Using Django REST Framework, I want to limit which values can be used in a related field in a creation.
For example consider this example (based on the filtering example on https://web.archive.org/web/20140515203013/http://www.django-rest-framework.org/api-guide/filtering.html, but changed to ListCreateAPIView):
class PurchaseList(generics.ListCreateAPIView)
model = Purchase
serializer_class = PurchaseSerializer
def get_queryset(self):
user = self.request.user
return Purchase.objects.filter(purchaser=user)
In this example, how do I ensure that on creation the purchaser may only be equal to self.request.user, and that this is the only value populated in the dropdown in the form in the browsable API renderer?
I ended up doing something similar to what Khamaileon suggested here. Basically I modified my serializer to peek into the request, which kind of smells wrong, but it gets the job done... Here's how it looks (examplified with the purchase-example):
class PurchaseSerializer(serializers.HyperlinkedModelSerializer):
def get_fields(self, *args, **kwargs):
fields = super(PurchaseSerializer, self).get_fields(*args, **kwargs)
fields['purchaser'].queryset = permitted_objects(self.context['view'].request.user, fields['purchaser'].queryset)
return fields
class Meta:
model = Purchase
permitted_objects is a function which takes a user and a query, and returns a filtered query which only contains objects that the user has permission to link to. This seems to work both for validation and for the browsable API dropdown fields.
Here's how I do it:
class PurchaseList(viewsets.ModelViewSet):
...
def get_serializer(self, *args, **kwargs):
serializer_class = self.get_serializer_class()
context = self.get_serializer_context()
return serializer_class(*args, request_user=self.request.user, context=context, **kwargs)
class PurchaseSerializer(serializers.ModelSerializer):
...
def __init__(self, *args, request_user=None, **kwargs):
super(PurchaseSerializer, self).__init__(*args, **kwargs)
self.fields['user'].queryset = User._default_manager.filter(pk=request_user.pk)
The example link does not seem to be available anymore, but by reading other comments, I assume that you are trying to filter the user relationship to purchases.
If i am correct, then i can say that there is now an official way to do this. Tested with django rest framework 3.10.1.
class UserPKField(serializers.PrimaryKeyRelatedField):
def get_queryset(self):
user = self.context['request'].user
queryset = User.objects.filter(...)
return queryset
class PurchaseSeriaizer(serializers.ModelSerializer):
users = UserPKField(many=True)
class Meta:
model = Purchase
fields = ('id', 'users')
This works as well with the browsable API.
Sources:
https://github.com/encode/django-rest-framework/issues/1985#issuecomment-328366412
https://medium.com/django-rest-framework/limit-related-data-choices-with-django-rest-framework-c54e96f5815e
I disliked the style of having to override the init method for every place where I need to have access to user data or the instance at runtime to limit the queryset. So I opted for this solution.
Here is the code inline.
from rest_framework import serializers
class LimitQuerySetSerializerFieldMixin:
"""
Serializer mixin with a special `get_queryset()` method that lets you pass
a callable for the queryset kwarg. This enables you to limit the queryset
based on data or context available on the serializer at runtime.
"""
def get_queryset(self):
"""
Return the queryset for a related field. If the queryset is a callable,
it will be called with one argument which is the field instance, and
should return a queryset or model manager.
"""
# noinspection PyUnresolvedReferences
queryset = self.queryset
if hasattr(queryset, '__call__'):
queryset = queryset(self)
if isinstance(queryset, (QuerySet, Manager)):
# Ensure queryset is re-evaluated whenever used.
# Note that actually a `Manager` class may also be used as the
# queryset argument. This occurs on ModelSerializer fields,
# as it allows us to generate a more expressive 'repr' output
# for the field.
# Eg: 'MyRelationship(queryset=ExampleModel.objects.all())'
queryset = queryset.all()
return queryset
class DynamicQuersetPrimaryKeyRelatedField(LimitQuerySetSerializerFieldMixin, serializers.PrimaryKeyRelatedField):
"""Evaluates callable queryset at runtime."""
pass
class MyModelSerializer(serializers.ModelSerializer):
"""
MyModel serializer with a primary key related field to 'MyRelatedModel'.
"""
def get_my_limited_queryset(self):
root = self.root
if root.instance is None:
return MyRelatedModel.objects.none()
return root.instance.related_set.all()
my_related_model = DynamicQuersetPrimaryKeyRelatedField(queryset=get_my_limited_queryset)
class Meta:
model = MyModel
The only drawback with this is that you would need to explicitly set the related serializer field instead of using the automatic field discovery provided by ModelSerializer. i would however expect something like this to be in rest_framework by default.
In django rest framework 3.0 the get_fields method was removed. But in a similar way you can do this in the init function of the serializer:
class PurchaseSerializer(serializers.HyperlinkedModelSerializer):
class Meta:
model = Purchase
def __init__(self, *args, **kwargs):
super(PurchaseSerializer, self).__init__(*args, **kwargs)
if 'request' in self.context:
self.fields['purchaser'].queryset = permitted_objects(self.context['view'].request.user, fields['purchaser'].queryset)
I added the if check since if you use PurchaseSerializer as field in another serializer on get methods, the request will not be passed to the context.
First to make sure you only allow "self.request.user" when you have an incoming http POST/PUT (this assumes the property on your serializer and model is named "user" literally)
def validate_user(self, attrs, source):
posted_user = attrs.get(source, None)
if posted_user:
raise serializers.ValidationError("invalid post data")
else:
user = self.context['request']._request.user
if not user:
raise serializers.ValidationError("invalid post data")
attrs[source] = user
return attrs
By adding the above to your model serializer you ensure that ONLY the request.user is inserted into your database.
2) -about your filter above (filter purchaser=user) I would actually recommend using a custom global filter (to ensure this is filtered globally). I do something for a software as a service app of my own and it helps to ensure each http request is filtered down (including an http 404 when someone tries to lookup a "object" they don't have access to see in the first place)
I recently patched this in the master branch so both list and singular views will filter this
https://github.com/tomchristie/django-rest-framework/commit/1a8f07def8094a1e34a656d83fc7bdba0efff184
3) - about the api renderer - are you having your customers use this directly? if not I would say avoid it. If you need this it might be possible to add a custom serlializer that would help to limit the input on the front-end
Upon request # gabn88, as you may know by now, with DRF 3.0 and above, there is no easy solution.
Even IF you do manage to figure out a solution, it won't be pretty and will most likely fail on subsequent versions of DRF as it will override a bunch of DRF source which will have changed by then.
I forget the exact implementation I used, but the idea is to create 2 fields on the serializer, one your normal serializer field (lets say PrimaryKeyRelatedField etc...), and another field a serializer method field, which the results will be swapped under certain cases (such as based on the request, the request user, or whatever). This would be done on the serializers constructor (ie: init)
Your serializer method field will return a custom query that you want.
You will pop and/or swap these fields results, so that the results of your serializer method field will be assigned to the normal/default serializer field (PrimaryKeyRelatedField etc...) accordingly. That way you always deal with that one key (your default field) while the other key remains transparent within your application.
Along with this info, all you really need is to modify this: http://www.django-rest-framework.org/api-guide/serializers/#dynamically-modifying-fields
I wrote a custom CustomQueryHyperlinkedRelatedField class to generalize this behavior:
class CustomQueryHyperlinkedRelatedField(serializers.HyperlinkedRelatedField):
def __init__(self, view_name=None, **kwargs):
self.custom_query = kwargs.pop('custom_query', None)
super(CustomQueryHyperlinkedRelatedField, self).__init__(view_name, **kwargs)
def get_queryset(self):
if self.custom_query and callable(self.custom_query):
qry = self.custom_query()(self)
else:
qry = super(CustomQueryHyperlinkedRelatedField, self).get_queryset()
return qry
#property
def choices(self):
qry = self.get_queryset()
return OrderedDict([
(
six.text_type(self.to_representation(item)),
six.text_type(item)
)
for item in qry
])
Usage:
class MySerializer(serializers.HyperlinkedModelSerializer):
....
somefield = CustomQueryHyperlinkedRelatedField(view_name='someview-detail',
queryset=SomeModel.objects.none(),
custom_query=lambda: MySerializer.some_custom_query)
#staticmethod
def some_custom_query(field):
return SomeModel.objects.filter(somefield=field.context['request'].user.email)
...
I did the following:
class MyModelSerializer(serializers.ModelSerializer):
myForeignKeyFieldName = MyForeignModel.objects.all()
def get_fields(self, *args, **kwargs):
fields = super(MyModelSerializer, self).get_fields()
qs = MyModel.objects.filter(room=self.instance.id)
fields['myForeignKeyFieldName'].queryset = qs
return fields
I looked for a solution where I can set the queryset upon creation of the field and don't have to add a separate field class. This is what I came up with:
class PurchaseSerializer(serializers.HyperlinkedModelSerializer):
class Meta:
model = Purchase
fields = ["purchaser"]
def get_purchaser_queryset(self):
user = self.context["request"].user
return Purchase.objects.filter(purchaser=user)
def get_extra_kwargs(self):
kwargs = super().get_extra_kwargs()
kwargs["purchaser"] = {"queryset": self.get_purchaser_queryset()}
return kwargs
The main issue for tracking suggestions regarding this seems to be drf#1985.
Here's a re-usable generic serializer field that can be used instead of defining a custom field for every use case.
class DynamicPrimaryKeyRelatedField(serializers.PrimaryKeyRelatedField):
"""A PrimaryKeyRelatedField with ability to set queryset at runtime.
Pass a function in the `queryset_fn` kwarg. It will be passed the serializer `context`.
The function should return a queryset.
"""
def __init__(self, queryset_fn=None, **kwargs):
assert queryset_fn is not None, "The `queryset_fn` argument is required."
self.queryset_fn = queryset_fn
super().__init__(**kwargs)
def get_queryset(self):
return self.queryset_fn(context=self.context)
Usage:
class MySerializer(serializers.ModelSerializer):
my_models = DynamicPrimaryKeyRelatedField(
queryset_fn=lambda context: MyModel.objects.visible_to_user(context["request"].user)
)
# ...
Same works for serializers.SlugRelatedField.
Related
I have the below serializer:
class OrderItemResponseSerializer(serializers.ModelSerializer):
prepack_qty = serializers.SerializerMethodField()
product_description = serializers.SerializerMethodField()
class Meta:
model = OrderItem
fields = (
"product",
"qty_ordered",
"qty_approved",
"status",
"prepack_qty",
"product_description"
)
def get_prepack_qty(self, obj):
return obj.product.prepack_quantity
def get_product_description(self, obj):
return obj.product.product_description
When I make a get request to /orders, I make a lot of SQL queries to the database because different orders may contain the same product. How can I cache the result of get_prepack_qty and get_product_description methods? I tried to use #cached_property this way:
class OrderItem(models.Model):
...
#cached_property
def item_product_description(self):
return self.product.product_description
But the number of requests to the database remained the same.
Well, first of all, I should say that what you have implemented in this piece of code below:
...
#cached_property
def item_product_description(self):
return self.product.product_description
And using #cached_property alone doesn't cache the data for you, you just created a property in the model for get_product_description serializer method, And this does not reduce the volume and number of your queries to the database at all; Of course you need .bind() method in your serializer method like below:
class BlobSerializer(SerializerCacheMixin, serializers.Serializer):
blobs = serializers.SerializerMethodField()
def get_blobs(self, instance):
# recursive serializer
serializer = self.__class__(instance.results, many=True)
serializer.bind('*', self)
return serializer.data
But in order to cache the result of this method As you asked in your question there is a good project in Pypi called drf-serializer-cache you can use it easily for this purpose, for example, the following piece of code is taken from the document of this project:
from drf_serializer_cache import SerializerCacheMixin
from rest_framework import serializer
class ResultSerializer(SerializerCacheMixin, serializers.Serializer):
results = serializers.SerializerMethodField()
def get_results(self, instance):
# recursive serializer
serializer = self.__class__(instance.results, many=True)
serializer.bind('*', self) # bind call is essential for >efficient cache!
return serializer.data
Also if you want to implement it yourself in your project Seeing the implementation of SerializerCacheMixin object in this project can help you a lot or even use it directly.
You can leverage Django's cache framework to cache the result of a SerializerMethodField. It would look something like this:
from django.core.cache import cache
class MySerializer(serializers.Serializer):
my_field = serializers.SerializerMethodField()
def get_my_field(self, obj):
# Get cached value
value = cache.get("my_field_%s" % obj.pk)
if value is None:
# Value is not cached, compute it
value = ...
# Cache the value
cache.set("my_field_%s" % obj.pk, value, 3600)
return value
This will cache the result of the field for 1 hour (3600 seconds), using a cache key that depends on the object's primary key.
You can of course adapt the caching logic to your exact use case.
I have a situation where I wish to utilize Django's autocomplete admin widget, that respects a referencing models field limitation.
For example I have the following Collection model that has the attribute kind with specified choices.
class Collection(models.Model):
...
COLLECTION_KINDS = (
('personal', 'Personal'),
('collaborative', 'Collaborative'),
)
name = models.CharField()
kind = models.CharField(choices=COLLECTION_KINDS)
...
Another model ScheduledCollection references Collection with a ForeignKey field that implements limit_choices_to option. The purpose of this model is to associate meta data to a Collection for a specific use case.
class ScheduledCollection(models.Model):
...
collection = models.ForeignKey(Collection, limit_choices_to={'kind': 'collaborative'})
start_date = models.DateField()
end_date = models.DateField()
...
Both models are registered with a ModelAdmin. The Collection model implements search_fields.
#register(models.Collection)
class CollectionAdmin(ModelAdmin):
...
search_fields = ['name']
...
The ScheduledCollection model implements autocomplete_fields
#register(models.ScheduledCollection)
class ScheduledCollectionAdmin(ModelAdmin):
...
autocomplete_fields = ['collection']
...
This works but not entirely as expected. The autocomplete retrieves results from a view generated by the Collection model. The limit_choices_to do not filter the results and are only enforced upon save.
It has been suggested to implement get_search_results or get_queryset on the CollectionAdmin model. I was able to do this and filter the results. However, this changes Collection search results across the board. I am unaware of how to attain more context within get_search_results or get_queryset to conditionally filter the results based upon a relationship.
In my case I would like to have several choices for Collection and several meta models with different limit_choices_to options and have the autocomplete feature respect these restrictions.
I don't expect this to work automagically and maybe this should be a feature request. At this point I am at a loss how to filter the results of a autocomplete with the respect to a choice limitation (or any condition).
Without using autocomplete_fields the Django admin's default <select> widget filters the results.
Triggering off the http referer was ugly so I made a better version: subclass the AutocompleteSelect and send extra query parameters to allow get_search_results to lookup the correct limit_choices_to automagically. Simply include this mixin in your ModelAdmin (for both source and target models). As a bonus it also adds a delay to the ajax requests so you don't spam the server as you type in the filter, makes the select wider and sets the search_fields attribute (to 'translations__name' which is correct for my system, customise for yours or omit and set individually on the ModelAdmins as before):
from django.contrib.admin import widgets
from django.utils.http import urlencode
from django.contrib.admin.options import ModelAdmin
class AutocompleteSelect(widgets.AutocompleteSelect):
"""
Improved version of django's autocomplete select that sends an extra query parameter with the model and field name
it is editing, allowing the search function to apply the appropriate filter.
Also wider by default, and adds a debounce to the ajax requests
"""
def __init__(self, rel, admin_site, attrs=None, choices=(), using=None, for_field=None):
super().__init__(rel, admin_site, attrs=attrs, choices=choices, using=using)
self.for_field = for_field
def build_attrs(self, base_attrs, extra_attrs=None):
attrs = super().build_attrs(base_attrs, extra_attrs=extra_attrs)
attrs.update({
'data-ajax--delay': 250,
'style': 'width: 50em;'
})
return attrs
def get_url(self):
url = super().get_url()
url += '?' + urlencode({
'app_label': self.for_field.model._meta.app_label,
'model_name': self.for_field.model._meta.model_name,
'field_name': self.for_field.name
})
return url
class UseAutocompleteSelectMixin():
"""
To avoid ForeignKey fields to Event (such as on ReportColumn) in admin from pre-loading all events
and thus being really slow, we turn them into autocomplete fields which load the events based on search text
via an ajax call that goes through this method.
Problem is this ignores the limit_choices_to of the original field as this ajax is a general 'search events'
without knowing the context of what field it is populating. Someone else has exact same problem:
https://stackoverflow.com/questions/55344987/how-to-filter-modeladmin-autocomplete-fields-results-with-the-context-of-limit-c
So fix this by adding extra query parameters on the autocomplete request,
and use these on the target ModelAdmin to lookup the correct limit_choices_to and filter with it.
"""
# Overrides django.contrib.admin.options.ModelAdmin#formfield_for_foreignkey
# Is identical except in case db_field.name is in autocomplete fields it constructs our improved AutocompleteSelect
# instead of django's and passes it extra for_field parameter
def formfield_for_foreignkey(self, db_field, request, **kwargs):
if db_field.name in self.get_autocomplete_fields(request):
db = kwargs.get('using')
kwargs['widget'] = AutocompleteSelect(db_field.remote_field, self.admin_site, using=db, for_field=db_field)
if 'queryset' not in kwargs:
queryset = self.get_field_queryset(db, db_field, request)
if queryset is not None:
kwargs['queryset'] = queryset
return db_field.formfield(**kwargs)
return super().formfield_for_foreignkey(db_field, request, **kwargs)
# In principle we could add this override in a different mixin as adding the formfield override above is needed on
# the source ModelAdmin, and this is needed on the target ModelAdmin, but there's do damage adding everywhere so combine them.
def get_search_results(self, request, queryset, search_term):
if 'app_label' in request.GET and 'model_name' in request.GET and 'field_name' in request.GET:
from django.apps import apps
model_class = apps.get_model(request.GET['app_label'], request.GET['model_name'])
limit_choices_to = model_class._meta.get_field(request.GET['field_name']).get_limit_choices_to()
if limit_choices_to:
queryset = queryset.filter(**limit_choices_to)
return super().get_search_results(request, queryset, search_term)
search_fields = ['translations__name']
I had the exact same problem. It's a bit hacky, but here's my solution:
Override get_search_results of the ModelAdmin you are searching for and want to filter
Use the request referer header to get the magical context you need to apply the appropriate filter based on the source of the relationship
Grab the limit_choices_to from the appropriate ForeignKey's _meta
Pre-filter the queryset and then pass to super method.
So for your models:
#register(models.Collection)
class CollectionAdmin(ModelAdmin):
...
search_fields = ['name']
def get_search_results(self, request, queryset, search_term):
if '<app_name>/scheduledcollection/' in request.META.get('HTTP_REFERER', ''):
limit_choices_to = ScheduledCollection._meta.get_field('collection').get_limit_choices_to()
queryset = queryset.filter(**limit_choices_to)
return super().get_search_results(request, queryset, search_term)
A shortcoming of this approach is the only context we have is the model being edited in admin, rather than which field of the model, so if your ScheduledCollection model has 2 collection autocomplete fields (say personal_collection and collaborative_collection) with different limit_choices_to we can't infer this from the referer header and treat them differently. Also inline admins will have the referer url based on the parent thing they are an inline for, rather than reflecting their own model. But it works in the basic cases.
Hopefully a new version of Django will have a cleaner solution, such as the autocomplete select widget sending an extra query parameter with the model and field name it is editing so that get_search_results can accurately look up the required filters instead of (potentially inaccurately) inferring from the referer header.
Here is another solution to get only a subset of choices in the auto-complete field. This solution does not change the default behavior for the main model (Collection), so you can still have other views using autocomplete with the full set in your app.
Here is how it works:
Proxy model for Collection with manager
Create a proxy model to represent a subset of Collection, e.g. CollaborativeCollection to represent collections that are of kind "collaborative". You will also need a manager to restrict the initial queryset of your proxy model to the intended subset.
class CollaborativeCollectionManager(models.Manager):
def get_queryset(self):
return (
super()
.get_queryset()
.filter(kind="collaborative")
)
class CollaborativeCollection(models.Model):
class Meta:
proxy = True
objects = CollaborativeCollectionManager()
Updating foreign key to use proxy model
Next update the foreign key in ScheduledCollection to use the proxy model instead. Note that you can remove the limit_choices_to feature if you don't need it for anything else.
class ScheduledCollection(models.Model):
...
collection = models.ForeignKey(CollaborativeCollection)
start_date = models.DateField()
end_date = models.DateField()
...
Define admin model for Proxy
Finally define the admin model for the proxy.
#admin.register(CollaborativeCollection)
class CollaborativeCollectionAdmin(admin.ModelAdmin):
search_fields = ["name"]
Note that instead of the manager, you could also define a custom get_search_results() in the admin model. However, I found that the manager approach appears to be more performant. And it also is conceptually more sounds, since with that all queries to CollaborativeCollection will only return collaborative collections.
My solution is to wrap get_url method on the widget.
Create a util method as shown below.
def wrap_get_url(original_get_url, extra_url_params: QueryDict) -> Callable:
def get_url_with_extra_url_params(*args, **kwargs) -> str:
url: str = original_get_url(*args, **kwargs)
scheme, netloc, url, params, query, fragment = tuple(urlparse(url))
query = QueryDict(query_string=query, mutable=True)
query.update(extra_url_params)
url_parts = (scheme, netloc, url, params, query.urlencode(), fragment)
return urlunparse(url_parts)
return get_url_with_extra_url_params
Create a custom form for your model admin.
class ExampleModelAdminForm(forms.ModelForm):
class Meta:
model = ExampleModel
exclude: List[str] = []
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
instance = getattr(self, "instance", None)
# Check for RelatedWidgetWrapper
if widget := getattr(self.fields["target_model"].widget, "widget", None):
query = QueryDict(mutable=True)
query["example_model_id"] = instance.pk
widget.get_url = wrap_get_url(
original_get_url=widget.get_url,
extra_url_params=query,
)
class ExampleModelAdmin(admin.ModelAdmin):
form = forms.ExampleModelAdminForm
autocomplete_fields = ("target_model",)
On the target model admin.
class TargetModelAdmin(admin.ModelAdmin):
search_fields = ("name", ) # Define your search fields
def get_search_results(self, request, queryset, search_term) -> tuple[QuerySet, bool]:
qs: QuerySet
duplicate: bool
qs, duplicate = super(TargetModelAdmin, self).get_search_results(request, queryset, search_term)
# Get Example model id from previous admin page in order to filter the queryset
if example_model_id := request.GET.get("example_account_id", None):
example_model: ExampleModel = ExampleModel.objects.get(
id=example_model_id
)
qs = qs.filter(field=example_model.field) # Filter your qs here
return qs, duplicate
With Django 3.2, the solution proposed by #Uberdude does not work anymore because AutocompleteSelect's constructor now takes a field rather than a relation.
Here is the updated code needed for the formfield_for_foreignkey method:
def formfield_for_foreignkey(self, db_field, request, **kwargs):
if db_field.name in self.get_autocomplete_fields(request) or\
db_field.name in self.get_autocomplete_cb_fields(request):
db = kwargs.get('using')
if db_field.name in self.get_autocomplete_cb_fields(request):
kwargs['widget'] = AutocompleteSelectCb(
db_field, self.admin_site, using=db, for_field=db_field)
else:
kwargs['widget'] = AutocompleteSelect(
db_field, self.admin_site, using=db, for_field=db_field)
if 'queryset' not in kwargs:
queryset = self.get_field_queryset(db, db_field, request)
if queryset is not None:
kwargs['queryset'] = queryset
return db_field.formfield(**kwargs)
return super().formfield_for_foreignkey(db_field, request, **kwargs)
I would like to provide different widgets to input form fields for the same type of model field in a Django admin inline.
I have implemented a version of the Entity-Attribute-Value paradigm in my shop application (I tried eav-django and it wasn't flexible enough). In my model it is Product-Parameter-Value (see Edit below).
Everything works as I want except that when including an admin inline for the Parameter-Value pair, the same input formfield is used for every value. I understand that this is the default Django admin behaviour because it uses the same formset for each Inline row.
I have a callback on my Parameter that I would like to use (get_value_formfield). I currently have:
class SpecificationValueAdminInline(admin.TabularInline):
model = SpecificationValue
fields = ('parameter', 'value')
readonly_fields = ('parameter',)
max_num = 0
def get_formset(self, request, instance, **kwargs):
"""Take a copy of the instance"""
self.parent_instance = instance
return super().get_formset(request, instance, **kwargs)
def formfield_for_dbfield(self, db_field, **kwargs):
"""Override admin function for requesting the formfield"""
if self.parent_instance and db_field.name == 'value':
# Notice first() on the end -->
sv_instance = SpecificationValue.objects.filter(
product=self.parent_instance).first()
formfield = sv_instance.parameter.get_value_formfield()
else:
formfield = super().formfield_for_dbfield(db_field, **kwargs)
return formfield
formfield_for_dbfield is only called once for each admin page.
How would I override the default behaviour so that formfield_for_dbfield is called once for each SpecificationValue instance, preferably passing the instance in each time?
Edit:
Here is the model layout:
class Product(Model):
specification = ManyToManyField('SpecificationParameter',
through='SpecificationValue')
class SpecificationParameter(Model):
"""Other normal model fields here"""
type = models.PositiveSmallIntegerField(choices=TUPLE)
def get_value_formfield(self):
"""
Return the type of form field for parameter instance
with the correct widget for the value
"""
class SpecificationValue(Model):
product = ForeignKey(Product)
parameter = ForeignKey(SpecificationParameter)
# To store and retrieve all types of value, overrides CharField
value = CustomValueField()
The way I eventually solved this is using the form = attribute of the Admin Inline. This skips the form generation code of the ModelAdmin:
class SpecificationValueForm(ModelForm):
class Meta:
model = SpecificationValue
def __init__(self, instance=None, **kwargs):
super().__init__(instance=instance, **kwargs)
if instance:
self.fields['value'] = instance.parameter.get_value_formfield()
else:
self.fields['value'].disabled = True
class SpecificationValueAdminInline(admin.TabularInline):
form = SpecificationValueForm
Using standard forms like this, widgets with choices (e.g. RadioSelect and CheckboxSelectMultiple) have list bullets next to them in the admin interface because the <ul> doesn't have the radiolist class. You can almost fix the RadioSelect by using AdminRadioSelect(attrs={'class': 'radiolist'}) but there isn't an admin version of the CheckboxSelectMultiple so I preferred consistency. Also there is an aligned class missing from the <fieldset> wrapper element.
Looks like I'll have to live with that!
Is it possible to build a custom model field/widget combination which displays a value but never writes anything back to the database? I would use this widget exclusively in the admin's forms.
I wrote my own field, which overwrites the formfield() method to declare its own widget class. It displays just fine, but as soon as the 'Save' button is clicked in the admin, I'm getting a validation error:
This field is required.
That makes sense, considering that my widget didn't render out a form field. However, what I'd like to do is basically remove this field from the update process: whenever used in the admin, it just shouldn't be mentioned in the SQL UPDATE at all.
Is that possible?
Here's a sketch of the code I have so far:
class MyWidget(Widget):
def render(self, name, value, attrs=None):
if value is None:
value = ""
else:
# pretty print the contents of value here
return '<table>' + ''.join(rows) + '</table>'
class MyField(JSONField):
def __init__(self, *args, **kwargs):
kwargs['null'] = False
kwargs['default'] = list
super(MyField, self).__init__(*args, **kwargs)
def formfield(self, **kwargs):
defaults = {
'form_class': JSONFormField,
'widget': MyWidget,
}
defaults.update(**kwargs)
return super(MyField, self).formfield(**defaults)
UPDATE 1: The use case is that the field represents an audit log. Internally, it will be written to regularly. The admin however never needs to write to it, it only has to render it out in a very readable format.
I'm not using any other ModelForms in the application, so the admin is the only form-user. I don't want to implement the behavior on the admin classes themselves, because this field will be reused across various models and is always supposed to behave the same way.
There are multiple ways to create a read-only field in the admin pages. Your requirements on the database storage are a bit fuzzy so I go through the options.
You have to register an AdminModel first in admin.py:
from django.contrib import admin
from yourapp.models import YourModel
class YourAdmin(admin.ModelAdmin):
pass
admin.site.register(YourModel, YourAdmin)
Now you can add different behavior to it. For example you can add the list of fields shown in the edit/add page:
class YourAdmin(admin.ModelAdmin):
fields = ['field1', 'field2']
This can be names of the model fields, model properties or model methods. Methods are displayed read-only.
If you want to have one field read-only explicitly add this:
class YourAdmin(admin.ModelAdmin):
fields = ['field1', 'field2']
readonly_fields = ['field2']
Then you have the option to overwrite the display of the field completely by adding a method with the same name. You will not even need a model field/method with that name, then:
class YourAdmin(admin.ModelAdmin):
fields = ['field1', 'field2']
readonly_fields = ['field2']
def field2(self, obj):
return '*** CLASSIFIED *** {}'.format(obj.field2)
With django.utils.safestring.mark_safe you can return HTML code as well.
All other options of the Admin are available, except the widget configuration as it applies to the writable fields only.
I might be a little confused as to what you want but you might want to look into model properties. Here is an example for my current project.
Code inside your model:
class Textbook(models.Model):
#other fields
#property
def NumWishes(self):
return self.wishlist_set.count()
Then you can just display it on the admin page.
class Textbook_table(admin.ModelAdmin):
fields = ["""attributes that are saved in the model"""]
list_display = ("""attributes that are saved in the model""", 'NumWishes'')
So now I can display NumWishes in the admin page but it doesn't need to be created with the model.
Hello in the class admin modify the permission method
#admin.register(my_model)
class My_modelAdmin(admin.ModelAdmin):
def has_delete_permission(self, request, obj=None):
return False
def has_change_permission(self, request, obj=None):
return False
I have a subscription model that looks like this
class Subscription(models.Model):
name = models.CharField(max_length=100)
quantity = models.IntegerField(max_length=20)
stripe_id = models.CharField(max_length=100)
user = models.ForeignKey(User)
I would like to create an endpoint that allows POST, PATCH, DELETE, GET
So I did the following things
views.py
class SubscriptionDetail(viewsets.ModelViewSet):
serializer_class = SubscriptionSerializer
permission_classes = (IsAuthenticated,)
queryset = Subscription.objects.all()
serializers.py
class SubscriptionSerializer(serializers.ModelSerializer):
class Meta:
model = Subscription
fields = ('name','quantity', 'stripe_id')
def update(self, instance, validated_data):
print "In update"
#how do I write create and delete?
urls.py
subscription = SubscriptionDetail.as_view({
'patch': 'update'
})
url(r'^rest-auth/subscription/$', subscription, name='something'),
Questions
Using the above when I send a PATCH request, I get an error. How can I fix this?
Expected view SubscriptionDetail to be called with a URL keyword
argument named "pk". Fix your URL conf, or set the .lookup_field
attribute on the view correctly.
While sending the patch request I would also like to send an 'email' field which is not on the subscription model. Is this possible to do? I need the email field in the POST (create) operation so that I know which user the subscription belongs to.
The easiest way is to do it this way.
keep the models class the same
views.py
from rest_framework import viewsets
#impost serializer and model class for subscription
class SubscriptionViewSet(viewsets.ModelViewSet):
serializer_class = SubscriptionSerializer
def get_queryset(self):
queryset = Subscription.objects.all()
#if you need to get subscription by name
name = self.request.QUERY_PARAMS.get('name', None)
if name is not None:
queryset = queryset.filter(name=name)
return queryset
serializers.py
class SubscriptionSerializer(serializers.ModelSerializer):
class Meta:
model = Subscription
fields = ('name','quantity', 'stripe_id')
# django will handle get, delete,patch, update for you ....
# for customization you can use def update or def create ... to do whatever you need
# def create(self, validated_data):
# you can handle the email here
# and something like subscription= Subscription (name=validated_data['name'],vendor=validated_data['quantity']...)
# subscription.save()
# it will save whatever you want
urls.py
#use the router to handle everything for you
from django.conf.urls import patterns, include, url
from rest_framework import routers
#import your classes
router = routers.DefaultRouter()
router.register(r'subscription', views.SubscriptionViewSet,base_name='subscription')
urlpatterns = patterns('',
url(r'^', include(router.urls)),
)
For the creation of an Object you must implement the create function as described in the official documentation, found here. For patching you could use the partial argument from within you view class:
SubscriptionSerializer(subscription, data={'something': u'another', partial=True)
For deletion of the a Subscription, that could be done when you get the delete call as so in your view class:
if request.METHOD == 'DELETE':
subscription = Subscription.objects.get(pk=pk)
subscription.delete()
See this tutorial for complete example
Further more I think that you should include the "id" field in the SubscriptionSerialiser Meta class, otherwise it will be difficult to do the updates/deletions. I hope this helped a little.
Cheers,
Tobbe
When you want to use a method that allow make these operations you have to use a #detail_route() where you can say as well which methods will you use, like in the docs is said:
#detail_route(methods=['post'])
def set_password(self, request, pk=None):
user = self.get_object()
serializer = PasswordSerializer(data=request.data)
...
So to be able to use them you should add the next decorator
#detail_route(methods=['post', 'patch'])
To add another parameters you can do it for the .save() parameter. You just have to indicate the name of this and them just override your .save() model to check if that email belongs or not to the user that is trying to do the subscription. Here I paste you what the Django Rest docs says:
" Passing additional attributes to .save()
...
You can do so by including additional keyword arguments when calling .save(). For example:
serializer.save(owner=request.user)
Here I leave you the link for more information:
http://www.django-rest-framework.org/api-guide/serializers/#passing-additional-attributes-to-save
Using the above when I send a PATCH request, I get an error. How can I fix this?
Expected view SubscriptionDetail to be called with a URL keyword
argument named "pk". Fix your URL conf, or set the .lookup_field
attribute on the view correctly.
The error is caused because unlike create request, patch/update require a pk to know which object to update. That is why you have to supply the pk value for it. So, your url for PUT, DELETE andPATCH must have at least named parameter like this -
subscription = SubscriptionDetail.as_view({
'patch': 'update'
})
url(r'^rest-auth/subscription/(?<pk>(\d+))$', subscription, name='something'),
an example url will be - rest-auth/subscription/10 where 10 is the pk or id of the object. Django Rest Framework will then load the object internally to be updated.
While sending the patch request I would also like to send an 'email' field which is not on the subscription model. Is this possible to do? I need the email field in the POST (create) operation so that I know which user the subscription belongs to.
To add custom parameters, first declare the property in serializer, it is better to keep it required=False, so that other request does not throw error -
class SubscriptionSerializer(serializers.ModelSerializer):
custom_field = serialiers.BooleanField(required=False)
class Meta:
model = Subscription
fields = ('name','quantity', 'stripe_id')
def update(self, instance, validated_data):
print "In update"
so far this is enough for the django rest framework to accept the field custom_field and you will find the value in update method. To get the value pop it from the attributes supplied by the framework like this -
def update(self, instance, validated_data):
custom_field = validated_data.pop('custom_field', None)
if custom_field is not None:
# do whatever you like with the field
return super().update(instance, validated_data)
# for python < 3.0 super(SubscriptionSerializer, self).update(instance, validated_data)
When you overrided (I don't know if that's the proper conjugation of overriding a method) the update method, you stopped the ability to PUT or PATCH and object. Your new method only prints out "In update" but doesn't save the instance. Look at the update method from the serializer.ModelSerializer object:
def update(self, instance, validated_data):
raise_errors_on_nested_writes('update', self, validated_data)
for attr, value in validated_data.items():
setattr(instance, attr, value)
instance.save()
return instance
Notice the last few lines where the instance is saved with the values and then returned. Remove your update method on the SubscriptionSerializer object. This let's your parent object's create, update, retrieve, and delete methods do their magic which supports PATCH and PUT updates. The next problem is that your urls.py is using the Django rather than the REST framework router. Change it to this:
from rest_framework.routers import DefaultRouter
router = DefaultRouter()
router.register(r'subscription', SubscriptionDetail)
That should solve the patch update problem.
I don't think you can add an email field in your patch method without the attribute on the subscription model. That's just a guess on my part, and I may be wrong. Does the email field map to anything on any object? Can you use a ForeignKey to map it?
I hope that works for you, good luck!
In view.py you just need set the class with:
class SubscriptionDetail(mixins.CreateModelMixin,
mixins.ListModelMixin,
mixins.RetrieveModelMixin,
mixins.UpdateModelMixin,
generics.GenericAPIView):
and add this to fix .lookup_field :
def update(self, request, *args, **kwargs):
log.error("OBJ update kwargs= %s , data = %s" % (kwargs, str(request.data)))
pk = request.data.get('id')
if (kwargs.get('pk') is not None):
kwargs['pk'] = request.data.get('id')
self.kwargs['pk'] = request.data.get('id')
return super().update(request, *args, **kwargs)
and add support to methods do you want :
def post(self, request, *args, **kwargs):
return self.create(request, *args, **kwargs)
def get(self, request, *args, **kwargs):
return self.list(request, *args, **kwargs)
# def get(self, request, *args, **kwargs):
# return self.retrieve(request, *args, **kwargs)
def put(self, request, *args, **kwargs):
return self.update(request, *args, **kwargs)
# def patch(self, request, *args, **kwargs):
# return self.partial_update(request, *args, **kwargs)
#
# def delete(self, request, *args, **kwargs):
# return self.destroy(request, *args, **kwargs)
only tweak that remains is get for list or get for retrieve on element but should be easy now add something if we have one pk we may call self.retrieve else we may call self.list