This package has been deprecated in favor of django-pictures.
First, make sure you understand the differences between the two packages and how to serve images in a modern web application via the picture-Element.
Next, follow the setup instructions for django-pictures.
Once you are set up, change your models to use the new PictureField
and provide the
aspect_ratios
you'd like to serve. Do create migrations just yet.
This step should be followed by changing your templates and frontend. The new placeholders feature for local development should help you to do this almost effortlessly.
Finally, run makemigrations
and replace the AlterField
operation with
AlterPictureField
.
We highly recommend to use Django's image_width
and image_height
fields, to avoid
unnecessary IO. If you can add these fields to your model, you can use the following
snippet to populate them:
import django.core.files.storage
from django.db import migrations, models
import pictures.models
from pictures.migrations import AlterPictureField
def forward(apps, schema_editor):
for obj in apps.get_model("my-app.MyModel").objects.all().iterator():
obj.image_width = obj.logo.width
obj.image_height = obj.logo.height
obj.save(update_fields=["image_height", "image_width"])
def backward(apps, schema_editor):
apps.get_model("my-app.MyModel").objects.all().update(
image_width=None,
image_height=None,
)
class Migration(migrations.Migration):
dependencies = [
('my-app', '0001_initial'),
]
operations = [
migrations.AddField(
model_name="mymodel",
name="image_height",
field=models.PositiveIntegerField(editable=False, null=True),
),
migrations.AddField(
model_name="mymodel",
name="image_width",
field=models.PositiveIntegerField(editable=False, null=True),
),
migrations.RunPython(forward, backward),
AlterPictureField(
model_name="mymodel",
name="image",
field=pictures.models.PictureField(
aspect_ratios=["3/2", "3/1"],
breakpoints={"desktop": 1024, "mobile": 576},
container_width=1200,
file_types=["WEBP"],
grid_columns=12,
height_field="image_height",
pixel_densities=[1, 2],
storage=django.core.files.storage.FileSystemStorage(),
upload_to="pictures/",
verbose_name="image",
width_field="image_width",
),
),
]
This is a drop-in replacement for the Django ImageField that provides a standardized way to handle image uploads. It is designed to be as easy to use as possible, and to provide a consistent interface for all image fields. It allows images to be presented in various size variants (eg:thumbnails, mid, and hi-res versions) and it provides a way to handle images that are too large with validators.
Django Standardized Image Field implements the following features:
- Django-Storages compatible (eg: S3, Azure, Google Cloud Storage, etc)
- Resizes images to different sizes
- Access thumbnails on model level, no template tags required
- Preserves original images
- Can be rendered asynchronously (ie as a Celery job)
- Restricts acceptable image dimensions
- Renames a file to a standardized name format (using a callable
upload_to
function, see below)
Simply install the latest stable package using the following command:
pip install django-stdimage
# or
pipenv install django-stdimage
and add 'stdimage'
to INSTALLED_APP
s in your settings.py, that's it!
Now it's instally you can use either: StdImageField
or JPEGField
.
StdImageField
works just like Django's own
ImageField
except that you can specify different size variations.
The JPEGField
is identical to the StdImageField
but all images are
converted to JPEGs, no matter what type the original file is.
Variations are specified within a dictionary. The key will be the attribute referencing the resized image. A variation can be defined both as a tuple or a dictionary.
Example:
from django.db import models
from stdimage import StdImageField, JPEGField
class MyModel(models.Model):
# works just like django's ImageField
image = StdImageField(upload_to='path/to/img')
# creates a thumbnail resized to maximum size to fit a 100x75 area
image = StdImageField(upload_to='path/to/img',
variations={'thumbnail': {'width': 100, 'height': 75}})
# is the same as dictionary-style call
image = StdImageField(upload_to='path/to/img', variations={'thumbnail': (100, 75)})
# JPEGField variations are converted to JPEGs.
jpeg = JPEGField(
upload_to='path/to/img',
variations={'full': (None, None), 'thumbnail': (100, 75)},
)
# creates a thumbnail resized to 100x100 croping if necessary
image = StdImageField(upload_to='path/to/img', variations={
'thumbnail': {"width": 100, "height": 100, "crop": True}
})
## Full ammo here. Please note all the definitions below are equal
image = StdImageField(upload_to='path/to/img', blank=True, variations={
'large': (600, 400),
'thumbnail': (100, 100, True),
'medium': (300, 200),
}, delete_orphans=True)
To use these variations in templates use myimagefield.variation_name
.
Example:
<a href="{{ object.myimage.url }}"><img alt="" src="{{ object.myimage.thumbnail.url }}"/></a>
You can use a function for the upload_to
argument. Using [Django Dynamic Filenames][dynamic_filenames].[dynamic_filenames]: https://github.com/codingjoe/django-dynamic-filenames
This allows images to be given unique paths and filenames based on the model instance.
Example
from django.db import models
from stdimage import StdImageField
from dynamic_filenames import FilePattern
upload_to_pattern = FilePattern(
filename_pattern='my_model/{app_label:.25}/{model_name:.30}/{uuid:base32}{ext}',
)
class MyModel(models.Model):
# works just like django's ImageField
image = StdImageField(upload_to=upload_to_pattern)
The StdImageField
doesn't implement any size validation out-of-the-box.
However, Validation can be specified using the validator attribute
and using a set of validators shipped with this package.
Validators can be used for both Forms and Models.
Example
from django.db import models
from stdimage.validators import MinSizeValidator, MaxSizeValidator
from stdimage.models import StdImageField
class MyClass(models.Model):
image1 = StdImageField(validators=[MinSizeValidator(800, 600)])
image2 = StdImageField(validators=[MaxSizeValidator(1028, 768)])
CAUTION: The MaxSizeValidator should be used with caution. As storage isn't expensive, you shouldn't restrict upload dimensions. If you seek prevent users form overflowing your memory you should restrict the HTTP upload body size.
Django dropped support for automated deletions in version 1.3.
Since version 5, this package supports a delete_orphans
argument. It will delete
orphaned files, should a file be deleted or replaced via a Django form and the object with
the StdImageField
be deleted. It will not delete files if the field value is changed or
reassigned programatically. In these rare cases, you will need to handle proper deletion
yourself.
from django.db import models
from stdimage.models import StdImageField
class MyModel(models.Model):
image = StdImageField(
upload_to='path/to/files',
variations={'thumbnail': (100, 75)},
delete_orphans=True,
blank=True,
)
Tools like celery allow to execute time-consuming tasks outside of the request. If you don't want to wait for your variations to be rendered in request, StdImage provides you the option to pass an async keyword and a 'render_variations' function that triggers the async task. Note that the callback is not transaction save, but the file variations will be present. The example below is based on celery.
tasks.py
:
from django.apps import apps
from celery import shared_task
from stdimage.utils import render_variations
@shared_task
def process_photo_image(file_name, variations, storage):
render_variations(file_name, variations, replace=True, storage=storage)
obj = apps.get_model('myapp', 'Photo').objects.get(image=file_name)
obj.processed = True
obj.save()
models.py
:
from django.db import models
from stdimage.models import StdImageField
from .tasks import process_photo_image
def image_processor(file_name, variations, storage):
process_photo_image.delay(file_name, variations, storage)
return False # prevent default rendering
class AsyncImageModel(models.Model):
image = StdImageField(
# above task definition can only handle one model object per image filename
upload_to='path/to/file/', # or use a function
render_variations=image_processor # pass boolean or callable
)
processed = models.BooleanField(default=False) # flag that could be used for view querysets
You might have added or changed variations to an existing field. That means you will need to render new variations. This can be accomplished using a management command.
python manage.py rendervariations 'app_name.model_name.field_name' [--replace] [-i/--ignore-missing]
The replace
option will replace all existing files.
The ignore-missing
option will suspend 'missing source file' errors and keep
rendering variations for other files. Otherwise, the command will stop on first missing file.