Update pyyaml to prevent arbitrary code execution

Before `pyyaml==5.1` the `yaml.load` function was vulnerable to
arbitrary code execution, because it loaded the full set of YAML. There
was an alternative `safe_load` but this was not the default and could
only load a sub set of the markup language. The new version of pyyaml
deprecates the old vulnerable code and provides the `FullLoader` that
can load the full set without being vulnerable.

aiida/cmdline/params/options/config.py

@@ -12,17 +12,16 @@
 .. py:module::config
     :synopsis: Convenience class for configuration file option
 """
-
-import yaml
 import click_config_file
+import yaml
 
 from .overridable import OverridableOption
 
 
 def yaml_config_file_provider(file_path, cmd_name):  # pylint: disable=unused-argument
     """Read yaml config file."""
     with open(file_path, 'r') as handle:
-        return yaml.load(handle)
+        return yaml.safe_load(handle)
 
 
 class ConfigFileOption(OverridableOption):

aiida/manage/caching.py

@@ -8,7 +8,6 @@
 # For further information please visit http://www.aiida.net               #
 ###########################################################################
 """Definition of caching mechanism and configuration for calculations."""
-
 import os
 import copy
 import warnings
@@ -55,7 +54,7 @@ def _get_config(config_file):
 
     try:
         with open(config_file, 'r', encoding='utf8') as handle:
-            config = yaml.load(handle)[profile.name]
+            config = yaml.safe_load(handle)[profile.name]
     except (OSError, IOError, KeyError):
         # No config file, or no config for this profile
         return DEFAULT_CONFIG

aiida/orm/utils/serialize.py

@@ -14,7 +14,6 @@
 checkpoints and messages in the RabbitMQ queue so do so with caution.  It is fine to add representers
 for new types though.
 """
-
 from functools import partial
 import yaml
 
@@ -177,8 +176,14 @@ def represent_data(self, data):
         return super().represent_data(data)
 
 
-class AiiDALoader(yaml.Loader):
-    """AiiDA specific yaml loader"""
+class AiiDALoader(yaml.FullLoader):
+    """AiiDA specific yaml loader
+
+    .. note:: we subclass the `FullLoader` which is the one that since `pyyaml>=5.1` is the loader that prevents
+        arbitrary code execution. Even though this is in principle only used internally, one could imagine someone
+        sharing a database with a maliciously crafted process instance dump, which when reloaded could execute arbitrary
+        code. This load prevents this: https://github.com/yaml/pyyaml/wiki/PyYAML-yaml.load(input)-Deprecation
+    """
 
 
 yaml.add_representer(Bundle, represent_bundle, Dumper=AiiDADumper)
@@ -217,6 +222,8 @@ def serialize(data, encoding=None):
 def deserialize(serialized):
     """Deserialize a yaml dump that represents a serialized data structure.
 
+    .. note:: no need to use `yaml.safe_load` here because the `Loader` will ensure that loading is safe.
+
     :param serialized: a yaml serialized string representation
     :return: the deserialized data structure
     """

docs/requirements_for_rtd.txt

@@ -35,7 +35,7 @@ pytest~=5.3
 python-dateutil~=2.8
 python-memcached~=1.59
 pytz~=2019.3
-pyyaml~=3.13
+pyyaml~=5.1
 reentry~=1.3
 seekpath~=1.9,>=1.9.3
 simplejson~=3.16

docs/source/nitpick-exceptions

@@ -242,6 +242,8 @@ py:class yaml.Dumper
 py:class yaml.Loader
 py:class yaml.dumper.Dumper
 py:class yaml.loader.Loader
+py:class yaml.FullLoader
+py:class yaml.loader.FullLoader
 
 py:class uuid.UUID
 

environment.yml

@@ -27,7 +27,7 @@ dependencies:
 - psycopg2~=2.8,>=2.8.3
 - python-dateutil~=2.8
 - pytz~=2019.3
-- pyyaml~=3.13
+- pyyaml~=5.1
 - reentry~=1.3
 - simplejson~=3.16
 - sqlalchemy-utils~=0.34.2

setup.json

@@ -41,7 +41,7 @@
     "pyblake2~=1.1; python_version<'3.6'",
     "python-dateutil~=2.8",
     "pytz~=2019.3",
-    "pyyaml~=3.13",
+    "pyyaml~=5.1",
     "reentry~=1.3",
     "simplejson~=3.16",
     "sqlalchemy-utils~=0.34.2",