django/docs/ref/templates/builtins.txt

==================================
Built-in template tags and filters
==================================

This document describes Django's built-in template tags and filters. It is
recommended that you use the :doc:`automatic documentation
</ref/contrib/admin/admindocs>`, if available, as this will also include
documentation for any custom tags or filters installed.

.. _ref-templates-builtins-tags:

Built-in tag reference
----------------------

.. highlightlang:: html+django

.. templatetag:: autoescape

autoescape
^^^^^^^^^^

Controls the current auto-escaping behavior. This tag takes either ``on`` or
``off`` as an argument and that determines whether auto-escaping is in effect
inside the block. The block is closed with an ``endautoescape`` ending tag.

When auto-escaping is in effect, all variable content has HTML escaping applied
to it before placing the result into the output (but after any filters have
been applied). This is equivalent to manually applying the :tfilter:`escape`
filter to each variable.

The only exceptions are variables that are already marked as "safe" from
escaping, either by the code that populated the variable, or because it has had
the :tfilter:`safe` or :tfilter:`escape` filters applied.

Sample usage::

    {% autoescape on %}
        {{ body }}
    {% endautoescape %}

.. templatetag:: block

block
^^^^^

Defines a block that can be overridden by child templates. See
:ref:`Template inheritance <template-inheritance>` for more information.

.. templatetag:: comment

comment
^^^^^^^

Ignores everything between ``{% comment %}`` and ``{% endcomment %}``.
An optional note may be inserted in the first tag. For example, this is
useful when commenting out code for documenting why the code was disabled.

Sample usage::

    <p>Rendered text with {{ pub_date|date:"c" }}</p>
    {% comment "Optional note" %}
        <p>Commented out text with {{ create_date|date:"c" }}</p>
    {% endcomment %}

``comment`` tags cannot be nested.

.. templatetag:: csrf_token

csrf_token
^^^^^^^^^^

This tag is used for CSRF protection, as described in the documentation for
:doc:`Cross Site Request Forgeries </ref/csrf>`.

.. templatetag:: cycle

cycle
^^^^^

Produces one of its arguments each time this tag is encountered. The first
argument is produced on the first encounter, the second argument on the second
encounter, and so forth. Once all arguments are exhausted, the tag cycles to
the first argument and produces it again.

This tag is particularly useful in a loop::

    {% for o in some_list %}
        <tr class="{% cycle 'row1' 'row2' %}">
            ...
        </tr>
    {% endfor %}

The first iteration produces HTML that refers to class ``row1``, the second to
``row2``, the third to ``row1`` again, and so on for each iteration of the
loop.

You can use variables, too. For example, if you have two template variables,
``rowvalue1`` and ``rowvalue2``, you can alternate between their values like
this::

    {% for o in some_list %}
        <tr class="{% cycle rowvalue1 rowvalue2 %}">
            ...
        </tr>
    {% endfor %}

Variables included in the cycle will be escaped.  You can disable auto-escaping
with::

    {% for o in some_list %}
        <tr class="{% autoescape off %}{% cycle rowvalue1 rowvalue2 %}{% endautoescape %}">
            ...
        </tr>
    {% endfor %}

You can mix variables and strings::

    {% for o in some_list %}
        <tr class="{% cycle 'row1' rowvalue2 'row3' %}">
            ...
        </tr>
    {% endfor %}

In some cases you might want to refer to the current value of a cycle
without advancing to the next value. To do this,
just give the ``{% cycle %}`` tag a name, using "as", like this::

    {% cycle 'row1' 'row2' as rowcolors %}

From then on, you can insert the current value of the cycle wherever you'd like
in your template by referencing the cycle name as a context variable. If you
want to move the cycle to the next value independently of the original
``cycle`` tag, you can use another ``cycle`` tag and specify the name of the
variable. So, the following template::

    <tr>
        <td class="{% cycle 'row1' 'row2' as rowcolors %}">...</td>
        <td class="{{ rowcolors }}">...</td>
    </tr>
    <tr>
        <td class="{% cycle rowcolors %}">...</td>
        <td class="{{ rowcolors }}">...</td>
    </tr>

would output::

    <tr>
        <td class="row1">...</td>
        <td class="row1">...</td>
    </tr>
    <tr>
        <td class="row2">...</td>
        <td class="row2">...</td>
    </tr>

You can use any number of values in a ``cycle`` tag, separated by spaces.
Values enclosed in single quotes (``'``) or double quotes (``"``) are treated
as string literals, while values without quotes are treated as template
variables.

By default, when you use the ``as`` keyword with the cycle tag, the
usage of ``{% cycle %}`` that initiates the cycle will itself produce
the first value in the cycle. This could be a problem if you want to
use the value in a nested loop or an included template. If you only want
to declare the cycle but not produce the first value, you can add a
``silent`` keyword as the last keyword in the tag. For example::

    {% for obj in some_list %}
        {% cycle 'row1' 'row2' as rowcolors silent %}
        <tr class="{{ rowcolors }}">{% include "subtemplate.html" %}</tr>
    {% endfor %}

This will output a list of ``<tr>`` elements with ``class``
alternating between ``row1`` and ``row2``. The subtemplate will have
access to ``rowcolors`` in its context and the value will match the class
of the ``<tr>`` that encloses it. If the ``silent`` keyword were to be
omitted, ``row1`` and ``row2`` would be emitted as normal text, outside the
``<tr>`` element.

When the silent keyword is used on a cycle definition, the silence
automatically applies to all subsequent uses of that specific cycle tag.
The following template would output *nothing*, even though the second
call to ``{% cycle %}`` doesn't specify ``silent``::

    {% cycle 'row1' 'row2' as rowcolors silent %}
    {% cycle rowcolors %}

.. deprecated:: 1.9

    The ``{% cycle %}`` tag supports the much inferior old syntax from previous
    Django versions. You shouldn't use this in any new projects, but for the
    sake of the people who are still using it, here's what it looks like::

        {% cycle row1,row2,row3 %}

    In this syntax, each value gets interpreted as a literal string, and there's
    no way to specify variable values, literal commas, or spaces. Support for
    this syntax will be removed in Django 2.0.

.. templatetag:: debug

debug
^^^^^

Outputs a whole load of debugging information, including the current context
and imported modules.

.. templatetag:: extends

extends
^^^^^^^

Signals that this template extends a parent template.

This tag can be used in two ways:

* ``{% extends "base.html" %}`` (with quotes) uses the literal value
  ``"base.html"`` as the name of the parent template to extend.

* ``{% extends variable %}`` uses the value of ``variable``. If the variable
  evaluates to a string, Django will use that string as the name of the
  parent template. If the variable evaluates to a ``Template`` object,
  Django will use that object as the parent template.

See :ref:`template-inheritance` for more information.

.. templatetag:: filter

filter
^^^^^^

Filters the contents of the block through one or more filters. Multiple
filters can be specified with pipes and filters can have arguments, just as
in variable syntax.

Note that the block includes *all* the text between the ``filter`` and
``endfilter`` tags.

Sample usage::

    {% filter force_escape|lower %}
        This text will be HTML-escaped, and will appear in all lowercase.
    {% endfilter %}

.. note::

    The :tfilter:`escape` and :tfilter:`safe` filters are not acceptable
    arguments. Instead, use the :ttag:`autoescape` tag to manage autoescaping
    for blocks of template code.

.. templatetag:: firstof

firstof
^^^^^^^

Outputs the first argument variable that is not ``False``. Outputs nothing if
all the passed variables are ``False``.

Sample usage::

    {% firstof var1 var2 var3 %}

This is equivalent to::

    {% if var1 %}
        {{ var1 }}
    {% elif var2 %}
        {{ var2 }}
    {% elif var3 %}
        {{ var3 }}
    {% endif %}

You can also use a literal string as a fallback value in case all
passed variables are False::

    {% firstof var1 var2 var3 "fallback value" %}

This tag auto-escapes variable values. You can disable auto-escaping with::

    {% autoescape off %}
        {% firstof var1 var2 var3 "<strong>fallback value</strong>" %}
    {% endautoescape %}

Or if only some variables should be escaped, you can use::

    {% firstof var1 var2|safe var3 "<strong>fallback value</strong>"|safe %}

You can use the syntax ``{% firstof var1 var2 var3 as value %}`` to store the
output inside a variable.

.. versionadded:: 1.9

    The "as" syntax was added.

.. templatetag:: for

for
^^^

Loops over each item in an array, making the item available in a context
variable. For example, to display a list of athletes provided in
``athlete_list``::

    <ul>
    {% for athlete in athlete_list %}
        <li>{{ athlete.name }}</li>
    {% endfor %}
    </ul>

You can loop over a list in reverse by using
``{% for obj in list reversed %}``.

If you need to loop over a list of lists, you can unpack the values
in each sublist into individual variables. For example, if your context
contains a list of (x,y) coordinates called ``points``, you could use the
following to output the list of points::

    {% for x, y in points %}
        There is a point at {{ x }},{{ y }}
    {% endfor %}

This can also be useful if you need to access the items in a dictionary.
For example, if your context contained a dictionary ``data``, the following
would display the keys and values of the dictionary::

    {% for key, value in data.items %}
        {{ key }}: {{ value }}
    {% endfor %}

The for loop sets a number of variables available within the loop:

==========================  ===============================================
Variable                    Description
==========================  ===============================================
``forloop.counter``         The current iteration of the loop (1-indexed)
``forloop.counter0``        The current iteration of the loop (0-indexed)
``forloop.revcounter``      The number of iterations from the end of the
                            loop (1-indexed)
``forloop.revcounter0``     The number of iterations from the end of the
                            loop (0-indexed)
``forloop.first``           True if this is the first time through the loop
``forloop.last``            True if this is the last time through the loop
``forloop.parentloop``      For nested loops, this is the loop surrounding
                            the current one
==========================  ===============================================

for ... empty
^^^^^^^^^^^^^

The ``for`` tag can take an optional ``{% empty %}`` clause whose text is
displayed if the given array is empty or could not be found::

    <ul>
    {% for athlete in athlete_list %}
        <li>{{ athlete.name }}</li>
    {% empty %}
        <li>Sorry, no athletes in this list.</li>
    {% endfor %}
    </ul>

The above is equivalent to -- but shorter, cleaner, and possibly faster
than -- the following::

    <ul>
      {% if athlete_list %}
        {% for athlete in athlete_list %}
          <li>{{ athlete.name }}</li>
        {% endfor %}
      {% else %}
        <li>Sorry, no athletes in this list.</li>
      {% endif %}
    </ul>

.. templatetag:: if

if
^^

The ``{% if %}`` tag evaluates a variable, and if that variable is "true" (i.e.
exists, is not empty, and is not a false boolean value) the contents of the
block are output::

    {% if athlete_list %}
        Number of athletes: {{ athlete_list|length }}
    {% elif athlete_in_locker_room_list %}
        Athletes should be out of the locker room soon!
    {% else %}
        No athletes.
    {% endif %}

In the above, if ``athlete_list`` is not empty, the number of athletes will be
displayed by the ``{{ athlete_list|length }}`` variable.

As you can see, the ``if`` tag may take one or several ``{% elif %}``
clauses, as well as an ``{% else %}`` clause that will be displayed if all
previous conditions fail. These clauses are optional.

Boolean operators
^^^^^^^^^^^^^^^^^

:ttag:`if` tags may use ``and``, ``or`` or ``not`` to test a number of
variables or to negate a given variable::

    {% if athlete_list and coach_list %}
        Both athletes and coaches are available.
    {% endif %}

    {% if not athlete_list %}
        There are no athletes.
    {% endif %}

    {% if athlete_list or coach_list %}
        There are some athletes or some coaches.
    {% endif %}

    {% if not athlete_list or coach_list %}
        There are no athletes or there are some coaches.
    {% endif %}

    {% if athlete_list and not coach_list %}
        There are some athletes and absolutely no coaches.
    {% endif %}

Use of both ``and`` and ``or`` clauses within the same tag is allowed, with
``and`` having higher precedence than ``or`` e.g.::

    {% if athlete_list and coach_list or cheerleader_list %}

will be interpreted like:

.. code-block:: python

    if (athlete_list and coach_list) or cheerleader_list

Use of actual parentheses in the :ttag:`if` tag is invalid syntax. If you need
them to indicate precedence, you should use nested :ttag:`if` tags.

:ttag:`if` tags may also use the operators ``==``, ``!=``, ``<``, ``>``,
``<=``, ``>=`` and ``in`` which work as follows:


``==`` operator
^^^^^^^^^^^^^^^

Equality. Example::

    {% if somevar == "x" %}
      This appears if variable somevar equals the string "x"
    {% endif %}

``!=`` operator
^^^^^^^^^^^^^^^

Inequality. Example::

    {% if somevar != "x" %}
      This appears if variable somevar does not equal the string "x",
      or if somevar is not found in the context
    {% endif %}

``<`` operator
^^^^^^^^^^^^^^

Less than. Example::

    {% if somevar < 100 %}
      This appears if variable somevar is less than 100.
    {% endif %}

``>`` operator
^^^^^^^^^^^^^^

Greater than. Example::

    {% if somevar > 0 %}
      This appears if variable somevar is greater than 0.
    {% endif %}

``<=`` operator
^^^^^^^^^^^^^^^

Less than or equal to. Example::

    {% if somevar <= 100 %}
      This appears if variable somevar is less than 100 or equal to 100.
    {% endif %}

``>=`` operator
^^^^^^^^^^^^^^^

Greater than or equal to. Example::

    {% if somevar >= 1 %}
      This appears if variable somevar is greater than 1 or equal to 1.
    {% endif %}

``in`` operator
^^^^^^^^^^^^^^^

Contained within. This operator is supported by many Python containers to test
whether the given value is in the container. The following are some examples
of how ``x in y`` will be interpreted::

    {% if "bc" in "abcdef" %}
      This appears since "bc" is a substring of "abcdef"
    {% endif %}

    {% if "hello" in greetings %}
      If greetings is a list or set, one element of which is the string
      "hello", this will appear.
    {% endif %}

    {% if user in users %}
      If users is a QuerySet, this will appear if user is an
      instance that belongs to the QuerySet.
    {% endif %}

``not in`` operator
^^^^^^^^^^^^^^^^^^^

Not contained within. This is the negation of the ``in`` operator.


The comparison operators cannot be 'chained' like in Python or in mathematical
notation. For example, instead of using::

    {% if a > b > c %}  (WRONG)

you should use::

    {% if a > b and b > c %}


Filters
^^^^^^^

You can also use filters in the :ttag:`if` expression. For example::

    {% if messages|length >= 100 %}
       You have lots of messages today!
    {% endif %}

Complex expressions
^^^^^^^^^^^^^^^^^^^

All of the above can be combined to form complex expressions. For such
expressions, it can be important to know how the operators are grouped when the
expression is evaluated - that is, the precedence rules. The precedence of the
operators, from lowest to highest, is as follows:

* ``or``
* ``and``
* ``not``
* ``in``
* ``==``, ``!=``, ``<``, ``>``, ``<=``, ``>=``

(This follows Python exactly). So, for example, the following complex
:ttag:`if` tag:

.. code-block:: django

    {% if a == b or c == d and e %}

...will be interpreted as:

.. code-block:: python

    (a == b) or ((c == d) and e)

If you need different precedence, you will need to use nested :ttag:`if` tags.
Sometimes that is better for clarity anyway, for the sake of those who do not
know the precedence rules.


.. templatetag:: ifchanged

ifchanged
^^^^^^^^^

Check if a value has changed from the last iteration of a loop.

The ``{% ifchanged %}`` block tag is used within a loop. It has two possible
uses.

1. Checks its own rendered contents against its previous state and only
   displays the content if it has changed. For example, this displays a list of
   days, only displaying the month if it changes::

        <h1>Archive for {{ year }}</h1>

        {% for date in days %}
            {% ifchanged %}<h3>{{ date|date:"F" }}</h3>{% endifchanged %}
            <a href="{{ date|date:"M/d"|lower }}/">{{ date|date:"j" }}</a>
        {% endfor %}

2. If given one or more variables, check whether any variable has changed.
   For example, the following shows the date every time it changes, while
   showing the hour if either the hour or the date has changed::

        {% for date in days %}
            {% ifchanged date.date %} {{ date.date }} {% endifchanged %}
            {% ifchanged date.hour date.date %}
                {{ date.hour }}
            {% endifchanged %}
        {% endfor %}

The ``ifchanged`` tag can also take an optional ``{% else %}`` clause that
will be displayed if the value has not changed::

        {% for match in matches %}
            <div style="background-color:
                {% ifchanged match.ballot_id %}
                    {% cycle "red" "blue" %}
                {% else %}
                    gray
                {% endifchanged %}
            ">{{ match }}</div>
        {% endfor %}

.. templatetag:: ifequal

ifequal
^^^^^^^

Output the contents of the block if the two arguments equal each other.

Example::

    {% ifequal user.pk comment.user_id %}
        ...
    {% endifequal %}

As in the :ttag:`if` tag, an ``{% else %}`` clause is optional.

The arguments can be hard-coded strings, so the following is valid::

    {% ifequal user.username "adrian" %}
        ...
    {% endifequal %}

An alternative to the ``ifequal`` tag is to use the :ttag:`if` tag and the
``==`` operator.

.. templatetag:: ifnotequal

ifnotequal
^^^^^^^^^^

Just like :ttag:`ifequal`, except it tests that the two arguments are not
equal.

An alternative to the ``ifnotequal`` tag is to use the :ttag:`if` tag and
the ``!=`` operator.

.. templatetag:: include

include
^^^^^^^

Loads a template and renders it with the current context. This is a way of
"including" other templates within a template.

The template name can either be a variable or a hard-coded (quoted) string,
in either single or double quotes.

This example includes the contents of the template ``"foo/bar.html"``::

    {% include "foo/bar.html" %}

This example includes the contents of the template whose name is contained in
the variable ``template_name``::

    {% include template_name %}

The variable may also be any object with a ``render()`` method that accepts a
context. This allows you to reference a compiled ``Template`` in your context.

An included template is rendered within the context of the template that
includes it. This example produces the output ``"Hello, John!"``:

* Context: variable ``person`` is set to ``"John"`` and variable ``greeting``
  is set to ``"Hello"``.

* Template::

    {% include "name_snippet.html" %}

* The ``name_snippet.html`` template::

    {{ greeting }}, {{ person|default:"friend" }}!

You can pass additional context to the template using keyword arguments::

    {% include "name_snippet.html" with person="Jane" greeting="Hello" %}

If you want to render the context only with the variables provided (or even
no variables at all), use the ``only`` option. No other variables are
available to the included template::

    {% include "name_snippet.html" with greeting="Hi" only %}

If the included template causes an exception while it's rendered (including
if it's missing or has syntax errors), the behavior varies depending on the
:class:`template engine's <django.template.Engine>` ``debug`` option (if not
set, this option defaults to the value of :setting:`DEBUG`). When debug mode is
turned on, an exception like :exc:`~django.template.TemplateDoesNotExist` or
:exc:`~django.template.TemplateSyntaxError` will be raised; otherwise
``{% include %}`` silences any exception that happens while rendering the
included template and returns an empty string.

.. note::
    The :ttag:`include` tag should be considered as an implementation of
    "render this subtemplate and include the HTML", not as "parse this
    subtemplate and include its contents as if it were part of the parent".
    This means that there is no shared state between included templates --
    each include is a completely independent rendering process.

    Blocks are evaluated *before* they are included. This means that a template
    that includes blocks from another will contain blocks that have *already
    been evaluated and rendered* - not blocks that can be overridden by, for
    example, an extending template.

.. templatetag:: load

load
^^^^

Loads a custom template tag set.

For example, the following template would load all the tags and filters
registered in ``somelibrary`` and ``otherlibrary`` located in package
``package``::

    {% load somelibrary package.otherlibrary %}

You can also selectively load individual filters or tags from a library, using
the ``from`` argument. In this example, the template tags/filters named ``foo``
and ``bar`` will be loaded from ``somelibrary``::

    {% load foo bar from somelibrary %}

See :doc:`Custom tag and filter libraries </howto/custom-template-tags>` for
more information.

.. templatetag:: lorem

lorem
^^^^^

.. versionadded:: 1.8

    The tag was previously located in :mod:`django.contrib.webdesign`.

Displays random "lorem ipsum" Latin text. This is useful for providing sample
data in templates.

Usage::

    {% lorem [count] [method] [random] %}

The ``{% lorem %}`` tag can be used with zero, one, two or three arguments.
The arguments are:

===========  =============================================================
Argument     Description
===========  =============================================================
``count``    A number (or variable) containing the number of paragraphs or
             words to generate (default is 1).
``method``   Either ``w`` for words, ``p`` for HTML paragraphs or ``b``
             for plain-text paragraph blocks (default is ``b``).
``random``   The word ``random``, which if given, does not use the common
             paragraph ("Lorem ipsum dolor sit amet...") when generating
             text.
===========  =============================================================

Examples:

* ``{% lorem %}`` will output the common "lorem ipsum" paragraph.
* ``{% lorem 3 p %}`` will output the common "lorem ipsum" paragraph
  and two random paragraphs each wrapped in HTML ``<p>`` tags.
* ``{% lorem 2 w random %}`` will output two random Latin words.

.. templatetag:: now

now
^^^

Displays the current date and/or time, using a format according to the given
string. Such string can contain format specifiers characters as described
in the :tfilter:`date` filter section.

Example::

    It is {% now "jS F Y H:i" %}

Note that you can backslash-escape a format string if you want to use the
"raw" value. In this example, both "o" and "f" are backslash-escaped, because
otherwise each is a format string that displays the year and the time,
respectively::

    It is the {% now "jS \o\f F" %}

This would display as "It is the 4th of September".

.. note::

    The format passed can also be one of the predefined ones
    :setting:`DATE_FORMAT`, :setting:`DATETIME_FORMAT`,
    :setting:`SHORT_DATE_FORMAT` or :setting:`SHORT_DATETIME_FORMAT`.
    The predefined formats may vary depending on the current locale and
    if :ref:`format-localization` is enabled, e.g.::

        It is {% now "SHORT_DATETIME_FORMAT" %}


You can also use the syntax ``{% now "Y" as current_year %}`` to store the
output inside a variable. This is useful if you want to use ``{% now %}``
inside a template tag like :ttag:`blocktrans` for example::

    {% now "Y" as current_year %}
    {% blocktrans %}Copyright {{ current_year }}{% endblocktrans %}

.. versionadded:: 1.8

The ability to use the "as" syntax was added.

.. templatetag:: regroup

regroup
^^^^^^^

Regroups a list of alike objects by a common attribute.

This complex tag is best illustrated by way of an example: say that "places" is a list of cities represented by dictionaries containing ``"name"``, ``"population"``, and ``"country"`` keys:

.. code-block:: python

    cities = [
        {'name': 'Mumbai', 'population': '19,000,000', 'country': 'India'},
        {'name': 'Calcutta', 'population': '15,000,000', 'country': 'India'},
        {'name': 'New York', 'population': '20,000,000', 'country': 'USA'},
        {'name': 'Chicago', 'population': '7,000,000', 'country': 'USA'},
        {'name': 'Tokyo', 'population': '33,000,000', 'country': 'Japan'},
    ]

...and you'd like to display a hierarchical list that is ordered by country, like this:

* India

  * Mumbai: 19,000,000
  * Calcutta: 15,000,000

* USA

  * New York: 20,000,000
  * Chicago: 7,000,000

* Japan

  * Tokyo: 33,000,000


You can use the ``{% regroup %}`` tag to group the list of cities by country.
The following snippet of template code would accomplish this::

    {% regroup cities by country as country_list %}

    <ul>
    {% for country in country_list %}
        <li>{{ country.grouper }}
        <ul>
            {% for item in country.list %}
              <li>{{ item.name }}: {{ item.population }}</li>
            {% endfor %}
        </ul>
        </li>
    {% endfor %}
    </ul>

Let's walk through this example. ``{% regroup %}`` takes three arguments: the
list you want to regroup, the attribute to group by, and the name of the
resulting list. Here, we're regrouping the ``cities`` list by the ``country``
attribute and calling the result ``country_list``.

``{% regroup %}`` produces a list (in this case, ``country_list``) of
**group objects**. Each group object has two attributes:

* ``grouper`` -- the item that was grouped by (e.g., the string "India" or
  "Japan").
* ``list`` -- a list of all items in this group (e.g., a list of all cities
  with country='India').

Note that ``{% regroup %}`` does not order its input! Our example relies on
the fact that the ``cities`` list was ordered by ``country`` in the first place.
If the ``cities`` list did *not* order its members by ``country``, the
regrouping would naively display more than one group for a single country. For
example, say the ``cities`` list was set to this (note that the countries are not
grouped together):

.. code-block:: python

    cities = [
        {'name': 'Mumbai', 'population': '19,000,000', 'country': 'India'},
        {'name': 'New York', 'population': '20,000,000', 'country': 'USA'},
        {'name': 'Calcutta', 'population': '15,000,000', 'country': 'India'},
        {'name': 'Chicago', 'population': '7,000,000', 'country': 'USA'},
        {'name': 'Tokyo', 'population': '33,000,000', 'country': 'Japan'},
    ]

With this input for ``cities``, the example ``{% regroup %}`` template code
above would result in the following output:

* India

  * Mumbai: 19,000,000

* USA

  * New York: 20,000,000

* India

  * Calcutta: 15,000,000

* USA

  * Chicago: 7,000,000

* Japan

  * Tokyo: 33,000,000

The easiest solution to this gotcha is to make sure in your view code that the
data is ordered according to how you want to display it.

Another solution is to sort the data in the template using the
:tfilter:`dictsort` filter, if your data is in a list of dictionaries::

    {% regroup cities|dictsort:"country" by country as country_list %}

Grouping on other properties
^^^^^^^^^^^^^^^^^^^^^^^^^^^^

Any valid template lookup is a legal grouping attribute for the regroup
tag, including methods, attributes, dictionary keys and list items. For
example, if the "country" field is a foreign key to a class with
an attribute "description," you could use::

    {% regroup cities by country.description as country_list %}

Or, if ``country`` is a field with ``choices``, it will have a
:meth:`~django.db.models.Model.get_FOO_display` method available as an
attribute, allowing  you to group on the display string rather than the
``choices`` key::

    {% regroup cities by get_country_display as country_list %}

``{{ country.grouper }}`` will now display the value fields from the
``choices`` set rather than the keys.

.. templatetag:: spaceless

spaceless
^^^^^^^^^

Removes whitespace between HTML tags. This includes tab
characters and newlines.

Example usage::

    {% spaceless %}
        <p>
            <a href="foo/">Foo</a>
        </p>
    {% endspaceless %}

This example would return this HTML::

    <p><a href="foo/">Foo</a></p>

Only space between *tags* is removed -- not space between tags and text. In
this example, the space around ``Hello`` won't be stripped::

    {% spaceless %}
        <strong>
            Hello
        </strong>
    {% endspaceless %}

.. templatetag:: ssi

ssi
^^^

.. deprecated:: 1.8

    This tag has been deprecated and will be removed in Django 2.0. Use the
    :ttag:`include` tag instead.

Outputs the contents of a given file into the page.

Like a simple :ttag:`include` tag, ``{% ssi %}`` includes the contents of
another file -- which must be specified using an absolute path -- in the
current page::

    {% ssi '/home/html/ljworld.com/includes/right_generic.html' %}

The first parameter of ``ssi`` can be a quoted literal or any other context
variable.

If the optional ``parsed`` parameter is given, the contents of the included
file are evaluated as template code, within the current context::

    {% ssi '/home/html/ljworld.com/includes/right_generic.html' parsed %}

Note that if you use ``{% ssi %}``, you'll need to define
``'allowed_include_roots'`` in the :setting:`OPTIONS <TEMPLATES-OPTIONS>` of
your template engine, as a security measure.

.. note::
    With the :ttag:`ssi` tag and the ``parsed`` parameter
    there is no shared state between files -- each include is a completely
    independent rendering process. This means it's not possible for example to
    define blocks or alter the context in the current page using the included
    file.

See also: :ttag:`{% include %}<include>`.

.. templatetag:: templatetag

templatetag
^^^^^^^^^^^

Outputs one of the syntax characters used to compose template tags.

Since the template system has no concept of "escaping", to display one of the
bits used in template tags, you must use the ``{% templatetag %}`` tag.

The argument tells which template bit to output:

==================  =======
Argument            Outputs
==================  =======
``openblock``       ``{%``
``closeblock``      ``%}``
``openvariable``    ``{{``
``closevariable``   ``}}``
``openbrace``       ``{``
``closebrace``      ``}``
``opencomment``     ``{#``
``closecomment``    ``#}``
==================  =======

Sample usage::

    {% templatetag openblock %} url 'entry_list' {% templatetag closeblock %}

.. templatetag:: url

url
^^^

Returns an absolute path reference (a URL without the domain name) matching a
given view function and optional parameters. Any special characters in the
resulting path will be encoded using :func:`~django.utils.encoding.iri_to_uri`.

This is a way to output links without violating the DRY principle by having to
hard-code URLs in your templates::

    {% url 'some-url-name' v1 v2 %}

The first argument is a path to a view function in the format
``package.package.module.function``. It can be a quoted literal or any other
context variable. Additional arguments are optional and
should be space-separated values that will be used as arguments in the URL.
The example above shows passing positional arguments. Alternatively you may
use keyword syntax::

    {% url 'some-url-name' arg1=v1 arg2=v2 %}

Do not mix both positional and keyword syntax in a single call. All arguments
required by the URLconf should be present.

For example, suppose you have a view, ``app_views.client``, whose URLconf
takes a client ID (here, ``client()`` is a method inside the views file
``app_views.py``). The URLconf line might look like this:

.. code-block:: python

    ('^client/([0-9]+)/$', 'app_views.client', name='app-views-client')

If this app's URLconf is included into the project's URLconf under a path
such as this:

.. code-block:: python

    ('^clients/', include('project_name.app_name.urls'))

...then, in a template, you can create a link to this view like this::

    {% url 'app-views-client' client.id %}

The template tag will output the string ``/clients/client/123/``.

If you're using :ref:`named URL patterns <naming-url-patterns>`, you can
refer to the name of the pattern in the ``url`` tag instead of using the
path to the view.

Note that if the URL you're reversing doesn't exist, you'll get an
:exc:`~django.core.urlresolvers.NoReverseMatch` exception raised, which will
cause your site to display an error page.

If you'd like to retrieve a URL without displaying it, you can use a slightly
different call::

    {% url 'some-url-name' arg arg2 as the_url %}

    <a href="{{ the_url }}">I'm linking to {{ the_url }}</a>

The scope of the variable created by the  ``as var`` syntax is the
``{% block %}`` in which the ``{% url %}`` tag appears.

This ``{% url ... as var %}`` syntax will *not* cause an error if the view is
missing. In practice you'll use this to link to views that are optional::

    {% url 'some-url-name' as the_url %}
    {% if the_url %}
      <a href="{{ the_url }}">Link to optional stuff</a>
    {% endif %}

If you'd like to retrieve a namespaced URL, specify the fully qualified name::

    {% url 'myapp:view-name' %}

This will follow the normal :ref:`namespaced URL resolution strategy
<topics-http-reversing-url-namespaces>`, including using any hints provided
by the context as to the current application.

.. deprecated:: 1.8

    The dotted Python path syntax is deprecated and will be removed in
    Django 2.0::

        {% url 'path.to.some_view' v1 v2 %}

.. warning::

    Don't forget to put quotes around the function path or pattern name,
    otherwise the value will be interpreted as a context variable!

.. templatetag:: verbatim

verbatim
^^^^^^^^

Stops the template engine from rendering the contents of this block tag.

A common use is to allow a JavaScript template layer that collides with
Django's syntax. For example::

    {% verbatim %}
        {{if dying}}Still alive.{{/if}}
    {% endverbatim %}

You can also designate a specific closing tag, allowing the use of
``{% endverbatim %}`` as part of the unrendered contents::

    {% verbatim myblock %}
        Avoid template rendering via the {% verbatim %}{% endverbatim %} block.
    {% endverbatim myblock %}

.. templatetag:: widthratio

widthratio
^^^^^^^^^^

For creating bar charts and such, this tag calculates the ratio of a given
value to a maximum value, and then applies that ratio to a constant.

For example::

    <img src="bar.png" alt="Bar"
         height="10" width="{% widthratio this_value max_value max_width %}" />

If ``this_value`` is 175, ``max_value`` is 200, and ``max_width`` is 100, the
image in the above example will be 88 pixels wide
(because 175/200 = .875; .875 * 100 = 87.5 which is rounded up to 88).

In some cases you might want to capture the result of ``widthratio`` in a
variable. It can be useful, for instance, in a :ttag:`blocktrans` like this::

    {% widthratio this_value max_value max_width as width %}
    {% blocktrans %}The width is: {{ width }}{% endblocktrans %}

.. templatetag:: with

with
^^^^

Caches a complex variable under a simpler name. This is useful when accessing
an "expensive" method (e.g., one that hits the database) multiple times.

For example::

    {% with total=business.employees.count %}
        {{ total }} employee{{ total|pluralize }}
    {% endwith %}

The populated variable (in the example above, ``total``) is only available
between the ``{% with %}`` and ``{% endwith %}`` tags.

You can assign more than one context variable::

    {% with alpha=1 beta=2 %}
        ...
    {% endwith %}

.. note:: The previous more verbose format is still supported:
   ``{% with business.employees.count as total %}``

.. _ref-templates-builtins-filters:

Built-in filter reference
-------------------------

.. templatefilter:: add

add
^^^

Adds the argument to the value.

For example::

    {{ value|add:"2" }}

If ``value`` is ``4``, then the output will be ``6``.

This filter will first try to coerce both values to integers. If this fails,
it'll attempt to add the values together anyway. This will work on some data
types (strings, list, etc.) and fail on others. If it fails, the result will
be an empty string.

For example, if we have::

    {{ first|add:second }}

and ``first`` is ``[1, 2, 3]`` and ``second`` is ``[4, 5, 6]``, then the
output will be ``[1, 2, 3, 4, 5, 6]``.

.. warning::

    Strings that can be coerced to integers will be **summed**, not
    concatenated, as in the first example above.

.. templatefilter:: addslashes

addslashes
^^^^^^^^^^

Adds slashes before quotes. Useful for escaping strings in CSV, for example.

For example::

    {{ value|addslashes }}

If ``value`` is ``"I'm using Django"``, the output will be
``"I\'m using Django"``.

.. templatefilter:: capfirst

capfirst
^^^^^^^^

Capitalizes the first character of the value. If the first character is not
a letter, this filter has no effect.

For example::

    {{ value|capfirst }}

If ``value`` is ``"django"``, the output will be ``"Django"``.

.. templatefilter:: center

center
^^^^^^

Centers the value in a field of a given width.

For example::

    "{{ value|center:"15" }}"

If ``value`` is ``"Django"``, the output will be ``"     Django    "``.

.. templatefilter:: cut

cut
^^^

Removes all values of arg from the given string.

For example::

    {{ value|cut:" " }}

If ``value`` is ``"String with spaces"``, the output will be
``"Stringwithspaces"``.

.. templatefilter:: date

date
^^^^

Formats a date according to the given format.

Uses a similar format as PHP's ``date()`` function (http://php.net/date)
with some differences.

.. note::
    These format characters are not used in Django outside of templates. They
    were designed to be compatible with PHP to ease transitioning for designers.

.. _date-and-time-formatting-specifiers:

Available format strings:

================  ========================================  =====================
Format character  Description                               Example output
================  ========================================  =====================
a                 ``'a.m.'`` or ``'p.m.'`` (Note that       ``'a.m.'``
                  this is slightly different than PHP's
                  output, because this includes periods
                  to match Associated Press style.)
A                 ``'AM'`` or ``'PM'``.                     ``'AM'``
b                 Month, textual, 3 letters, lowercase.     ``'jan'``
B                 Not implemented.
c                 ISO 8601 format. (Note: unlike others     ``2008-01-02T10:30:00.000123+02:00``,
                  formatters, such as "Z", "O" or "r",      or ``2008-01-02T10:30:00.000123`` if the datetime is naive
                  the "c" formatter will not add timezone
                  offset if value is a naive datetime
                  (see :class:`datetime.tzinfo`).
d                 Day of the month, 2 digits with           ``'01'`` to ``'31'``
                  leading zeros.
D                 Day of the week, textual, 3 letters.      ``'Fri'``
e                 Timezone name. Could be in any format,
                  or might return an empty string,          ``''``, ``'GMT'``, ``'-500'``, ``'US/Eastern'``, etc.
                  depending on the datetime.
E                 Month, locale specific alternative
                  representation usually used for long
                  date representation.                      ``'listopada'`` (for Polish locale, as opposed to ``'Listopad'``)
f                 Time, in 12-hour hours and minutes,       ``'1'``, ``'1:30'``
                  with minutes left off if they're zero.
                  Proprietary extension.
F                 Month, textual, long.                     ``'January'``
g                 Hour, 12-hour format without leading      ``'1'`` to ``'12'``
                  zeros.
G                 Hour, 24-hour format without leading      ``'0'`` to ``'23'``
                  zeros.
h                 Hour, 12-hour format.                     ``'01'`` to ``'12'``
H                 Hour, 24-hour format.                     ``'00'`` to ``'23'``
i                 Minutes.                                  ``'00'`` to ``'59'``
I                 Daylight Savings Time, whether it's       ``'1'`` or ``'0'``
                  in effect or not.
j                 Day of the month without leading          ``'1'`` to ``'31'``
                  zeros.
l                 Day of the week, textual, long.           ``'Friday'``
L                 Boolean for whether it's a leap year.     ``True`` or ``False``
m                 Month, 2 digits with leading zeros.       ``'01'`` to ``'12'``
M                 Month, textual, 3 letters.                ``'Jan'``
n                 Month without leading zeros.              ``'1'`` to ``'12'``
N                 Month abbreviation in Associated Press    ``'Jan.'``, ``'Feb.'``, ``'March'``, ``'May'``
                  style. Proprietary extension.
o                 ISO-8601 week-numbering year,             ``'1999'``
                  corresponding to
                  the ISO-8601 week number (W)
O                 Difference to Greenwich time in hours.    ``'+0200'``
P                 Time, in 12-hour hours, minutes and       ``'1 a.m.'``, ``'1:30 p.m.'``, ``'midnight'``, ``'noon'``, ``'12:30 p.m.'``
                  'a.m.'/'p.m.', with minutes left off
                  if they're zero and the special-case
                  strings 'midnight' and 'noon' if
                  appropriate. Proprietary extension.
r                 :rfc:`2822` formatted date.               ``'Thu, 21 Dec 2000 16:01:07 +0200'``
s                 Seconds, 2 digits with leading zeros.     ``'00'`` to ``'59'``
S                 English ordinal suffix for day of the     ``'st'``, ``'nd'``, ``'rd'`` or ``'th'``
                  month, 2 characters.
t                 Number of days in the given month.        ``28`` to ``31``
T                 Time zone of this machine.                ``'EST'``, ``'MDT'``
u                 Microseconds.                             ``000000`` to ``999999``
U                 Seconds since the Unix Epoch
                  (January 1 1970 00:00:00 UTC).
w                 Day of the week, digits without           ``'0'`` (Sunday) to ``'6'`` (Saturday)
                  leading zeros.
W                 ISO-8601 week number of year, with        ``1``, ``53``
                  weeks starting on Monday.
y                 Year, 2 digits.                           ``'99'``
Y                 Year, 4 digits.                           ``'1999'``
z                 Day of the year.                          ``0`` to ``365``
Z                 Time zone offset in seconds. The          ``-43200`` to ``43200``
                  offset for timezones west of UTC is
                  always negative, and for those east of
                  UTC is always positive.
================  ========================================  =====================

For example::

    {{ value|date:"D d M Y" }}

If ``value`` is a :py:class:`~datetime.datetime` object (e.g., the result of
``datetime.datetime.now()``), the output will be the string
``'Wed 09 Jan 2008'``.

The format passed can be one of the predefined ones :setting:`DATE_FORMAT`,
:setting:`DATETIME_FORMAT`, :setting:`SHORT_DATE_FORMAT` or
:setting:`SHORT_DATETIME_FORMAT`, or a custom format that uses the format
specifiers shown in the table above. Note that predefined formats may vary
depending on the current locale.

Assuming that :setting:`USE_L10N` is ``True`` and :setting:`LANGUAGE_CODE` is,
for example, ``"es"``, then for::

    {{ value|date:"SHORT_DATE_FORMAT" }}

the output would be the string ``"09/01/2008"`` (the ``"SHORT_DATE_FORMAT"``
format specifier for the ``es`` locale as shipped with Django is ``"d/m/Y"``).

When used without a format string::

    {{ value|date }}

...the formatting string defined in the :setting:`DATE_FORMAT` setting will be
used, without applying any localization.

You can combine ``date`` with the :tfilter:`time` filter to render a full
representation of a ``datetime`` value. E.g.::

    {{ value|date:"D d M Y" }} {{ value|time:"H:i" }}

.. templatefilter:: default

default
^^^^^^^

If value evaluates to ``False``, uses the given default. Otherwise, uses the
value.

For example::

    {{ value|default:"nothing" }}

If ``value`` is ``""`` (the empty string), the output will be ``nothing``.

.. templatefilter:: default_if_none

default_if_none
^^^^^^^^^^^^^^^

If (and only if) value is ``None``, uses the given default. Otherwise, uses the
value.

Note that if an empty string is given, the default value will *not* be used.
Use the :tfilter:`default` filter if you want to fallback for empty strings.

For example::

    {{ value|default_if_none:"nothing" }}

If ``value`` is ``None``, the output will be the string ``"nothing"``.

.. templatefilter:: dictsort

dictsort
^^^^^^^^

Takes a list of dictionaries and returns that list sorted by the key given in
the argument.

For example::

    {{ value|dictsort:"name" }}

If ``value`` is:

.. code-block:: python

    [
        {'name': 'zed', 'age': 19},
        {'name': 'amy', 'age': 22},
        {'name': 'joe', 'age': 31},
    ]

then the output would be:

.. code-block:: python

    [
        {'name': 'amy', 'age': 22},
        {'name': 'joe', 'age': 31},
        {'name': 'zed', 'age': 19},
    ]

You can also do more complicated things like::

    {% for book in books|dictsort:"author.age" %}
        * {{ book.title }} ({{ book.author.name }})
    {% endfor %}

If ``books`` is:

.. code-block:: python

    [
        {'title': '1984', 'author': {'name': 'George', 'age': 45}},
        {'title': 'Timequake', 'author': {'name': 'Kurt', 'age': 75}},
        {'title': 'Alice', 'author': {'name': 'Lewis', 'age': 33}},
    ]

then the output would be::

    * Alice (Lewis)
    * 1984 (George)
    * Timequake (Kurt)

.. templatefilter:: dictsortreversed

dictsortreversed
^^^^^^^^^^^^^^^^

Takes a list of dictionaries and returns that list sorted in reverse order by
the key given in the argument. This works exactly the same as the above filter,
but the returned value will be in reverse order.

.. templatefilter:: divisibleby

divisibleby
^^^^^^^^^^^

Returns ``True`` if the value is divisible by the argument.

For example::

    {{ value|divisibleby:"3" }}

If ``value`` is ``21``, the output would be ``True``.

.. templatefilter:: escape

escape
^^^^^^

Escapes a string's HTML. Specifically, it makes these replacements:

* ``<`` is converted to ``&lt;``
* ``>`` is converted to ``&gt;``
* ``'`` (single quote) is converted to ``&#39;``
* ``"`` (double quote) is converted to ``&quot;``
* ``&`` is converted to ``&amp;``

The escaping is only applied when the string is output, so it does not matter
where in a chained sequence of filters you put ``escape``: it will always be
applied as though it were the last filter. If you want escaping to be applied
immediately, use the :tfilter:`force_escape` filter.

Applying ``escape`` to a variable that would normally have auto-escaping
applied to the result will only result in one round of escaping being done. So
it is safe to use this function even in auto-escaping environments. If you want
multiple escaping passes to be applied, use the :tfilter:`force_escape` filter.

For example, you can apply ``escape`` to fields when :ttag:`autoescape` is off::

    {% autoescape off %}
        {{ title|escape }}
    {% endautoescape %}

.. templatefilter:: escapejs

escapejs
^^^^^^^^

Escapes characters for use in JavaScript strings. This does *not* make the
string safe for use in HTML, but does protect you from syntax errors when using
templates to generate JavaScript/JSON.

For example::

    {{ value|escapejs }}

If ``value`` is ``"testing\r\njavascript \'string" <b>escaping</b>"``,
the output will be ``"testing\\u000D\\u000Ajavascript \\u0027string\\u0022 \\u003Cb\\u003Eescaping\\u003C/b\\u003E"``.

.. templatefilter:: filesizeformat

filesizeformat
^^^^^^^^^^^^^^

Formats the value like a 'human-readable' file size (i.e. ``'13 KB'``,
``'4.1 MB'``, ``'102 bytes'``, etc).

For example::

    {{ value|filesizeformat }}

If ``value`` is 123456789, the output would be ``117.7 MB``.

.. admonition:: File sizes and SI units

    Strictly speaking, ``filesizeformat`` does not conform to the International
    System of Units which recommends using KiB, MiB, GiB, etc. when byte sizes
    are calculated in powers of 1024 (which is the case here). Instead, Django
    uses traditional unit names (KB, MB, GB, etc.) corresponding to names that
    are more commonly used.

.. templatefilter:: first

first
^^^^^

Returns the first item in a list.

For example::

    {{ value|first }}

If ``value`` is the list ``['a', 'b', 'c']``, the output will be ``'a'``.

.. templatefilter:: floatformat

floatformat
^^^^^^^^^^^

When used without an argument, rounds a floating-point number to one decimal
place -- but only if there's a decimal part to be displayed. For example:

============  ===========================  ========
``value``     Template                     Output
============  ===========================  ========
``34.23234``  ``{{ value|floatformat }}``  ``34.2``
``34.00000``  ``{{ value|floatformat }}``  ``34``
``34.26000``  ``{{ value|floatformat }}``  ``34.3``
============  ===========================  ========

If used with a numeric integer argument, ``floatformat`` rounds a number to
that many decimal places. For example:

============  =============================  ==========
``value``     Template                       Output
============  =============================  ==========
``34.23234``  ``{{ value|floatformat:3 }}``  ``34.232``
``34.00000``  ``{{ value|floatformat:3 }}``  ``34.000``
``34.26000``  ``{{ value|floatformat:3 }}``  ``34.260``
============  =============================  ==========

Particularly useful is passing 0 (zero) as the argument which will round the
float to the nearest integer.

============  ================================  ==========
``value``     Template                          Output
============  ================================  ==========
``34.23234``  ``{{ value|floatformat:"0" }}``   ``34``
``34.00000``  ``{{ value|floatformat:"0" }}``   ``34``
``39.56000``  ``{{ value|floatformat:"0" }}``   ``40``
============  ================================  ==========

If the argument passed to ``floatformat`` is negative, it will round a number
to that many decimal places -- but only if there's a decimal part to be
displayed. For example:

============  ================================  ==========
``value``     Template                          Output
============  ================================  ==========
``34.23234``  ``{{ value|floatformat:"-3" }}``  ``34.232``
``34.00000``  ``{{ value|floatformat:"-3" }}``  ``34``
``34.26000``  ``{{ value|floatformat:"-3" }}``  ``34.260``
============  ================================  ==========

Using ``floatformat`` with no argument is equivalent to using ``floatformat``
with an argument of ``-1``.

.. templatefilter:: force_escape

force_escape
^^^^^^^^^^^^

Applies HTML escaping to a string (see the :tfilter:`escape` filter for
details). This filter is applied *immediately* and returns a new, escaped
string. This is useful in the rare cases where you need multiple escaping or
want to apply other filters to the escaped results. Normally, you want to use
the :tfilter:`escape` filter.

For example, if you want to catch the ``<p>`` HTML elements created by
the :tfilter:`linebreaks` filter::

    {% autoescape off %}
        {{ body|linebreaks|force_escape }}
    {% endautoescape %}

.. templatefilter:: get_digit

get_digit
^^^^^^^^^

Given a whole number, returns the requested digit, where 1 is the right-most
digit, 2 is the second-right-most digit, etc. Returns the original value for
invalid input (if input or argument is not an integer, or if argument is less
than 1). Otherwise, output is always an integer.

For example::

    {{ value|get_digit:"2" }}

If ``value`` is ``123456789``, the output will be ``8``.

.. templatefilter:: iriencode

iriencode
^^^^^^^^^

Converts an IRI (Internationalized Resource Identifier) to a string that is
suitable for including in a URL. This is necessary if you're trying to use
strings containing non-ASCII characters in a URL.

It's safe to use this filter on a string that has already gone through the
:tfilter:`urlencode` filter.

For example::

    {{ value|iriencode }}

If ``value`` is ``"?test=1&me=2"``, the output will be ``"?test=1&amp;me=2"``.

.. templatefilter:: join

join
^^^^

Joins a list with a string, like Python's ``str.join(list)``

For example::

    {{ value|join:" // " }}

If ``value`` is the list ``['a', 'b', 'c']``, the output will be the string
``"a // b // c"``.

.. templatefilter:: last

last
^^^^

Returns the last item in a list.

For example::

    {{ value|last }}

If ``value`` is the list ``['a', 'b', 'c', 'd']``, the output will be the
string ``"d"``.

.. templatefilter:: length

length
^^^^^^

Returns the length of the value. This works for both strings and lists.

For example::

    {{ value|length }}

If ``value`` is ``['a', 'b', 'c', 'd']`` or ``"abcd"``, the output will be
``4``.

.. versionchanged:: 1.8

    The filter returns ``0`` for an undefined variable. Previously, it returned
    an empty string.

.. templatefilter:: length_is

length_is
^^^^^^^^^

Returns ``True`` if the value's length is the argument, or ``False`` otherwise.

For example::

    {{ value|length_is:"4" }}

If ``value`` is ``['a', 'b', 'c', 'd']`` or ``"abcd"``, the output will be
``True``.

.. templatefilter:: linebreaks

linebreaks
^^^^^^^^^^

Replaces line breaks in plain text with appropriate HTML; a single
newline becomes an HTML line break (``<br />``) and a new line
followed by a blank line becomes a paragraph break (``</p>``).

For example::

    {{ value|linebreaks }}

If ``value`` is ``Joel\nis a slug``, the output will be ``<p>Joel<br />is a
slug</p>``.

.. templatefilter:: linebreaksbr

linebreaksbr
^^^^^^^^^^^^

Converts all newlines in a piece of plain text to HTML line breaks
(``<br />``).

For example::

    {{ value|linebreaksbr }}

If ``value`` is ``Joel\nis a slug``, the output will be ``Joel<br />is a
slug``.

.. templatefilter:: linenumbers

linenumbers
^^^^^^^^^^^

Displays text with line numbers.

For example::

    {{ value|linenumbers }}

If ``value`` is::

    one
    two
    three

the output will be::

    1. one
    2. two
    3. three

.. templatefilter:: ljust

ljust
^^^^^

Left-aligns the value in a field of a given width.

**Argument:** field size

For example::

    "{{ value|ljust:"10" }}"

If ``value`` is ``Django``, the output will be ``"Django    "``.

.. templatefilter:: lower

lower
^^^^^

Converts a string into all lowercase.

For example::

    {{ value|lower }}

If ``value`` is ``Still MAD At Yoko``, the output will be
``still mad at yoko``.

.. templatefilter:: make_list

make_list
^^^^^^^^^

Returns the value turned into a list. For a string, it's a list of characters.
For an integer, the argument is cast into an unicode string before creating a
list.

For example::

    {{ value|make_list }}

If ``value`` is the string ``"Joel"``, the output would be the list
``['J', 'o', 'e', 'l']``. If ``value`` is ``123``, the output will be the
list ``['1', '2', '3']``.

.. templatefilter:: phone2numeric

phone2numeric
^^^^^^^^^^^^^

Converts a phone number (possibly containing letters) to its numerical
equivalent.

The input doesn't have to be a valid phone number. This will happily convert
any string.

For example::

    {{ value|phone2numeric }}

If ``value`` is ``800-COLLECT``, the output will be ``800-2655328``.

.. templatefilter:: pluralize

pluralize
^^^^^^^^^

Returns a plural suffix if the value is not 1. By default, this suffix is
``'s'``.

Example::

    You have {{ num_messages }} message{{ num_messages|pluralize }}.

If ``num_messages`` is ``1``, the output will be ``You have 1 message.``
If ``num_messages`` is ``2``  the output will be ``You have 2 messages.``

For words that require a suffix other than ``'s'``, you can provide an alternate
suffix as a parameter to the filter.

Example::

    You have {{ num_walruses }} walrus{{ num_walruses|pluralize:"es" }}.

For words that don't pluralize by simple suffix, you can specify both a
singular and plural suffix, separated by a comma.

Example::

    You have {{ num_cherries }} cherr{{ num_cherries|pluralize:"y,ies" }}.

.. note:: Use :ttag:`blocktrans` to pluralize translated strings.

.. templatefilter:: pprint

pprint
^^^^^^

A wrapper around :func:`pprint.pprint` -- for debugging, really.

.. templatefilter:: random

random
^^^^^^

Returns a random item from the given list.

For example::

    {{ value|random }}

If ``value`` is the list ``['a', 'b', 'c', 'd']``, the output could be ``"b"``.

.. templatefilter:: removetags

removetags
^^^^^^^^^^

.. deprecated:: 1.8

    ``removetags`` cannot guarantee HTML safe output and has been deprecated due
    to security concerns. Consider using `bleach`_ instead.

.. _bleach: http://bleach.readthedocs.org/en/latest/

Removes a space-separated list of [X]HTML tags from the output.

For example::

    {{ value|removetags:"b span" }}

If ``value`` is ``"<b>Joel</b> <button>is</button> a <span>slug</span>"`` the
unescaped output will be ``"Joel <button>is</button> a slug"``.

Note that this filter is case-sensitive.

If ``value`` is ``"<B>Joel</B> <button>is</button> a <span>slug</span>"`` the
unescaped output will be ``"<B>Joel</B> <button>is</button> a slug"``.

.. admonition:: No safety guarantee

    Note that ``removetags`` doesn't give any guarantee about its output being
    HTML safe. In particular, it doesn't work recursively, so an input like
    ``"<sc<script>ript>alert('XSS')</sc</script>ript>"`` won't be safe even if
    you apply ``|removetags:"script"``. So if the input is user provided,
    **NEVER** apply the ``safe`` filter to a ``removetags`` output. If you are
    looking for something more robust, you can use the ``bleach`` Python
    library, notably its `clean`_ method.

.. _clean: http://bleach.readthedocs.org/en/latest/clean.html

.. templatefilter:: rjust

rjust
^^^^^

Right-aligns the value in a field of a given width.

**Argument:** field size

For example::

    "{{ value|rjust:"10" }}"

If ``value`` is ``Django``, the output will be ``"    Django"``.

.. templatefilter:: safe

safe
^^^^

Marks a string as not requiring further HTML escaping prior to output. When
autoescaping is off, this filter has no effect.

.. note::

    If you are chaining filters, a filter applied after ``safe`` can
    make the contents unsafe again. For example, the following code
    prints the variable as is, unescaped:

    .. code-block:: html+django

        {{ var|safe|escape }}

.. templatefilter:: safeseq

safeseq
^^^^^^^

Applies the :tfilter:`safe` filter to each element of a sequence. Useful in
conjunction with other filters that operate on sequences, such as
:tfilter:`join`. For example::

    {{ some_list|safeseq|join:", " }}

You couldn't use the :tfilter:`safe` filter directly in this case, as it would
first convert the variable into a string, rather than working with the
individual elements of the sequence.

.. templatefilter:: slice

slice
^^^^^

Returns a slice of the list.

Uses the same syntax as Python's list slicing. See
http://www.diveintopython3.net/native-datatypes.html#slicinglists
for an introduction.

Example::

    {{ some_list|slice:":2" }}

If ``some_list`` is ``['a', 'b', 'c']``, the output will be ``['a', 'b']``.

.. templatefilter:: slugify

slugify
^^^^^^^

Converts to ASCII. Converts spaces to hyphens. Removes characters that aren't
alphanumerics, underscores, or hyphens. Converts to lowercase. Also strips
leading and trailing whitespace.

For example::

    {{ value|slugify }}

If ``value`` is ``"Joel is a slug"``, the output will be ``"joel-is-a-slug"``.

.. templatefilter:: stringformat

stringformat
^^^^^^^^^^^^

Formats the variable according to the argument, a string formatting specifier.
This specifier uses Python string formatting syntax, with the exception that
the leading "%" is dropped.

See https://docs.python.org/library/stdtypes.html#string-formatting-operations
for documentation of Python string formatting

For example::

    {{ value|stringformat:"E" }}

If ``value`` is ``10``, the output will be ``1.000000E+01``.

.. templatefilter:: striptags

striptags
^^^^^^^^^

Makes all possible efforts to strip all [X]HTML tags.

For example::

    {{ value|striptags }}

If ``value`` is ``"<b>Joel</b> <button>is</button> a <span>slug</span>"``, the
output will be ``"Joel is a slug"``.

.. admonition:: No safety guarantee

    Note that ``striptags`` doesn't give any guarantee about its output being
    HTML safe, particularly with non valid HTML input. So **NEVER** apply the
    ``safe`` filter to a ``striptags`` output. If you are looking for something
    more robust, you can use the ``bleach`` Python library, notably its
    `clean`_ method.

.. _clean: http://bleach.readthedocs.org/en/latest/clean.html

.. templatefilter:: time

time
^^^^

Formats a time according to the given format.

Given format can be the predefined one :setting:`TIME_FORMAT`, or a custom
format, same as the :tfilter:`date` filter. Note that the predefined format
is locale-dependent.

For example::

    {{ value|time:"H:i" }}

If ``value`` is equivalent to ``datetime.datetime.now()``, the output will be
the string ``"01:23"``.

Another example:

Assuming that :setting:`USE_L10N` is ``True`` and :setting:`LANGUAGE_CODE` is,
for example, ``"de"``, then for::

    {{ value|time:"TIME_FORMAT" }}

the output will be the string ``"01:23:00"`` (The ``"TIME_FORMAT"`` format
specifier for the ``de`` locale as shipped with Django is ``"H:i:s"``).

The ``time`` filter will only accept parameters in the format string that
relate to the time of day, not the date (for obvious reasons). If you need to
format a ``date`` value, use the :tfilter:`date` filter instead (or along
``time`` if you need to render a full :py:class:`~datetime.datetime` value).

There is one exception the above rule: When passed a ``datetime`` value with
attached timezone information (a :ref:`time-zone-aware
<naive_vs_aware_datetimes>` ``datetime`` instance) the ``time`` filter will
accept the timezone-related :ref:`format specifiers
<date-and-time-formatting-specifiers>` ``'e'``, ``'O'`` , ``'T'`` and ``'Z'``.

When used without a format string::

    {{ value|time }}

...the formatting string defined in the :setting:`TIME_FORMAT` setting will be
used, without applying any localization.

.. templatefilter:: timesince

timesince
^^^^^^^^^

Formats a date as the time since that date (e.g., "4 days, 6 hours").

Takes an optional argument that is a variable containing the date to use as
the comparison point (without the argument, the comparison point is *now*).
For example, if ``blog_date`` is a date instance representing midnight on 1
June 2006, and ``comment_date`` is a date instance for 08:00 on 1 June 2006,
then the following would return "8 hours"::

    {{ blog_date|timesince:comment_date }}

Comparing offset-naive and offset-aware datetimes will return an empty string.

Minutes is the smallest unit used, and "0 minutes" will be returned for any
date that is in the future relative to the comparison point.

.. templatefilter:: timeuntil

timeuntil
^^^^^^^^^

Similar to ``timesince``, except that it measures the time from now until the
given date or datetime. For example, if today is 1 June 2006 and
``conference_date`` is a date instance holding 29 June 2006, then
``{{ conference_date|timeuntil }}`` will return "4 weeks".

Takes an optional argument that is a variable containing the date to use as
the comparison point (instead of *now*). If ``from_date`` contains 22 June
2006, then the following will return "1 week"::

    {{ conference_date|timeuntil:from_date }}

Comparing offset-naive and offset-aware datetimes will return an empty string.

Minutes is the smallest unit used, and "0 minutes" will be returned for any
date that is in the past relative to the comparison point.

.. templatefilter:: title

title
^^^^^

Converts a string into titlecase by making words start with an uppercase
character and the remaining characters lowercase. This tag makes no effort to
keep "trivial words" in lowercase.

For example::

    {{ value|title }}

If ``value`` is ``"my FIRST post"``, the output will be ``"My First Post"``.

.. templatefilter:: truncatechars

truncatechars
^^^^^^^^^^^^^

Truncates a string if it is longer than the specified number of characters.
Truncated strings will end with a translatable ellipsis sequence ("...").

**Argument:** Number of characters to truncate to

For example::

    {{ value|truncatechars:9 }}

If ``value`` is ``"Joel is a slug"``, the output will be ``"Joel i..."``.

.. templatefilter:: truncatechars_html

truncatechars_html
^^^^^^^^^^^^^^^^^^

Similar to :tfilter:`truncatechars`, except that it is aware of HTML tags. Any
tags that are opened in the string and not closed before the truncation point
are closed immediately after the truncation.

For example::

    {{ value|truncatechars_html:9 }}

If ``value`` is ``"<p>Joel is a slug</p>"``, the output will be
``"<p>Joel i...</p>"``.

Newlines in the HTML content will be preserved.

.. templatefilter:: truncatewords

truncatewords
^^^^^^^^^^^^^

Truncates a string after a certain number of words.

**Argument:** Number of words to truncate after

For example::

    {{ value|truncatewords:2 }}

If ``value`` is ``"Joel is a slug"``, the output will be ``"Joel is ..."``.

Newlines within the string will be removed.

.. templatefilter:: truncatewords_html

truncatewords_html
^^^^^^^^^^^^^^^^^^

Similar to :tfilter:`truncatewords`, except that it is aware of HTML tags. Any
tags that are opened in the string and not closed before the truncation point,
are closed immediately after the truncation.

This is less efficient than :tfilter:`truncatewords`, so should only be used
when it is being passed HTML text.

For example::

    {{ value|truncatewords_html:2 }}

If ``value`` is ``"<p>Joel is a slug</p>"``, the output will be
``"<p>Joel is ...</p>"``.

Newlines in the HTML content will be preserved.

.. templatefilter:: unordered_list

unordered_list
^^^^^^^^^^^^^^

Recursively takes a self-nested list and returns an HTML unordered list --
WITHOUT opening and closing <ul> tags.

The list is assumed to be in the proper format. For example, if ``var``
contains ``['States', ['Kansas', ['Lawrence', 'Topeka'], 'Illinois']]``, then
``{{ var|unordered_list }}`` would return::

    <li>States
    <ul>
            <li>Kansas
            <ul>
                    <li>Lawrence</li>
                    <li>Topeka</li>
            </ul>
            </li>
            <li>Illinois</li>
    </ul>
    </li>

.. deprecated:: 1.8

    An older, more restrictive and verbose input format is also supported:
    ``['States', [['Kansas', [['Lawrence', []], ['Topeka', []]]], ['Illinois', []]]]``.
    Support for this syntax will be removed in Django 2.0.

.. templatefilter:: upper

upper
^^^^^

Converts a string into all uppercase.

For example::

    {{ value|upper }}

If ``value`` is ``"Joel is a slug"``, the output will be ``"JOEL IS A SLUG"``.

.. templatefilter:: urlencode

urlencode
^^^^^^^^^

Escapes a value for use in a URL.

For example::

    {{ value|urlencode }}

If ``value`` is ``"http://www.example.org/foo?a=b&c=d"``, the output will be
``"http%3A//www.example.org/foo%3Fa%3Db%26c%3Dd"``.

An optional argument containing the characters which should not be escaped can
be provided.

If not provided, the '/' character is assumed safe. An empty string can be
provided when *all* characters should be escaped. For example::

    {{ value|urlencode:"" }}

If ``value`` is ``"http://www.example.org/"``, the output will be
``"http%3A%2F%2Fwww.example.org%2F"``.

.. templatefilter:: urlize

urlize
^^^^^^

Converts URLs and email addresses in text into clickable links.

This template tag works on links prefixed with ``http://``, ``https://``, or
``www.``. For example, ``http://goo.gl/aia1t`` will get converted but
``goo.gl/aia1t`` won't.

It also supports domain-only links ending in one of the original top level
domains (``.com``, ``.edu``, ``.gov``, ``.int``, ``.mil``, ``.net``, and
``.org``). For example, ``djangoproject.com`` gets converted.

.. versionchanged:: 1.8

    Support for domain-only links that include characters after the top-level
    domain (e.g. ``djangoproject.com/`` and ``djangoproject.com/download/``)
    was added.

Links can have trailing punctuation (periods, commas, close-parens) and leading
punctuation (opening parens), and ``urlize`` will still do the right thing.

Links generated by ``urlize`` have a ``rel="nofollow"`` attribute added
to them.

For example::

    {{ value|urlize }}

If ``value`` is ``"Check out www.djangoproject.com"``, the output will be
``"Check out <a href="http://www.djangoproject.com"
rel="nofollow">www.djangoproject.com</a>"``.

In addition to web links, ``urlize`` also converts email addresses into
``mailto:`` links. If ``value`` is
``"Send questions to foo@example.com"``, the output will be
``"Send questions to <a href="mailto:foo@example.com">foo@example.com</a>"``.

The ``urlize`` filter also takes an optional parameter ``autoescape``. If
``autoescape`` is ``True``, the link text and URLs will be escaped using
Django's built-in :tfilter:`escape` filter. The default value for
``autoescape`` is ``True``.

.. note::

    If ``urlize`` is applied to text that already contains HTML markup,
    things won't work as expected. Apply this filter only to plain text.

.. templatefilter:: urlizetrunc

urlizetrunc
^^^^^^^^^^^

Converts URLs and email addresses into clickable links just like urlize_, but truncates URLs
longer than the given character limit.

**Argument:** Number of characters that link text should be truncated to,
including the ellipsis that's added if truncation is necessary.

For example::

    {{ value|urlizetrunc:15 }}

If ``value`` is ``"Check out www.djangoproject.com"``, the output would be
``'Check out <a href="http://www.djangoproject.com"
rel="nofollow">www.djangopr...</a>'``.

As with urlize_, this filter should only be applied to plain text.

.. templatefilter:: wordcount

wordcount
^^^^^^^^^

Returns the number of words.

For example::

    {{ value|wordcount }}

If ``value`` is ``"Joel is a slug"``, the output will be ``4``.

.. templatefilter:: wordwrap

wordwrap
^^^^^^^^

Wraps words at specified line length.

**Argument:** number of characters at which to wrap the text

For example::

    {{ value|wordwrap:5 }}

If ``value`` is ``Joel is a slug``, the output would be::

    Joel
    is a
    slug

.. templatefilter:: yesno

yesno
^^^^^

Maps values for ``True``, ``False``, and (optionally) ``None``, to the strings
"yes", "no", "maybe", or a custom mapping passed as a comma-separated list, and
returns one of those strings according to the value:

For example::

    {{ value|yesno:"yeah,no,maybe" }}

==========  ======================  ===========================================
Value       Argument                Outputs
==========  ======================  ===========================================
``True``                            ``yes``
``True``    ``"yeah,no,maybe"``     ``yeah``
``False``   ``"yeah,no,maybe"``     ``no``
``None``    ``"yeah,no,maybe"``     ``maybe``
``None``    ``"yeah,no"``           ``no`` (converts ``None`` to ``False``
                                    if no mapping for ``None`` is given)
==========  ======================  ===========================================

Internationalization tags and filters
-------------------------------------

Django provides template tags and filters to control each aspect of
:doc:`internationalization </topics/i18n/index>` in templates. They allow for
granular control of translations, formatting, and time zone conversions.

i18n
^^^^

This library allows specifying translatable text in templates.
To enable it, set :setting:`USE_I18N` to ``True``, then load it with
``{% load i18n %}``.

See :ref:`specifying-translation-strings-in-template-code`.

l10n
^^^^

This library provides control over the localization of values in templates.
You only need to load the library using ``{% load l10n %}``, but you'll often
set :setting:`USE_L10N` to ``True`` so that localization is active by default.

See :ref:`topic-l10n-templates`.

tz
^^

This library provides control over time zone conversions in templates.
Like ``l10n``, you only need to load the library using ``{% load tz %}``,
but you'll usually also set :setting:`USE_TZ` to ``True`` so that conversion
to local time happens by default.

See :ref:`time-zones-in-templates`.

Other tags and filters libraries
--------------------------------

Django comes with a couple of other template-tag libraries that you have to
enable explicitly in your :setting:`INSTALLED_APPS` setting and enable in your
template with the :ttag:`{% load %}<load>` tag.

django.contrib.humanize
^^^^^^^^^^^^^^^^^^^^^^^

A set of Django template filters useful for adding a "human touch" to data. See
:doc:`/ref/contrib/humanize`.

static
^^^^^^

.. templatetag:: static

static
""""""

.. highlight:: html+django

To link to static files that are saved in :setting:`STATIC_ROOT` Django ships
with a :ttag:`static` template tag. You can use this regardless if you're
using :class:`~django.template.RequestContext` or not.

.. code-block:: html+django

    {% load static %}
    <img src="{% static "images/hi.jpg" %}" alt="Hi!" />

It is also able to consume standard context variables, e.g. assuming a
``user_stylesheet`` variable is passed to the template:

.. code-block:: html+django

    {% load static %}
    <link rel="stylesheet" href="{% static user_stylesheet %}" type="text/css" media="screen" />

If you'd like to retrieve a static URL without displaying it, you can use a
slightly different call:

.. code-block:: html+django

    {% load static %}
    {% static "images/hi.jpg" as myphoto %}
    <img src="{{ myphoto }}"></img>

.. note::

    The :mod:`staticfiles<django.contrib.staticfiles>` contrib app also ships
    with a :ttag:`static template tag<staticfiles-static>` which uses
    ``staticfiles'`` :setting:`STATICFILES_STORAGE` to build the URL of the
    given path (rather than simply using :func:`urllib.parse.urljoin` with the
    :setting:`STATIC_URL` setting and the given path). Use that instead if you
    have an advanced use case such as :ref:`using a cloud service to serve
    static files<staticfiles-from-cdn>`::

        {% load static from staticfiles %}
        <img src="{% static "images/hi.jpg" %}" alt="Hi!" />

.. templatetag:: get_static_prefix

get_static_prefix
"""""""""""""""""

.. highlight:: html+django

You should prefer the :ttag:`static` template tag, but if you need more control
over exactly where and how :setting:`STATIC_URL` is injected into the template,
you can use the :ttag:`get_static_prefix` template tag::

    {% load static %}
    <img src="{% get_static_prefix %}images/hi.jpg" alt="Hi!" />

There's also a second form you can use to avoid extra processing if you need
the value multiple times::

    {% load static %}
    {% get_static_prefix as STATIC_PREFIX %}

    <img src="{{ STATIC_PREFIX }}images/hi.jpg" alt="Hi!" />
    <img src="{{ STATIC_PREFIX }}images/hi2.jpg" alt="Hello!" />

.. templatetag:: get_media_prefix

get_media_prefix
""""""""""""""""

.. highlight:: html+django

Similar to the :ttag:`get_static_prefix`, ``get_media_prefix`` populates a
template variable with the media prefix :setting:`MEDIA_URL`, e.g.::

    {% load static %}
    <body data-media-url="{% get_media_prefix %}">

By storing the value in a data attribute, we ensure it's escaped appropriately
if we want to use it in a JavaScript context.