view roundup/cgi/TranslationService.py @ 5201:a9ace22e0a2f

issue 2550690 - Adding anti-csrf measures to roundup following https://www.owasp.org/index.php/Cross-Site_Request_Forgery_(CSRF)_Prevention_Cheat_Sheet and https://seclab.stanford.edu/websec/csrf/csrf.pdf Basically implement Synchronizer (CSRF) Tokens per form on a page. Single use (destroyed once used). Random input data for the token includes: system random implementation in python using /dev/urandom (fallback to random based on timestamp as the seed. Not as good, but should be ok for the short lifetime of the token??) the id (in cpython it's the memory address) of the object requesting a token. In theory this depends on memory layout, the history of the process (how many previous objects have been allocated from the heap etc.) I claim without any proof that for long running processes this is another source of randomness. For short running processes with little activity it could be guessed. last the floating point time.time() value is added. This may only have 1 second resolution so may be guessable. Hopefully for a short lived (2 week by default) token this is sufficient. Also in the current implementation the user is notified when validation fails and is told why. This allows the roundup admin to find the log entry (at error level) and try to resolve the issue. In the future user notification may change but for now this is probably best.
author John Rouillard <rouilj@ieee.org>
date Sat, 18 Mar 2017 16:59:01 -0400
parents 6e3e4f24c753
children 56c9bcdea47f
line wrap: on
line source

# TranslationService for Roundup templates
#
# This module is free software, you may redistribute it
# and/or modify under the same terms as Python.
#
# This module provides National Language Support
# for Roundup templating - much like roundup.i18n
# module for Roundup command line interface.
# The only difference is that translator objects
# returned by get_translation() have one additional
# method which is used by TAL engines:
#
#   translate(domain, msgid, mapping, context, target_language, default)
#

from roundup import i18n
from roundup.cgi.PageTemplates import Expressions, PathIterator, TALES
from roundup.cgi.TAL import TALInterpreter

### Translation classes

class TranslationServiceMixin:

    OUTPUT_ENCODING = "utf-8"

    def translate(self, domain, msgid, mapping=None,
        context=None, target_language=None, default=None
    ):
        _msg = self.gettext(msgid)
        #print ("TRANSLATE", msgid, _msg, mapping, context)
        _msg = TALInterpreter.interpolate(_msg, mapping)
        return _msg

    def gettext(self, msgid):
        if not isinstance(msgid, unicode):
            msgid = unicode(msgid, 'utf8')
        msgtrans=self.ugettext(msgid)
        return msgtrans.encode(self.OUTPUT_ENCODING)

    def ngettext(self, singular, plural, number):
        if not isinstance(singular, unicode):
            singular = unicode(singular, 'utf8')
        if not isinstance(plural, unicode):
            plural = unicode(plural, 'utf8')
        msgtrans=self.ungettext(singular, plural, number)
        return msgtrans.encode(self.OUTPUT_ENCODING)

class TranslationService(TranslationServiceMixin, i18n.RoundupTranslations):
    pass

class NullTranslationService(TranslationServiceMixin,
        i18n.RoundupNullTranslations):
    def ugettext(self, message):
        if self._fallback:
            return self._fallback.ugettext(message)
        # Sometimes the untranslatable message is a UTF-8 encoded string
        # (thanks to PageTemplate's internals).
        if not isinstance(message, unicode):
            return unicode(message, 'utf8')
        return message

### TAL patching
#
# Template Attribute Language (TAL) uses only global translation service,
# which is not thread-safe.  We will use context variable 'i18n'
# to access request-dependent transalation service (with domain
# and target language set during initializations of the roundup
# client interface.
#

class Context(TALES.Context):

    def __init__(self, compiler, contexts):
        TALES.Context.__init__(self, compiler, contexts)
        if not self.contexts.get('i18n', None):
            # if the context contains no TranslationService,
            # create default one
            self.contexts['i18n'] = get_translation()
        self.i18n = self.contexts['i18n']

    def translate(self, domain, msgid, mapping=None,
                  context=None, target_language=None, default=None):
        if context is None:
            context = self.contexts.get('here')
        return self.i18n.translate(domain, msgid,
            mapping=mapping, context=context, default=default,
            target_language=target_language)

class Engine(TALES.Engine):

    def getContext(self, contexts=None, **kwcontexts):
        if contexts is not None:
            if kwcontexts:
                kwcontexts.update(contexts)
            else:
                kwcontexts = contexts
        return Context(self, kwcontexts)

# patching TAL like this is a dirty hack,
# but i see no other way to specify different Context class
Expressions._engine = Engine(PathIterator.Iterator)
Expressions.installHandlers(Expressions._engine)

### main API function

def get_translation(language=None, tracker_home=None,
    translation_class=TranslationService,
    null_translation_class=NullTranslationService
):
    """Return Translation object for given language and domain

    Arguments 'translation_class' and 'null_translation_class'
    specify the classes that are instantiated for existing
    and non-existing translations, respectively.
    """
    return i18n.get_translation(language=language,
        tracker_home=tracker_home,
        translation_class=translation_class,
        null_translation_class=null_translation_class)

# vim: set et sts=4 sw=4 :

Roundup Issue Tracker: http://roundup-tracker.org/