diff options
Diffstat (limited to 'user_guide_src/source')
| -rw-r--r-- | user_guide_src/source/changelog.rst | 3 | ||||
| -rw-r--r-- | user_guide_src/source/libraries/security.rst | 17 |
2 files changed, 18 insertions, 2 deletions
diff --git a/user_guide_src/source/changelog.rst b/user_guide_src/source/changelog.rst index 5c233efac..64a768977 100644 --- a/user_guide_src/source/changelog.rst +++ b/user_guide_src/source/changelog.rst @@ -508,9 +508,10 @@ Release Date: Not Released - :doc:`Security Library <libraries/security>` changes include: - - Added method ``strip_image_tags()``. - Added ``$config['csrf_regeneration']``, which makes CSRF token regeneration optional. - Added ``$config['csrf_exclude_uris']``, allowing for exclusion of URIs from the CSRF protection (regular expressions are supported). + - Added method ``strip_image_tags()``. + - Added method ``get_random_bytes()`` and switched CSRF & XSS token generation to use it. - Modified method ``sanitize_filename()`` to read a public ``$filename_bad_chars`` property for getting the invalid characters list. - Return status code of 403 instead of a 500 if CSRF protection is enabled but a token is missing from a request. diff --git a/user_guide_src/source/libraries/security.rst b/user_guide_src/source/libraries/security.rst index c8d69d16f..0c51e342b 100644 --- a/user_guide_src/source/libraries/security.rst +++ b/user_guide_src/source/libraries/security.rst @@ -163,4 +163,19 @@ Class Reference This method acts a lot like PHP's own native ``html_entity_decode()`` function in ENT_COMPAT mode, only it tries to detect HTML entities that don't end in a semicolon because some browsers allow that. - If the ``$charset`` parameter is left empty, then your configured ``$config['charset']`` value will be used.
\ No newline at end of file + If the ``$charset`` parameter is left empty, then your configured ``$config['charset']`` value will be used. + + .. method:: get_random_bytes($length) + + :param int $length: Output length + :returns: A binary stream of random bytes or FALSE on failure + :rtype: string + + A convenience method for getting proper random bytes via ``mcrypt_create_iv()``, + ``/dev/urandom`` or ``openssl_random_pseudo_bytes()`` (in that order), if one + of them is available. + + Used for generating CSRF and XSS tokens. + + .. note:: The output is NOT guaranteed to be cryptographically secure, + just the best attempt at that.
\ No newline at end of file |