summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorAymeric Augustin <aymeric.augustin@m4x.org>2014-12-23 22:29:01 +0100
committerAymeric Augustin <aymeric.augustin@m4x.org>2014-12-27 18:02:34 +0100
commit6d52f6f8e688b5c4e70be8352eb02c05fea60e85 (patch)
tree253c8fe96adf7780790e85e8f4f0c1e8daeb5a37
parent5c5eb5fea4d7dcd2b0eed982021cfa8aeee2efd8 (diff)
Fixed #23831 -- Supported strings escaped by third-party libs in Django.
Refs #7261 -- Made strings escaped by Django usable in third-party libs. The changes in mark_safe and mark_for_escaping are straightforward. The more tricky part is to handle correctly objects that implement __html__. Historically escape() has escaped SafeData. Even if that doesn't seem a good behavior, changing it would create security concerns. Therefore support for __html__() was only added to conditional_escape() where this concern doesn't exist. Then using conditional_escape() instead of escape() in the Django template engine makes it understand data escaped by other libraries. Template filter |escape accounts for __html__() when it's available. |force_escape forces the use of Django's HTML escaping implementation. Here's why the change in render_value_in_context() is safe. Before Django 1.7 conditional_escape() was implemented as follows: if isinstance(text, SafeData): return text else: return escape(text) render_value_in_context() never called escape() on SafeData. Therefore replacing escape() with conditional_escape() doesn't change the autoescaping logic as it was originally intended. This change should be backported to Django 1.7 because it corrects a feature added in Django 1.7. Thanks mitsuhiko for the report.
-rw-r--r--django/template/base.py4
-rw-r--r--django/template/debug.py4
-rw-r--r--django/utils/html.py7
-rw-r--r--django/utils/safestring.py8
-rw-r--r--tests/utils_tests/test_safestring.py33
5 files changed, 44 insertions, 12 deletions
diff --git a/django/template/base.py b/django/template/base.py
index 8a9099baa5..4ac20b13e1 100644
--- a/django/template/base.py
+++ b/django/template/base.py
@@ -19,7 +19,7 @@ from django.utils.translation import ugettext_lazy, pgettext_lazy
from django.utils.safestring import (SafeData, EscapeData, mark_safe,
mark_for_escaping)
from django.utils.formats import localize
-from django.utils.html import escape
+from django.utils.html import conditional_escape
from django.utils.module_loading import module_has_submodule
from django.utils import six
from django.utils.timezone import template_localtime
@@ -887,7 +887,7 @@ def render_value_in_context(value, context):
value = force_text(value)
if ((context.autoescape and not isinstance(value, SafeData)) or
isinstance(value, EscapeData)):
- return escape(value)
+ return conditional_escape(value)
else:
return value
diff --git a/django/template/debug.py b/django/template/debug.py
index cf9c2a59fc..465d7ff926 100644
--- a/django/template/debug.py
+++ b/django/template/debug.py
@@ -1,6 +1,6 @@
from django.template.base import Lexer, Parser, tag_re, NodeList, VariableNode, TemplateSyntaxError
from django.utils.encoding import force_text
-from django.utils.html import escape
+from django.utils.html import conditional_escape
from django.utils.safestring import SafeData, EscapeData
from django.utils.formats import localize
from django.utils.timezone import template_localtime
@@ -98,6 +98,6 @@ class DebugVariableNode(VariableNode):
e.django_template_source = self.source
raise
if (context.autoescape and not isinstance(output, SafeData)) or isinstance(output, EscapeData):
- return escape(output)
+ return conditional_escape(output)
else:
return output
diff --git a/django/utils/html.py b/django/utils/html.py
index 3c03210c11..a596662d22 100644
--- a/django/utils/html.py
+++ b/django/utils/html.py
@@ -44,6 +44,10 @@ def escape(text):
"""
Returns the given text with ampersands, quotes and angle brackets encoded
for use in HTML.
+
+ This function always escapes its input, even if it's already escaped and
+ marked as such. This may result in double-escaping. If this is a concern,
+ use conditional_escape() instead.
"""
return mark_safe(force_text(text).replace('&', '&amp;').replace('<', '&lt;')
.replace('>', '&gt;').replace('"', '&quot;').replace("'", '&#39;'))
@@ -76,6 +80,9 @@ escapejs = allow_lazy(escapejs, six.text_type, SafeText)
def conditional_escape(text):
"""
Similar to escape(), except that it doesn't operate on pre-escaped strings.
+
+ This function relies on the __html__ convention used both by Django's
+ SafeData class and by third-party libraries like markupsafe.
"""
if hasattr(text, '__html__'):
return text.__html__()
diff --git a/django/utils/safestring.py b/django/utils/safestring.py
index 50b0c03686..ab4d8149c9 100644
--- a/django/utils/safestring.py
+++ b/django/utils/safestring.py
@@ -36,9 +36,9 @@ else:
class SafeData(object):
def __html__(self):
"""
- Returns the html representation of a string.
+ Returns the html representation of a string for interoperability.
- Allows interoperability with other template engines.
+ This allows other template engines to understand Django's SafeData.
"""
return self
@@ -121,7 +121,7 @@ def mark_safe(s):
Can be called multiple times on a single string.
"""
- if isinstance(s, SafeData):
+ if hasattr(s, '__html__'):
return s
if isinstance(s, bytes) or (isinstance(s, Promise) and s._delegate_bytes):
return SafeBytes(s)
@@ -138,7 +138,7 @@ def mark_for_escaping(s):
Can be called multiple times on a single string (the resulting escaping is
only applied once).
"""
- if isinstance(s, (SafeData, EscapeData)):
+ if hasattr(s, '__html__') or isinstance(s, EscapeData):
return s
if isinstance(s, bytes) or (isinstance(s, Promise) and s._delegate_bytes):
return EscapeBytes(s)
diff --git a/tests/utils_tests/test_safestring.py b/tests/utils_tests/test_safestring.py
index 053d9f42fa..e23851815b 100644
--- a/tests/utils_tests/test_safestring.py
+++ b/tests/utils_tests/test_safestring.py
@@ -13,6 +13,13 @@ lazystr = lazy(force_text, six.text_type)
lazybytes = lazy(force_bytes, bytes)
+class customescape(six.text_type):
+ def __html__(self):
+ # implement specific and obviously wrong escaping
+ # in order to be able to tell for sure when it runs
+ return self.replace('<', '<<').replace('>', '>>')
+
+
class SafeStringTest(TestCase):
def assertRenderEqual(self, tpl, expected, **context):
context = Context(context)
@@ -25,6 +32,14 @@ class SafeStringTest(TestCase):
self.assertRenderEqual('{{ s }}', 'a&b', s=s)
self.assertRenderEqual('{{ s|force_escape }}', 'a&amp;b', s=s)
+ def test_mark_safe_object_implementing_dunder_html(self):
+ e = customescape('<a&b>')
+ s = mark_safe(e)
+ self.assertIs(s, e)
+
+ self.assertRenderEqual('{{ s }}', '<<a&b>>', s=s)
+ self.assertRenderEqual('{{ s|force_escape }}', '&lt;a&amp;b&gt;', s=s)
+
def test_mark_safe_lazy(self):
s = lazystr('a&b')
b = lazybytes(b'a&b')
@@ -42,11 +57,25 @@ class SafeStringTest(TestCase):
self.assertRenderEqual('{{ s }}', '<obj>', s=s)
+ def test_mark_safe_result_implements_dunder_html(self):
+ self.assertEqual(mark_safe('a&b').__html__(), 'a&b')
+
+ def test_mark_safe_lazy_result_implements_dunder_html(self):
+ self.assertEqual(mark_safe(lazystr('a&b')).__html__(), 'a&b')
+
def test_mark_for_escaping(self):
s = mark_for_escaping('a&b')
self.assertRenderEqual('{{ s }}', 'a&amp;b', s=s)
self.assertRenderEqual('{{ s }}', 'a&amp;b', s=mark_for_escaping(s))
+ def test_mark_for_escaping_object_implementing_dunder_html(self):
+ e = customescape('<a&b>')
+ s = mark_for_escaping(e)
+ self.assertIs(s, e)
+
+ self.assertRenderEqual('{{ s }}', '<<a&b>>', s=s)
+ self.assertRenderEqual('{{ s|force_escape }}', '&lt;a&amp;b&gt;', s=s)
+
def test_mark_for_escaping_lazy(self):
s = lazystr('a&b')
b = lazybytes(b'a&b')
@@ -55,10 +84,6 @@ class SafeStringTest(TestCase):
self.assertIsInstance(mark_for_escaping(b), EscapeData)
self.assertRenderEqual('{% autoescape off %}{{ s }}{% endautoescape %}', 'a&amp;b', s=mark_for_escaping(s))
- def test_html(self):
- s = '<h1>interop</h1>'
- self.assertEqual(s, mark_safe(s).__html__())
-
def test_mark_for_escaping_object_implementing_dunder_str(self):
class Obj(object):
def __str__(self):