if len(sentences) > 0:
selected_sentence = random.choice(sentences)
selected_sentence = preprocess_sentence(
selected_sentence)
if id in tuning_ids:
def get_all_prediction_files(self) -> List[str]:
files = get_files(self.results_dir())
filtered = [file for file in files if file.endswith(".txt")]
return filtered
import random
import shutil
import project
from src.helper.files import get_files, read_lines, write_lines
from src.settings import paths, symbols
from src.helper.pickle import dump_object
if __name__ == "__main__":
random.seed(42)
step = sys.argv[1]
if step == "split":
path = "/home/hertel/tokenization-repair-dumps/nastase/acl-201302_word-resegmented/raw/"
files = sorted(get_files(path))
print(len(files), "files")
random.shuffle(files)
n_test = 100
out_path = "/home/hertel/tokenization-repair-dumps/acl_corpus/"
for i, filename in enumerate(files):
print(filename)
if i < n_test:
subdir = "development/"
elif i < 2 * n_test:
subdir = "test/"
else:
subdir = "training/"
shutil.copy(path + filename, out_path + subdir + filename)
