custom-model-fields.txt

===========================
Writing custom model fields
===========================

.. currentmodule:: django.db.models

Introduction
============

The :doc:`model reference </topics/db/models>` documentation explains how to use
Django's standard field classes -- :class:`~django.db.models.CharField`,
:class:`~django.db.models.DateField`, etc. For many purposes, those classes are
all you'll need. Sometimes, though, the Django version won't meet your precise
requirements, or you'll want to use a field that is entirely different from
those shipped with Django.

Django's built-in field types don't cover every possible database column type --
only the common types, such as ``VARCHAR`` and ``INTEGER``. For more obscure
column types, such as geographic polygons or even user-created types such as
`PostgreSQL custom types`_, you can define your own Django ``Field`` subclasses.

.. _PostgreSQL custom types: http://www.postgresql.org/docs/8.2/interactive/sql-createtype.html

Alternatively, you may have a complex Python object that can somehow be
serialized to fit into a standard database column type. This is another case
where a ``Field`` subclass will help you use your object with your models.

Our example object
------------------

Creating custom fields requires a bit of attention to detail. To make things
easier to follow, we'll use a consistent example throughout this document:
wrapping a Python object representing the deal of cards in a hand of Bridge_.
Don't worry, you don't have know how to play Bridge to follow this example.
You only need to know that 52 cards are dealt out equally to four players, who
are traditionally called *north*, *east*, *south* and *west*.  Our class looks
something like this::

    class Hand(object):
        """A hand of cards (bridge style)"""

        def __init__(self, north, east, south, west):
            # Input parameters are lists of cards ('Ah', '9s', etc)
            self.north = north
            self.east = east
            self.south = south
            self.west = west

        # ... (other possibly useful methods omitted) ...

.. _Bridge: http://en.wikipedia.org/wiki/Contract_bridge

This is just an ordinary Python class, with nothing Django-specific about it.
We'd like to be able to do things like this in our models (we assume the
``hand`` attribute on the model is an instance of ``Hand``)::

    example = MyModel.objects.get(pk=1)
    print example.hand.north

    new_hand = Hand(north, east, south, west)
    example.hand = new_hand
    example.save()

We assign to and retrieve from the ``hand`` attribute in our model just like
any other Python class. The trick is to tell Django how to handle saving and
loading such an object.

In order to use the ``Hand`` class in our models, we **do not** have to change
this class at all. This is ideal, because it means you can easily write
model support for existing classes where you cannot change the source code.

.. note::
    You might only be wanting to take advantage of custom database column
    types and deal with the data as standard Python types in your models;
    strings, or floats, for example. This case is similar to our ``Hand``
    example and we'll note any differences as we go along.

Background theory
=================

Database storage
----------------

The simplest way to think of a model field is that it provides a way to take a
normal Python object -- string, boolean, ``datetime``, or something more
complex like ``Hand`` -- and convert it to and from a format that is useful
when dealing with the database (and serialization, but, as we'll see later,
that falls out fairly naturally once you have the database side under control).

Fields in a model must somehow be converted to fit into an existing database
column type. Different databases provide different sets of valid column types,
but the rule is still the same: those are the only types you have to work
with. Anything you want to store in the database must fit into one of
those types.

Normally, you're either writing a Django field to match a particular database
column type, or there's a fairly straightforward way to convert your data to,
say, a string.

For our ``Hand`` example, we could convert the card data to a string of 104
characters by concatenating all the cards together in a pre-determined order --
say, all the *north* cards first, then the *east*, *south* and *west* cards. So
``Hand`` objects can be saved to text or character columns in the database.

What does a field class do?
---------------------------

.. class:: Field

All of Django's fields (and when we say *fields* in this document, we always
mean model fields and not :doc:`form fields </ref/forms/fields>`) are subclasses
of :class:`django.db.models.Field`. Most of the information that Django records
about a field is common to all fields -- name, help text, uniqueness and so
forth. Storing all that information is handled by ``Field``. We'll get into the
precise details of what ``Field`` can do later on; for now, suffice it to say
that everything descends from ``Field`` and then customizes key pieces of the
class behavior.

It's important to realize that a Django field class is not what is stored in
your model attributes. The model attributes contain normal Python objects. The
field classes you define in a model are actually stored in the ``Meta`` class
when the model class is created (the precise details of how this is done are
unimportant here). This is because the field classes aren't necessary when
you're just creating and modifying attributes. Instead, they provide the
machinery for converting between the attribute value and what is stored in the
database or sent to the :doc:`serializer </topics/serialization>`.

Keep this in mind when creating your own custom fields. The Django ``Field``
subclass you write provides the machinery for converting between your Python
instances and the database/serializer values in various ways (there are
differences between storing a value and using a value for lookups, for
example). If this sounds a bit tricky, don't worry -- it will become clearer in
the examples below. Just remember that you will often end up creating two
classes when you want a custom field:

    * The first class is the Python object that your users will manipulate.
      They will assign it to the model attribute, they will read from it for
      displaying purposes, things like that. This is the ``Hand`` class in our
      example.

    * The second class is the ``Field`` subclass. This is the class that knows
      how to convert your first class back and forth between its permanent
      storage form and the Python form.

Writing a field subclass
========================

When planning your :class:`~django.db.models.Field` subclass, first give some
thought to which existing :class:`~django.db.models.Field` class your new field
is most similar to. Can you subclass an existing Django field and save yourself
some work? If not, you should subclass the :class:`~django.db.models.Field`
class, from which everything is descended.

Initializing your new field is a matter of separating out any arguments that are
specific to your case from the common arguments and passing the latter to the
:meth:`~django.db.models.Field.__init__` method of
:class:`~django.db.models.Field` (or your parent class).

In our example, we'll call our field ``HandField``. (It's a good idea to call
your :class:`~django.db.models.Field` subclass ``<Something>Field``, so it's
easily identifiable as a :class:`~django.db.models.Field` subclass.) It doesn't
behave like any existing field, so we'll subclass directly from
:class:`~django.db.models.Field`::

    from django.db import models

    class HandField(models.Field):

        description = "A hand of cards (bridge style)"

        def __init__(self, *args, **kwargs):
            kwargs['max_length'] = 104
            super(HandField, self).__init__(*args, **kwargs)

Our ``HandField`` accepts most of the standard field options (see the list
below), but we ensure it has a fixed length, since it only needs to hold 52
card values plus their suits; 104 characters in total.

.. note::
    Many of Django's model fields accept options that they don't do anything
    with. For example, you can pass both
    :attr:`~django.db.models.Field.editable` and
    :attr:`~django.db.models.Field.auto_now` to a
    :class:`django.db.models.DateField` and it will simply ignore the
    :attr:`~django.db.models.Field.editable` parameter
    (:attr:`~django.db.models.Field.auto_now` being set implies
    ``editable=False``). No error is raised in this case.

    This behavior simplifies the field classes, because they don't need to
    check for options that aren't necessary. They just pass all the options to
    the parent class and then don't use them later on. It's up to you whether
    you want your fields to be more strict about the options they select, or
    to use the simpler, more permissive behavior of the current fields.

.. method:: Field.__init__

The :meth:`~django.db.models.Field.__init__` method takes the following
parameters:

    * :attr:`~django.db.models.Field.verbose_name`
    * :attr:`~django.db.models.Field.name`
    * :attr:`~django.db.models.Field.primary_key`
    * :attr:`~django.db.models.Field.max_length`
    * :attr:`~django.db.models.Field.unique`
    * :attr:`~django.db.models.Field.blank`
    * :attr:`~django.db.models.Field.null`
    * :attr:`~django.db.models.Field.db_index`
    * :attr:`~django.db.models.Field.rel`: Used for related fields (like
      :class:`ForeignKey`). For advanced use only.
    * :attr:`~django.db.models.Field.default`
    * :attr:`~django.db.models.Field.editable`
    * :attr:`~django.db.models.Field.serialize`: If ``False``, the field will
      not be serialized when the model is passed to Django's :doc:`serializers
      </topics/serialization>`. Defaults to ``True``.
    * :attr:`~django.db.models.Field.unique_for_date`
    * :attr:`~django.db.models.Field.unique_for_month`
    * :attr:`~django.db.models.Field.unique_for_year`
    * :attr:`~django.db.models.Field.choices`
    * :attr:`~django.db.models.Field.help_text`
    * :attr:`~django.db.models.Field.db_column`
    * :attr:`~django.db.models.Field.db_tablespace`: Currently only used with
      the Oracle backend and only for index creation. You can usually ignore
      this option.
    * :attr:`~django.db.models.Field.auto_created`: True if the field was
      automatically created, as for the `OneToOneField` used by model
      inheritance. For advanced use only.

All of the options without an explanation in the above list have the same
meaning they do for normal Django fields. See the :doc:`field documentation
</ref/models/fields>` for examples and details.

The ``SubfieldBase`` metaclass
------------------------------

.. class:: django.db.models.SubfieldBase

As we indicated in the introduction_, field subclasses are often needed for
two reasons: either to take advantage of a custom database column type, or to
handle complex Python types. Obviously, a combination of the two is also
possible. If you're only working with custom database column types and your
model fields appear in Python as standard Python types direct from the
database backend, you don't need to worry about this section.

If you're handling custom Python types, such as our ``Hand`` class, we need to
make sure that when Django initializes an instance of our model and assigns a
database value to our custom field attribute, we convert that value into the
appropriate Python object. The details of how this happens internally are a
little complex, but the code you need to write in your ``Field`` class is
simple: make sure your field subclass uses a special metaclass:

For example::

    class HandField(models.Field):

        description = "A hand of cards (bridge style)"

        __metaclass__ = models.SubfieldBase

        def __init__(self, *args, **kwargs):
            # ...

This ensures that the :meth:`.to_python` method, documented below, will always
be called when the attribute is initialized.

ModelForms and custom fields
~~~~~~~~~~~~~~~~~~~~~~~~~~~~

If you use :class:`~django.db.models.SubfieldBase`, :meth:`.to_python`
will be called every time an instance of the field is assigned a
value. This means that whenever a value may be assigned to the field,
you need to ensure that it will be of the correct datatype, or that
you handle any exceptions.

This is especially important if you use :doc:`ModelForms
</topics/forms/modelforms>`. When saving a ModelForm, Django will use
form values to instantiate model instances. However, if the cleaned
form data can't be used as valid input to the field, the normal form
validation process will break.

Therefore, you must ensure that the form field used to represent your
custom field performs whatever input validation and data cleaning is
necessary to convert user-provided form input into a
`to_python()`-compatible model field value. This may require writing a
custom form field, and/or implementing the :meth:`.formfield` method on
your field to return a form field class whose `to_python()` returns the
correct datatype.

Documenting your custom field
-----------------------------

.. attribute:: Field.description

As always, you should document your field type, so users will know what it is.
In addition to providing a docstring for it, which is useful for developers,
you can also allow users of the admin app to see a short description of the
field type via the :doc:`django.contrib.admindocs
</ref/contrib/admin/admindocs>` application. To do this simply provide
descriptive text in a ``description`` class attribute of your custom field. In
the above example, the description displayed by the ``admindocs``
application for a ``HandField`` will be 'A hand of cards (bridge style)'.

Useful methods
--------------

Once you've created your :class:`~django.db.models.Field` subclass and set up
the ``__metaclass__``, you might consider overriding a few standard methods,
depending on your field's behavior. The list of methods below is in
approximately decreasing order of importance, so start from the top.

Custom database types
~~~~~~~~~~~~~~~~~~~~~

.. method:: Field.db_type(self, connection)

.. versionadded:: 1.2
   The ``connection`` argument was added to support multiple databases.

Returns the database column data type for the :class:`~django.db.models.Field`,
taking into account the connection object, and the settings associated with it.

Say you've created a PostgreSQL custom type called ``mytype``. You can use this
field with Django by subclassing ``Field`` and implementing the
:meth:`.db_type` method, like so::

    from django.db import models

    class MytypeField(models.Field):
        def db_type(self, connection):
            return 'mytype'

Once you have ``MytypeField``, you can use it in any model, just like any other
``Field`` type::

    class Person(models.Model):
        name = models.CharField(max_length=80)
        gender = models.CharField(max_length=1)
        something_else = MytypeField()

If you aim to build a database-agnostic application, you should account for
differences in database column types. For example, the date/time column type
in PostgreSQL is called ``timestamp``, while the same column in MySQL is called
``datetime``. The simplest way to handle this in a :meth:`.db_type`
method is to check the ``connection.settings_dict['ENGINE']`` attribute.

For example::

    class MyDateField(models.Field):
        def db_type(self, connection):
            if connection.settings_dict['ENGINE'] == 'django.db.backends.mysql':
                return 'datetime'
            else:
                return 'timestamp'

The :meth:`.db_type` method is only called by Django when the framework
constructs the ``CREATE TABLE`` statements for your application -- that is,
when you first create your tables. It's not called at any other time, so it can
afford to execute slightly complex code, such as the
``connection.settings_dict`` check in the above example.

Some database column types accept parameters, such as ``CHAR(25)``, where the
parameter ``25`` represents the maximum column length. In cases like these,
it's more flexible if the parameter is specified in the model rather than being
hard-coded in the ``db_type()`` method. For example, it wouldn't make much
sense to have a ``CharMaxlength25Field``, shown here::

    # This is a silly example of hard-coded parameters.
    class CharMaxlength25Field(models.Field):
        def db_type(self, connection):
            return 'char(25)'

    # In the model:
    class MyModel(models.Model):
        # ...
        my_field = CharMaxlength25Field()

The better way of doing this would be to make the parameter specifiable at run
time -- i.e., when the class is instantiated. To do that, just implement
:meth:`django.db.models.Field.__init__`, like so::

    # This is a much more flexible example.
    class BetterCharField(models.Field):
        def __init__(self, max_length, *args, **kwargs):
            self.max_length = max_length
            super(BetterCharField, self).__init__(*args, **kwargs)

        def db_type(self, connection):
            return 'char(%s)' % self.max_length

    # In the model:
    class MyModel(models.Model):
        # ...
        my_field = BetterCharField(25)

Finally, if your column requires truly complex SQL setup, return ``None`` from
:meth:`.db_type`. This will cause Django's SQL creation code to skip
over this field. You are then responsible for creating the column in the right
table in some other way, of course, but this gives you a way to tell Django to
get out of the way.

Converting database values to Python objects
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

.. method:: Field.to_python(self, value)

Converts a value as returned by your database (or a serializer) to a Python
object.

The default implementation simply returns ``value``, for the common case in
which the database backend already returns data in the correct format (as a
Python string, for example).

If your custom :class:`~django.db.models.Field` class deals with data structures
that are more complex than strings, dates, integers or floats, then you'll need
to override this method. As a general rule, the method should deal gracefully
with any of the following arguments:

    * An instance of the correct type (e.g., ``Hand`` in our ongoing example).

    * A string (e.g., from a deserializer).

    * Whatever the database returns for the column type you're using.

In our ``HandField`` class, we're storing the data as a VARCHAR field in the
database, so we need to be able to process strings and ``Hand`` instances in
:meth:`.to_python`::

    import re

    class HandField(models.Field):
        # ...

        def to_python(self, value):
            if isinstance(value, Hand):
                return value

            # The string case.
            p1 = re.compile('.{26}')
            p2 = re.compile('..')
            args = [p2.findall(x) for x in p1.findall(value)]
            return Hand(*args)

Notice that we always return a ``Hand`` instance from this method. That's the
Python object type we want to store in the model's attribute.

**Remember:** If your custom field needs the :meth:`to_python` method to be
called when it is created, you should be using `The SubfieldBase metaclass`_
mentioned earlier. Otherwise :meth:`.to_python` won't be called
automatically.

Converting Python objects to query values
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

.. method:: Field.get_prep_value(self, value)

.. versionadded:: 1.2
   This method was factored out of ``get_db_prep_value()``

This is the reverse of :meth:`.to_python` when working with the
database backends (as opposed to serialization). The ``value``
parameter is the current value of the model's attribute (a field has
no reference to its containing model, so it cannot retrieve the value
itself), and the method should return data in a format that has been
prepared for use as a parameter in a query.

This conversion should *not* include any database-specific
conversions. If database-specific conversions are required, they
should be made in the call to :meth:`.get_db_prep_value`.

For example::

    class HandField(models.Field):
        # ...

        def get_prep_value(self, value):
            return ''.join([''.join(l) for l in (value.north,
                    value.east, value.south, value.west)])

Converting query values to database values
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

.. method:: Field.get_db_prep_value(self, value, connection, prepared=False)

.. versionadded:: 1.2
   The ``connection`` and ``prepared`` arguments were added to support multiple databases.

Some data types (for example, dates) need to be in a specific format
before they can be used by a database backend.
:meth:`.get_db_prep_value` is the method where those conversions should
be made. The specific connection that will be used for the query is
passed as the ``connection`` parameter. This allows you to use
backend-specific conversion logic if it is required.

The ``prepared`` argument describes whether or not the value has
already been passed through :meth:`.get_prep_value` conversions. When
``prepared`` is False, the default implementation of
:meth:`.get_db_prep_value` will call :meth:`.get_prep_value` to do
initial data conversions before performing any database-specific
processing.

.. method:: Field.get_db_prep_save(self, value, connection)

.. versionadded:: 1.2
   The ``connection`` argument was added to support multiple databases.

Same as the above, but called when the Field value must be *saved* to
the database. As the default implementation just calls
:meth:`.get_db_prep_value`, you shouldn't need to implement this method
unless your custom field needs a special conversion when being saved
that is not the same as the conversion used for normal query
parameters (which is implemented by :meth:`.get_db_prep_value`).

Preprocessing values before saving
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

.. method:: Field.pre_save(self, model_instance, add)

This method is called just prior to :meth:`.get_db_prep_save` and should return
the value of the appropriate attribute from ``model_instance`` for this field.
The attribute name is in ``self.attname`` (this is set up by
:class:`~django.db.models.Field`). If the model is being saved to the database
for the first time, the ``add`` parameter will be ``True``, otherwise it will be
``False``.

You only need to override this method if you want to preprocess the value
somehow, just before saving. For example, Django's
:class:`~django.db.models.DateTimeField` uses this method to set the attribute
correctly in the case of :attr:`~django.db.models.Field.auto_now` or
:attr:`~django.db.models.Field.auto_now_add`.

If you do override this method, you must return the value of the attribute at
the end. You should also update the model's attribute if you make any changes
to the value so that code holding references to the model will always see the
correct value.

Preparing values for use in database lookups
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

As with value conversions, preparing a value for database lookups is a
two phase process.

.. method:: Field.get_prep_lookup(self, lookup_type, value)

.. versionadded:: 1.2
   This method was factored out of ``get_db_prep_lookup()``

:meth:`.get_prep_lookup` performs the first phase of lookup preparation,
performing generic data validity checks

Prepares the ``value`` for passing to the database when used in a lookup (a
``WHERE`` constraint in SQL). The ``lookup_type`` will be one of the valid
Django filter lookups: ``exact``, ``iexact``, ``contains``, ``icontains``,
``gt``, ``gte``, ``lt``, ``lte``, ``in``, ``startswith``, ``istartswith``,
``endswith``, ``iendswith``, ``range``, ``year``, ``month``, ``day``,
``isnull``, ``search``, ``regex``, and ``iregex``.

Your method must be prepared to handle all of these ``lookup_type`` values and
should raise either a ``ValueError`` if the ``value`` is of the wrong sort (a
list when you were expecting an object, for example) or a ``TypeError`` if
your field does not support that type of lookup. For many fields, you can get
by with handling the lookup types that need special handling for your field
and pass the rest to the :meth:`.get_db_prep_lookup` method of the parent class.

If you needed to implement ``get_db_prep_save()``, you will usually need to
implement ``get_prep_lookup()``. If you don't, ``get_prep_value`` will be
called by the default implementation, to manage ``exact``, ``gt``, ``gte``,
``lt``, ``lte``, ``in`` and ``range`` lookups.

You may also want to implement this method to limit the lookup types that could
be used with your custom field type.

Note that, for ``range`` and ``in`` lookups, ``get_prep_lookup`` will receive
a list of objects (presumably of the right type) and will need to convert them
to a list of things of the right type for passing to the database. Most of the
time, you can reuse ``get_prep_value()``, or at least factor out some common
pieces.

For example, the following code implements ``get_prep_lookup`` to limit the
accepted lookup types to ``exact`` and ``in``::

    class HandField(models.Field):
        # ...

        def get_prep_lookup(self, lookup_type, value):
            # We only handle 'exact' and 'in'. All others are errors.
            if lookup_type == 'exact':
                return self.get_prep_value(value)
            elif lookup_type == 'in':
                return [self.get_prep_value(v) for v in value]
            else:
                raise TypeError('Lookup type %r not supported.' % lookup_type)

.. method:: Field.get_db_prep_lookup(self, lookup_type, value, connection, prepared=False)

.. versionadded:: 1.2
   The ``connection`` and ``prepared`` arguments were added to support multiple databases.

Performs any database-specific data conversions required by a lookup.
As with :meth:`.get_db_prep_value`, the specific connection that will
be used for the query is passed as the ``connection`` parameter.
The ``prepared`` argument describes whether the value has already been
prepared with :meth:`.get_prep_lookup`.

Specifying the form field for a model field
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

.. method:: Field.formfield(self, form_class=forms.CharField, **kwargs)

Returns the default form field to use when this field is displayed in a model.
This method is called by the :class:`~django.forms.ModelForm` helper.

All of the ``kwargs`` dictionary is passed directly to the form field's
:meth:`~django.forms.Field__init__` method. Normally, all you need to do is
set up a good default for the ``form_class`` argument and then delegate further
handling to the parent class. This might require you to write a custom form
field (and even a form widget). See the :doc:`forms documentation
</topics/forms/index>` for information about this, and take a look at the code in
:mod:`django.contrib.localflavor` for some examples of custom widgets.

Continuing our ongoing example, we can write the :meth:`.formfield` method as::

    class HandField(models.Field):
        # ...

        def formfield(self, **kwargs):
            # This is a fairly standard way to set up some defaults
            # while letting the caller override them.
            defaults = {'form_class': MyFormField}
            defaults.update(kwargs)
            return super(HandField, self).formfield(**defaults)

This assumes we've imported a ``MyFormField`` field class (which has its own
default widget). This document doesn't cover the details of writing custom form
fields.

.. _helper functions: ../forms/#generating-forms-for-models
.. _forms documentation: ../forms/

Emulating built-in field types
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

.. method:: Field.get_internal_type(self)

Returns a string giving the name of the :class:`~django.db.models.Field`
subclass we are emulating at the database level. This is used to determine the
type of database column for simple cases.

If you have created a :meth:`.db_type` method, you don't need to worry about
:meth:`.get_internal_type` -- it won't be used much. Sometimes, though, your
database storage is similar in type to some other field, so you can use that
other field's logic to create the right column.

For example::

    class HandField(models.Field):
        # ...

        def get_internal_type(self):
            return 'CharField'

No matter which database backend we are using, this will mean that ``syncdb``
and other SQL commands create the right column type for storing a string.

If :meth:`.get_internal_type` returns a string that is not known to Django for
the database backend you are using -- that is, it doesn't appear in
``django.db.backends.<db_name>.creation.DATA_TYPES`` -- the string will still be
used by the serializer, but the default :meth:`.db_type` method will return
``None``. See the documentation of :meth:`.db_type` for reasons why this might be
useful. Putting a descriptive string in as the type of the field for the
serializer is a useful idea if you're ever going to be using the serializer
output in some other place, outside of Django.

Converting field data for serialization
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

.. method:: Field.value_to_string(self, obj)

This method is used by the serializers to convert the field into a string for
output. Calling :meth:`Field._get_val_from_obj(obj)` is the best way to get the
value to serialize. For example, since our ``HandField`` uses strings for its
data storage anyway, we can reuse some existing conversion code::

    class HandField(models.Field):
        # ...

        def value_to_string(self, obj):
            value = self._get_val_from_obj(obj)
            return self.get_db_prep_value(value)

Some general advice
--------------------

Writing a custom field can be a tricky process, particularly if you're doing
complex conversions between your Python types and your database and
serialization formats. Here are a couple of tips to make things go more
smoothly:

    1. Look at the existing Django fields (in
       :file:`django/db/models/fields/__init__.py`) for inspiration. Try to find
       a field that's similar to what you want and extend it a little bit,
       instead of creating an entirely new field from scratch.

    2. Put a :meth:`__str__` or :meth:`__unicode__` method on the class you're
       wrapping up as a field. There are a lot of places where the default
       behavior of the field code is to call
       :func:`~django.utils.encoding.force_unicode` on the value. (In our
       examples in this document, ``value`` would be a ``Hand`` instance, not a
       ``HandField``). So if your :meth:`__unicode__` method automatically
       converts to the string form of your Python object, you can save yourself
       a lot of work.


Writing a ``FileField`` subclass
=================================

In addition to the above methods, fields that deal with files have a few other
special requirements which must be taken into account. The majority of the
mechanics provided by ``FileField``, such as controlling database storage and
retrieval, can remain unchanged, leaving subclasses to deal with the challenge
of supporting a particular type of file.

Django provides a ``File`` class, which is used as a proxy to the file's
contents and operations. This can be subclassed to customize how the file is
accessed, and what methods are available. It lives at
``django.db.models.fields.files``, and its default behavior is explained in the
:doc:`file documentation </ref/files/file>`.

Once a subclass of ``File`` is created, the new ``FileField`` subclass must be
told to use it. To do so, simply assign the new ``File`` subclass to the special
``attr_class`` attribute of the ``FileField`` subclass.

A few suggestions
------------------

In addition to the above details, there are a few guidelines which can greatly
improve the efficiency and readability of the field's code.

    1. The source for Django's own ``ImageField`` (in
       ``django/db/models/fields/files.py``) is a great example of how to
       subclass ``FileField`` to support a particular type of file, as it
       incorporates all of the techniques described above.

    2. Cache file attributes wherever possible. Since files may be stored in
       remote storage systems, retrieving them may cost extra time, or even
       money, that isn't always necessary. Once a file is retrieved to obtain
       some data about its content, cache as much of that data as possible to
       reduce the number of times the file must be retrieved on subsequent
       calls for that information.