schema_utils.py

# Copyright 2014 The Oppia Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#      http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS-IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

"""Utility functions for managing schemas and schema-based validation.

A schema is a way to specify the type of an object. For example, one might
want to require that an object is an integer, or that it is a dict with two
keys named 'abc' and 'def', each with values that are unicode strings. This
file contains utilities for validating schemas and for checking that objects
follow the definitions given by the schemas.

The objects that can be described by these schemas must be composable from the
following Python types: bool, dict, float, int, list, unicode.
"""

import numbers
import re
import urllib
import urlparse

from core.domain import html_cleaner  # pylint: disable=relative-import

SCHEMA_KEY_ITEMS = 'items'
SCHEMA_KEY_LEN = 'len'
SCHEMA_KEY_PROPERTIES = 'properties'
SCHEMA_KEY_TYPE = 'type'
SCHEMA_KEY_POST_NORMALIZERS = 'post_normalizers'
SCHEMA_KEY_CHOICES = 'choices'
SCHEMA_KEY_NAME = 'name'
SCHEMA_KEY_SCHEMA = 'schema'
SCHEMA_KEY_OBJ_TYPE = 'obj_type'
SCHEMA_KEY_VALIDATORS = 'validators'

SCHEMA_TYPE_BOOL = 'bool'
SCHEMA_TYPE_CUSTOM = 'custom'
SCHEMA_TYPE_DICT = 'dict'
SCHEMA_TYPE_FLOAT = 'float'
SCHEMA_TYPE_HTML = 'html'
SCHEMA_TYPE_INT = 'int'
SCHEMA_TYPE_LIST = 'list'
SCHEMA_TYPE_UNICODE = 'unicode'


def normalize_against_schema(obj, schema, apply_custom_validators=True):
    """Validate the given object using the schema, normalizing if necessary.

    Returns:
        the normalized object.

    Raises:
        AssertionError: if the object fails to validate against the schema.
    """
    normalized_obj = None

    if schema[SCHEMA_KEY_TYPE] == SCHEMA_TYPE_BOOL:
        assert isinstance(obj, bool), ('Expected bool, received %s' % obj)
        normalized_obj = obj
    elif schema[SCHEMA_KEY_TYPE] == SCHEMA_TYPE_CUSTOM:
        # Importing this at the top of the file causes a circular dependency.
        # TODO(sll): Either get rid of custom objects or find a way to merge
        # them into the schema framework -- probably the latter.
        from core.domain import obj_services  # pylint: disable=relative-import
        obj_class = obj_services.Registry.get_object_class_by_type(
            schema[SCHEMA_KEY_OBJ_TYPE])
        normalized_obj = obj_class.normalize(obj)
    elif schema[SCHEMA_KEY_TYPE] == SCHEMA_TYPE_DICT:
        assert isinstance(obj, dict), ('Expected dict, received %s' % obj)
        expected_dict_keys = [
            p[SCHEMA_KEY_NAME] for p in schema[SCHEMA_KEY_PROPERTIES]]
        assert set(obj.keys()) == set(expected_dict_keys), (
            'Missing keys: %s, Extra keys: %s' % (
                list(set(expected_dict_keys) - set(obj.keys())),
                list(set(obj.keys()) - set(expected_dict_keys))))

        normalized_obj = {}
        for prop in schema[SCHEMA_KEY_PROPERTIES]:
            key = prop[SCHEMA_KEY_NAME]
            normalized_obj[key] = normalize_against_schema(
                obj[key], prop[SCHEMA_KEY_SCHEMA])
    elif schema[SCHEMA_KEY_TYPE] == SCHEMA_TYPE_FLOAT:
        obj = float(obj)
        assert isinstance(obj, numbers.Real), (
            'Expected float, received %s' % obj)
        normalized_obj = obj
    elif schema[SCHEMA_KEY_TYPE] == SCHEMA_TYPE_INT:
        obj = int(obj)
        assert isinstance(obj, numbers.Integral), (
            'Expected int, received %s' % obj)
        assert isinstance(obj, int), ('Expected int, received %s' % obj)
        normalized_obj = obj
    elif schema[SCHEMA_KEY_TYPE] == SCHEMA_TYPE_HTML:
        assert isinstance(obj, basestring), (
            'Expected unicode HTML string, received %s' % obj)
        obj = unicode(obj)
        assert isinstance(obj, unicode), (
            'Expected unicode, received %s' % obj)
        normalized_obj = html_cleaner.clean(obj)
    elif schema[SCHEMA_KEY_TYPE] == SCHEMA_TYPE_LIST:
        assert isinstance(obj, list), ('Expected list, received %s' % obj)
        item_schema = schema[SCHEMA_KEY_ITEMS]
        if SCHEMA_KEY_LEN in schema:
            assert len(obj) == schema[SCHEMA_KEY_LEN]
        normalized_obj = [
            normalize_against_schema(item, item_schema) for item in obj
        ]
    elif schema[SCHEMA_KEY_TYPE] == SCHEMA_TYPE_UNICODE:
        assert isinstance(obj, basestring), (
            'Expected unicode string, received %s' % obj)
        obj = unicode(obj)
        assert isinstance(obj, unicode), (
            'Expected unicode, received %s' % obj)
        normalized_obj = obj
    else:
        raise Exception('Invalid schema type: %s' % schema[SCHEMA_KEY_TYPE])

    if SCHEMA_KEY_CHOICES in schema:
        assert normalized_obj in schema[SCHEMA_KEY_CHOICES], (
            'Received %s which is not in the allowed range of choices: %s' %
            (normalized_obj, schema[SCHEMA_KEY_CHOICES]))

    # When type normalization is finished, apply the post-normalizers in the
    # given order.
    if SCHEMA_KEY_POST_NORMALIZERS in schema:
        for normalizer in schema[SCHEMA_KEY_POST_NORMALIZERS]:
            kwargs = dict(normalizer)
            del kwargs['id']
            normalized_obj = Normalizers.get(normalizer['id'])(
                normalized_obj, **kwargs)

    # Validate the normalized object.
    if apply_custom_validators:
        if SCHEMA_KEY_VALIDATORS in schema:
            for validator in schema[SCHEMA_KEY_VALIDATORS]:
                kwargs = dict(validator)
                del kwargs['id']
                assert _Validators.get(
                    validator['id'])(normalized_obj, **kwargs), (
                        'Validation failed: %s (%s) for object %s' % (
                            validator['id'], kwargs, normalized_obj))

    return normalized_obj


class Normalizers(object):
    """Various normalizers.

    A normalizer is a function that takes an object, attempts to normalize
    it to a canonical representation, and/or performs validity checks on the
    object pre- and post-normalization. If the normalization succeeds, the
    function returns the transformed object; if it fails, it raises an
    exception.

    Some normalizers require additional arguments. It is the responsibility of
    callers of normalizer functions to ensure that the arguments they supply to
    the normalizer are valid. What exactly this entails is provided in the
    docstring for each normalizer.
    """

    @classmethod
    def get(cls, normalizer_id):
        if not hasattr(cls, normalizer_id):
            raise Exception('Invalid normalizer id: %s' % normalizer_id)
        return getattr(cls, normalizer_id)

    @staticmethod
    def normalize_spaces(obj):
        """Collapses multiple spaces into single spaces.

        Args:
          obj: a string.

        Returns:
          a string that is the same as `obj`, except that each block of
          whitespace is collapsed into a single space character.
        """
        return ' '.join(obj.split())

    @staticmethod
    def sanitize_url(obj):
        """Takes a string representing a URL and sanitizes it.

        Args:
          obj: a string representing a URL.

        Returns:
          An empty string if the URL does not start with http:// or https://.
          Otherwise, returns the original URL.
        """
        url_components = urlparse.urlsplit(obj)
        quoted_url_components = (
            urllib.quote(component) for component in url_components)
        raw = urlparse.urlunsplit(quoted_url_components)

        acceptable = html_cleaner.filter_a('href', obj)
        assert acceptable, (
            'Invalid URL: Sanitized URL should start with '
            '\'http://\' or \'https://\'; received %s' % raw)
        return raw


class _Validators(object):
    """Various validators.

    A validator is a function that takes an object and returns True if it is
    valid, and False if it isn't.

    Validators should only be accessed from the checker methods in
    schema_utils.py and schema_utils_test.py, since these methods do
    preliminary checks on the arguments passed to the validator.
    """
    @classmethod
    def get(cls, validator_id):
        if not hasattr(cls, validator_id):
            raise Exception('Invalid validator id: %s' % validator_id)
        return getattr(cls, validator_id)

    @staticmethod
    def has_length_at_least(obj, min_value):
        """Returns True iff the given object (a list) has at least
        `min_value` elements.
        """
        return len(obj) >= min_value

    @staticmethod
    def has_length_at_most(obj, max_value):
        """Returns True iff the given object (a list) has at most
        `max_value` elements.
        """
        return len(obj) <= max_value

    @staticmethod
    def is_nonempty(obj):
        """Returns True iff the given object (a string) is nonempty."""
        return bool(obj)

    @staticmethod
    def is_uniquified(obj):
        """Returns True iff the given object (a list) has no duplicates."""
        return sorted(list(set(obj))) == sorted(obj)

    @staticmethod
    def is_at_least(obj, min_value):
        """Ensures that `obj` (an int/float) is at least `min_value`."""
        return obj >= min_value

    @staticmethod
    def is_at_most(obj, max_value):
        """Ensures that `obj` (an int/float) is at most `max_value`."""
        return obj <= max_value

    @staticmethod
    def is_regex(obj):
        """Ensures that `obj` (a string) defines a valid regex."""
        raise NotImplementedError

    @staticmethod
    def matches_regex(obj, regex):
        """Ensures that `obj` (a string) matches the given regex."""
        raise NotImplementedError

    @staticmethod
    def is_valid_email(obj):
        """Ensures that `obj` (a string) is a valid email."""
        return bool(re.search(r"^[\w\.\+\-]+\@[\w]+\.[a-z]{2,3}$", obj))