From a9ba41bea6b001c02dcba74a1e984b6f2e9474db Mon Sep 17 00:00:00 2001 From: TrellixVulnTeam Date: Mon, 3 Oct 2022 04:07:36 +0000 Subject: [PATCH 1/2] Adding tarfile member sanitization to extractall() --- model-training/transcribe.py | 42 ++++++++++++++++++++++++++++++++++-- 1 file changed, 40 insertions(+), 2 deletions(-) diff --git a/model-training/transcribe.py b/model-training/transcribe.py index f4adc315..058b092e 100644 --- a/model-training/transcribe.py +++ b/model-training/transcribe.py @@ -81,7 +81,26 @@ def _asr_on_tarpaths_chunk(chunk_args): os.makedirs(tmp_dir) try: with tarfile.open(audio_tarpath) as audio_tarfile: - audio_tarfile.extractall(path=tmp_dir) + 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) + + + safe_extract(audio_tarfile, path=tmp_dir) except OSError as exc: # Skip long-named audio files if exc.errno == errno.ENAMETOOLONG: @@ -91,7 +110,26 @@ def _asr_on_tarpaths_chunk(chunk_args): m for m in members if len(m.name) <= os.pathconf("/", "PC_NAME_MAX") ] - audio_tarfile.extractall(path=tmp_dir, members=healthy_members) + 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) + + + safe_extract(audio_tarfile, path=tmp_dir, members=healthy_members) warnings.warn( f"Some audio files were dropped from {audio_tarpath} " "because they have long file names" From ab8e06520cb2198348cd72107f5a13c7910b3df1 Mon Sep 17 00:00:00 2001 From: TrellixVulnTeam Date: Tue, 4 Oct 2022 23:21:49 +0000 Subject: [PATCH 2/2] Adding numeric_owner as keyword arguement --- model-training/transcribe.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/model-training/transcribe.py b/model-training/transcribe.py index 058b092e..25472b2d 100644 --- a/model-training/transcribe.py +++ b/model-training/transcribe.py @@ -97,7 +97,7 @@ def safe_extract(tar, path=".", members=None, *, numeric_owner=False): if not is_within_directory(path, member_path): raise Exception("Attempted Path Traversal in Tar File") - tar.extractall(path, members, numeric_owner) + tar.extractall(path, members, numeric_owner=numeric_owner) safe_extract(audio_tarfile, path=tmp_dir) @@ -126,7 +126,7 @@ def safe_extract(tar, path=".", members=None, *, numeric_owner=False): if not is_within_directory(path, member_path): raise Exception("Attempted Path Traversal in Tar File") - tar.extractall(path, members, numeric_owner) + tar.extractall(path, members, numeric_owner=numeric_owner) safe_extract(audio_tarfile, path=tmp_dir, members=healthy_members)