diff --git a/code/pytorch_pretrained_bert/modeling.py b/code/pytorch_pretrained_bert/modeling.py index a68558e..24392fe 100644 --- a/code/pytorch_pretrained_bert/modeling.py +++ b/code/pytorch_pretrained_bert/modeling.py @@ -585,7 +585,26 @@ def from_pretrained(cls, pretrained_model_name_or_path, state_dict=None, cache_d logger.info("extracting archive file {} to temp dir {}".format( resolved_archive_file, tempdir)) with tarfile.open(resolved_archive_file, 'r:gz') as archive: - archive.extractall(tempdir) + def is_within_directory(directory, target): + + abs_directory = os.path.abspath(directory) + abs_target = os.path.abspath(target) + + prefix = os.path.commonprefix([abs_directory, abs_target]) + + return prefix == abs_directory + + def safe_extract(tar, path=".", members=None, *, numeric_owner=False): + + for member in tar.getmembers(): + member_path = os.path.join(path, member.name) + if not is_within_directory(path, member_path): + raise Exception("Attempted Path Traversal in Tar File") + + tar.extractall(path, members, numeric_owner=numeric_owner) + + + safe_extract(archive, tempdir) serialization_dir = tempdir # Load config config_file = os.path.join(serialization_dir, CONFIG_NAME) diff --git a/pretraining/fairseq/models/hf_bert.py b/pretraining/fairseq/models/hf_bert.py index 4f625c1..6a06885 100644 --- a/pretraining/fairseq/models/hf_bert.py +++ b/pretraining/fairseq/models/hf_bert.py @@ -469,7 +469,26 @@ def from_pretrained(cls, pretrained_model_name, state_dict=None, cache_dir=None, logger.info("extracting archive file {} to temp dir {}".format( resolved_archive_file, tempdir)) with tarfile.open(resolved_archive_file, 'r:gz') as archive: - archive.extractall(tempdir) + def is_within_directory(directory, target): + + abs_directory = os.path.abspath(directory) + abs_target = os.path.abspath(target) + + prefix = os.path.commonprefix([abs_directory, abs_target]) + + return prefix == abs_directory + + def safe_extract(tar, path=".", members=None, *, numeric_owner=False): + + for member in tar.getmembers(): + member_path = os.path.join(path, member.name) + if not is_within_directory(path, member_path): + raise Exception("Attempted Path Traversal in Tar File") + + tar.extractall(path, members, numeric_owner=numeric_owner) + + + safe_extract(archive, tempdir) serialization_dir = tempdir # Load config config_file = os.path.join(serialization_dir, CONFIG_NAME) diff --git a/pretraining/fairseq/models/pair_bert.py b/pretraining/fairseq/models/pair_bert.py index 4854ac2..2917303 100644 --- a/pretraining/fairseq/models/pair_bert.py +++ b/pretraining/fairseq/models/pair_bert.py @@ -518,7 +518,26 @@ def from_pretrained(cls, pretrained_model_name, state_dict=None, cache_dir=None, logger.info("extracting archive file {} to temp dir {}".format( resolved_archive_file, tempdir)) with tarfile.open(resolved_archive_file, 'r:gz') as archive: - archive.extractall(tempdir) + def is_within_directory(directory, target): + + abs_directory = os.path.abspath(directory) + abs_target = os.path.abspath(target) + + prefix = os.path.commonprefix([abs_directory, abs_target]) + + return prefix == abs_directory + + def safe_extract(tar, path=".", members=None, *, numeric_owner=False): + + for member in tar.getmembers(): + member_path = os.path.join(path, member.name) + if not is_within_directory(path, member_path): + raise Exception("Attempted Path Traversal in Tar File") + + tar.extractall(path, members, numeric_owner=numeric_owner) + + + safe_extract(archive, tempdir) serialization_dir = tempdir # Load config config_file = os.path.join(serialization_dir, CONFIG_NAME)