This makes gzip output deterministic, which allows
ConditionalGetMiddleware to reliably compare ETags on gzipped
content (views using the gzip_page() decorator in particular).
Allowed ServerFormatter to handle simple string messages or messages with
formatting arguments. The formatter will set the server_time variable on
the log record if it's required by the format string but wasn't passed in
extra parameters.
Added format_lazy function to django.utils.text module.
Useful when dealing with relative complex lazy string concatenations
(e.g. in urls.py when translating urls in regular expressions).
Refs #26572, #17281. The RFC doesn't limit total length, just the line length
which is already taken care of by Python itself.
Thanks Tim Graham for the review.
Shallow copying of `django.utils.functional.LazyObject` or its subclasses has
been broken in a couple of different ways in the past, most recently due to
35355a4.
When Django reraises an exception, it sets the __cause__ attribute even
in Python 2, mimicking Python's 3 behavior for "raise Foo from Bar".
However, Python 3 also ensures that all exceptions have a __traceback__
attribute and thus the "traceback2" Python 2 module (backport of Python
3's "traceback" module) relies on the fact that whenever you have a
__cause__ attribute, the recorded exception also has a __traceback__
attribute.
This is breaking testtools which is using traceback2 (see
https://github.com/testing-cabal/testtools/issues/162).
This commit fixes this inconsistency by ensuring that Django sets
the __traceback__ attribute on any exception stored in a __cause__
attribute of a reraised exception.
This only happens if USE_TZ = False and pytz is installed (perhaps not
the most logical combination, but who am I to jugde?)
Refs #23714 which essentially fixed the same problem when USE_TZ = True.
Thanks Florian and Carl for insisting until I wrote a complete patch.
Pickling a `SimpleLazyObject` wrapping a model did not work correctly; in
particular it did not add the `_django_version` attribute added in 42736ac8.
Now it will handle this and other custom `__reduce__` methods correctly.
Without an explicit 'level', only messages at WARNING or higher
are handled. This makes the config consistent with the docs
which say, "The django catch-all logger sends all messages at
the INFO level or higher to the console."
The ``item_enclosures`` hook returns a list of ``Enclosure`` objects which is
then used by the feed builder. If the feed is a RSS feed, an exception is
raised as RSS feeds don't allow multiple enclosures per feed item.
The ``item_enclosures`` hook defaults to an empty list or, if the
``item_enclosure_url`` hook is defined, to a list with a single ``Enclosure``
built from the ``item_enclosure_url``, ``item_enclosure_length``, and
``item_enclosure_mime_type`` hooks.
With this change, it's expected to survive anything except errors
that make it impossible to import the settings. It's too complex
to fallback to a sensible behavior with a broken settings module.
Harcoding things about runserver in ManagementUtility.execute is
atrocious but it's the only way out of the chicken'n'egg problem:
the current implementation of the autoreloader primarily watches
imported Python modules -- and then a few other things that were
bolted on top of this design -- but we want it to kick in even if
the project contains import-time errors and django.setup() fails.
At some point we should throw away this code and replace it by an
off-the-shelf autoreloader that watches the working directory and
re-runs `django-admin runserver` whenever something changes.
* When some old files contain errors, the second call to
gen_filenames() should return them.
* When some new files contain errors, the first call to
gen_filenames(only_new=True) should return them.
Refactored tests to use a sample project.
Updated extraction:
* Removed special handling of single percent signs.
* When extracting messages from template text, doubled all percent signs
so they are not interpreted by gettext as string format flags. All
strings extracted by gettext, if containing a percent sign, will now
be labeled "#, python-format".
Updated translation:
* Used "%%" for "%" in template text before calling gettext.
* Updated {% trans %} rendering to restore "%" from "%%".
Commit 15f82c7 ("used pyinotify as change detection system when
available") introduced a regression where editing a file in vim with
default settings (writebackup=auto) no longer causes the dev server
to be restarted. On a write, vim moves the monitored file to a backup
path and then creates a new file in the original. The new file is not
monitored as it has a different inode. Fixed this by also watching for
inotify events IN_DELETE_SELF and IN_MOVE_SELF.
Implemented __str__() to return the string-representation of the
proxied object, not the proxy itself, if the lazy object didn't have
a string-like object in its resultclasses.
This avoids calling date.tzinfo.utcoffset(date) twice. It's also
robust to cases where that function returns None -- which never
happens in practice :-)
This patch does three major things:
* Merges the django.template.debug implementation into django.template.base.
* Simplifies the debug implementation.
The old implementation copied debug information to every token and node.
The django_template_source attribute was set in multiple places, some
quite hacky, like django.template.defaulttags.ForNode.
Debug information is now annotated in two high-level places:
* Template.compile_nodelist for errors during parsing
* Node.render_annotated for errors during rendering
These were chosen because they have access to the template and context
as well as to all exceptions that happen during either the parse or
render phase.
* Moves the contextual line traceback information creation from
django.views.debug into django.template.base.Template. The debug views now
only deal with the presentation of the debug information.
There's no reason to assume that sys.path[0] is an appropriate location
for generating code. Specifically that doesn't work with extend_sys_path
which puts the additional directories at the end of sys.path.
In order to create a new migrations module, instead of using an
arbitrary entry from sys.path, import as much as possible from the path
to the module, then create missing submodules from there.
Without this change, the tests introduced in the following commit fail,
which seems sufficient to prevent regressions for such a refactoring.
The stated reason for its introduction in d18d37ce no longer applies
since Django's code repository was switched from Subversion to git.
Furthermore it never had any effect because shutil.rmtree ignores its
onerror argument when ignore_errors is True.
The reason for its use in template management commands is unclear.
This dramatically improves performance on PyPy. The following benchmark:
python -mtimeit -s "from django.utils.functional import allow_lazy; from django.utils.translation import ugettext_lazy; f = allow_lazy(lambda s: s, str)" "f(ugettext_lazy('abc'))"
goes from 390us per loop to 165us.
The function no longer flushes zfile after each write as doing so can
lead to the gzipped streamed content being larger than the original
content; each flush adds a 5/6 byte type 0 block. Removing this means
buf.read() may return nothing, so only yield if that has some data.
Testing shows without the flush() the buffer is being flushed every 17k
or so and compresses the same as if it had been done as a whole string.
Since this package is going to hold both the implementation of the Django
Template Language and the infrastructure for Multiple Template Engines,
it should be untied from the DTL as much as possible within our
backwards-compatibility policy.
Only public APIs (i.e. APIs mentioned in the documentation) were left.
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.
mark_safe and mark_for_escaping should have been kept similar.
On Python 2 this change has no effect. On Python 3 it fixes the use case
shown in the regression test for mark_for_escaping, which used to raise
a TypeError. The regression test for mark_safe is just for completeness.
A field for storing periods of time - modeled in Python by timedelta. It
is stored in the native interval data type on PostgreSQL and as a bigint
of microseconds on other backends.
Also includes significant changes to the internals of time related maths
in expressions, including the removal of DateModifierNode.
Thanks to Tim and Josh in particular for reviews.
Added a test for the condition safe_join is designed to prevent.
Previously, a generic ValueError was raised. It was impossible to tell
an intentional exception raised to implement safe_join's contract from
an unintentional exception caused by incorrect inputs or unexpected
conditions. That resulted in bizarre exception catching patterns, which
this patch removes.
Since safe_join is a private API and since the change is unlikely to
create security issues for users who use it anyway -- at worst, an
uncaught SuspiciousFileOperation exception will bubble up -- it isn't
documented.
This also defines QuerySet.__bool__ for consistency though this should not have any consequence as bool(qs) used to fallback on QuerySet.__len__ in Py3.
Translating an empty string used to return the gettext catalog
metadata instead of the empty string.
Thanks Ned Batchelder for the suggestion, Tim Graham for the review
and Anton Berezin and Claude Paroz for contributions to the patch.
This reverts commit 66757fee7e.
Discussions have led to think that this functionality does not
bring significant benefits to justify the added complexity.
Read also discussions on ticket #22734.
This fixes a regression introduced in 6d302f639.
Thanks lorinkoz at gmail.com for the report, Collin Anderson
for the initial patch and Simon Charette for the review.
This fixes the Chinese language issues described in #23005 but
also provides for other fallback exceptions by updating the
LANG_INFO structure.
Thanks caxekis at gmail.com for the report and Tim Graham for the
review.
The rate at which we've increased this has not been keeping up with hardware (and software) improvements, and we're now considerably behind where we should be. The delta between our performance and an optimized implementation's performance prevents us from improving that further, but hopefully once Python 2.7.8 and 3.4+ get into more hands we can more aggressively increase this number.
And follow more closely the class of characters defined in the
RFC 3986.
Thanks Erik van Zijst for the report and the initial patch, and
Tim Graham for the review.
This follows commits 80f4487 and 01399fa; original patch had to be
reverted because it wasn't Python 2.6 compatible and we need it to
be in order to build docs on the djangoproject.com server.
This fix should be replaced by @lru_cache as soon as we drop
Python 2.6 compatibility.
Thanks Florian Apolloner for the review and Alexander Schepanovski
for the original patch.
This reverts commit 80f4487 temporarily, because that commit prevented
the djangoproject.com server from building the docs, because it still
uses Python 2.6.
Python 2.7.7 includes compare_digest in the hmac module, but it requires
both arguments to have the same type. This is usually not a problem on
Python 3 since everything is text, but we have mixed unicode and str on
Python 2 -- hence make sure everything is bytes before feeding it into
compare_digest.
Previously the FORMAT_MODULE_PATH setting only accepted one string (dotted
module path). A feature has been added to also allow a list of strings.
This is useful when using several reusable third party apps that define new
formats. We can now use them all and we can even override some of the formats
by providing a project-wide format module.
Previously the FORMAT_MODULE_PATH setting only accepted one string (dotted
module path).
This is useful when using several reusable third party apps that define new
formats. We can now use them all and we can even override some of the formats
by providing a project-wide format module.
While Node class has a useful `__str__`, its `__repr__` is not that
useful. Added a `__repr__` that makes use of the current `__str__`.
This is especially useful since the more popular `Q` class inherits
`tree.Node`. Also created new tests that cover most of `Node` class
functionality.
This is a bit faster than ours, which is good, because it lets you increase
the iteration counts.
This will be used on Python 3.4+, and, pending the acceptance of PEP466, on
newer Python 2.7s.
There may be more than 100 (default maxsize) commonly seen xx-yy values
on some sites. The additional memory consumption isn't significant.
Also added a comment explaining why this cache must have a maxsize.
Since d2e242d16c6dde6f4736086fb38057424bed3edb made isinstance()
calls work correctly on LazyObject, we can simplify the
implementation of is_local_storage added in
7e27885c6e.
By removing the 'supported' keyword from the detection methods and only relying
on a cached settings.LANGUAGES, the speed of said methods has been improved;
around 4x raw performance. This allows us to stop checking Python's incomplete
list of locales, and rely on a less restrictive regular expression for
accepting certain locales.
HTTP Accept-Language is defined as being case-insensitive, based on this fact
extra performance improvements have been made; it wouldn't make sense to
check for case differences.
When a method decorator was used in conjunction with a decorator
implemented as a descriptor, method_decorator did not correctly respect
the method binding.
Thanks for Graham Dumpleton for the report and initial patch.
This reverts commit 2ee447fb5f.
That commit introduced a regression (#21882) and didn't really
do what it was supposed to: while it did delay the evaluation
of lazy objects passed to mark_safe(), they weren't actually
marked as such so they could end up being escaped twice.
Refs #21882.
This is the result of Christopher Medrela's 2013 Summer of Code project.
Thanks also to Preston Holmes, Tim Graham, Anssi Kääriäinen, Florian
Apolloner, and Alex Gaynor for review notes along the way.
Also: Fixes#8579, fixes#3055, fixes#19844.
Python 3.5 will change the default value of convert_charrefs, so 3.4
gives warnings if it's not present. This is slightly technical as 2.7
doesn't have the kwarg. Thankfully, we already have a bunch of
workarounds for different versions.
Since the app registry is always populated before the first request is
processed, the situation described in #18251 for the old app cache
cannot happen any more.
Refs #18251, #21628.
Copied attributes into the decorated method and special case __name__
copy as this will not be present on a Class object. Added regression
test to decorator suite.
It was not used inside Django, is not tested or documented. Consequently
remove without deprecation path.
Thanks to @vajrasky for bringing it to our attention.
Claude Paroz
Kevin Christopher Henry
Nick Pope
Denis Cornehl
Reto Aebersold
Rinat Khabibiev
Sergey Fursov
Jon Dufresne
Tim Graham
Jani Tiainen
Olexander Yermakov
Mattias Loverot
Przemysław Suliga
Ville Skyttä
Will Hardy
jasisz
Carl Meyer
Vytis Banaitis
Thejaswi Puthraya
Scott Vitale
Chesco Igual
Ketan Bhatt
Florian Apolloner
Andre Cruz
Vasiliy Faronov
Marko Benko
Jeremy Lainé
Amine Yaiche
Mark Striemer
Nick Malakhov
Alexey Kotlyarov
Marcin Markiewicz
Ben Kraft
userimack
Flavio Curella
Benjamin Bach
Iacopo Spalletti
Raphaël Hertzog
Gagaro
Attila Tovt
Jaap Roes
Marti Raudsepp
Dwight Gunning
Aymeric Augustin
Neal Todd
Matt Deacalion Stevens
Unai Zalakain
fabrizio ettore messina
Matt Robenolt
Zan Anderle
Alexandre Pocquet
Maxime Lorant
Doug Beck
Rigel Di Scala
Edward Henderson
Vlastimil Zíma
darkryder
Chris Bainbridge
Marten Kenbeek
Moritz Sichert
Tomasz Kontusz
Matthew Somerville
Raphael Michel
Tomáš Ehrlich
zauddelig
Piotr Jakimiak
Markus Bertheau
Michal Čihař
Josh Smeaton
Curtis Maloney
Bas Peschier
Preston Timmons
Rik
Benjamin Wohlwend
Aymeric Augustin
Lukas Klein
Alex Gaynor
Varun Sharma
Adam Taylor
Gavin Wahl
Collin Anderson
Oscar Ramirez
Marc Tamlyn
Michael Hall
Berker Peksag
Eric Rouleau
Tom V
Markus Holtermann
David Hoffman
John-Scott Atlakson
Thomas Chaumeny
Anubhav Joshi
Jaap Roes
Seth Hill
Yuri Kriachko
Thomas Chaumeny
Ian Foote
Sitesh Shrivastava
Baptiste Mispelon
LarryBrid
LarryBrid
Loic Bistuer
Alexander Schepanovski
Andrew Godwin
Shai Berger
Richard Eames
Alexandr Shurigin
Martin Brochhaus
Markus Amalthea Magnuson
Moayad Mardini
Erik Romijn
Simon Charette
Tomasz Wysocki
Chris Beaven
Rodolfo Carvalho
Bouke Haarsma
MattBlack85
Jeremy
Jannis Leidel
Vajrasky Kok
Russell Keith-Magee
dpwrussell
ijl