matthewwithanm / django-imagekit Goto Github PK

When using the Admin thumbnails the ticket located in /virtualenv/.../templates/imagekit/admin are not found.
If I move those into my local template directory it works fine. Somehow it's not finding that template directory.

I eventually want to have a nice CI thing going here (first: more tests). Tox seems to be the flavor of the day when it comes to hassle-free testing, so I'll just leave this here for myself.

http://tox.testrun.org/en/latest/

Both attempts end with:

IOError: [Errno 2] No such file or directory: '/home/ned/project/sbo/ve/build/django-imagekit/README.rst'

Specifying ==0.4.0 installs fine.

Ran into and issue when using this on a cloud based hosting solution that had PIL already installed.

Read the following which can explain better than I could:

http://jaredforsyth.com/blog/2010/apr/28/accessinit-hash-collision-3-both-1-and-1/
http://projects.scipy.org/scipy/ticket/1374

Need to update to use the from PIL... import style

In v.1.0.3 on pypi the options parameter has been removed from the ImageSpec init.py, so installing via pip and using the instructions on the readme on github throws errors. Change the readme here to reflect the new syntax:

formatted_image = ImageSpec(image_field='original_image', format='JPEG',
        quality=90)

For some reason I'm getting the absolute path to the images. It should point to the media folder I specified in MEDIA_URL = 'http://mysite.com/mymedia/' in settings.py. Any ideas where the problem is?

It'd be ideal if the pip version of this was at least 1.0 instead of 0.4. The version on github and the version on pip are almost totally different now but the install instructions stil recommend pip.

In a model,
I have an attached spec with a "Format" processor so that my thumbnail images are all in jpeg.

I first upload a jpeg file say "a.PNG". spec works fine and produces "a_thumbnail.jpg" file with the format: (file_name - a)_(spec_name - thumbnail).(processor_output_extension - JPEG)

But,
when i upload a file after the first one say "a.TIF" (same name - different extension).
spec works again output file format will be: "a_thumbnail.jpg" again. Since this file already generated for previous image, it just uses this one, which is really bad!

A solution may be to include "original_file_extension" in the cached file name.
Technically speaking:
options.py line 17:
cache_filename_format = "%(filename)s_%(specname)s_%(original_extension)s.%(extension)s"

delete() on a model is not called when performing bulk actions of when a model is deleted via a relation. This means the cache is not being cleaned up.

Would it be possible to hook-up the cleanup using a post_delete signal instead?

The intent is great — as I understand it, refresh the spec files when instance is saved because the image may have been updated — but it wreaks havoc when the image spec url accessor isn't used in templates and pre_cache is not set to True.

When pre-cache is True, the spec files are generated immediately after deletion. But if a post_save handler is the desired method to generate spec files when the instance is created (to work well with celery, for instance), and calling the image size directly in templates isn't ideal (in my use-case, because I'm using cloudfiles for spec file storage and there's a second or three lag during this operation) then images vanish on future instance saves and never get regenerated.

In order to show a thumbnail image, add following lines to admin.py;

class PhotoAdmin(admin.ModelAdmin):
admin_thumbnail = AdminThumbnail(image_field='thumbnail')
list_display = ('caption, 'admin_thumbnail')
fields = ['admin_thumbnail', 'original_image', 'caption']

The thumbnail is shown in the list and the detail page.
But when adding a new photo, the error occurs;

File "/usr/lib/python2.6/site-packages/django/template/debug.py" in render_node

1.         result = node.render(context)
   

   File "/usr/lib/python2.6/site-packages/django/template/debug.py" in render

2.         output = self.filter_expression.resolve(context)
   

   File "/usr/lib/python2.6/site-packages/django/template/base.py" in resolve

3.             obj = self.var.resolve(context)
   

   File "/usr/lib/python2.6/site-packages/django/template/base.py" in resolve

4.         value = self._resolve_lookup(context)
   

   File "/usr/lib/python2.6/site-packages/django/template/base.py" in _resolve_lookup

5.                         current = current()
   

   File "/usr/lib/python2.6/site-packages/django/contrib/admin/helpers.py" in contents

6.         f, attr, value = lookup_field(field, obj, model_admin)
   

   File "/usr/lib/python2.6/site-packages/django/contrib/admin/util.py" in lookup_field

7.         value = attr(obj)
   

   File "/usr/lib/python2.6/site-packages/imagekit/admin.py" in call

8.             obj, self.image_field))
   

Exception Type: TemplateSyntaxError at /admin/lupines/photo/add/
Exception Value: Caught Exception while rendering: The property - None is not defined on thumbnail.

And local values are;

self imagekit.admin.AdminThumbnail object at 0xa6409ac
obj Photo: - None
thumbnail ImageSpecFile: None

A nasty bug in the extension suggestion algorithm can cause images to be unnecessarily regenerated and even duplicated on the filesystem.

The reason is that IK relies on PIL's EXTENSION list to map formats to extensions. However, under normal circumstances, this list isn't populated until an image is opened or saved. Therefore, the first time a spec is accessed (assuming you haven't already used PIL to do something else), you can wind up with a different (wrong) suggested extension. IK will then check for the existence of the file with the wrong extension. Of course, it won't exist, so IK will generate a new image. Since Image.open will have been called in the course of generating the image, the new suggested filename will be correct. This process will be repeated exactly each subsequent time (since the file with the incorrect extension still won't exist) but since a file with the correct extension will already exist, the storage backend will append a count, resulting in a filesystem full of duplicate images!

I guess this is half question, half request.
So, let's say I have a model class PersonPhoto(ImageModel) (among others) that will be used in various places on a site, in various shapes. It has ImageSpecs of FacebookStyleThumb, SmallPortrait, SmallLandscape. The width and height of these would be, say, 50,50, 120,170, and 170,120 respectively.

I would love to take PersonPhoto instances and be able to use some fancy js (jCrop) to define cropboxes for the various ImageSpecs, enforcing the aspect ratio defined by their width/height. Then save that cropbox in the DB. I have it working, but it's hacky and am now trying to do it a "better" way.

To store the cropbox dimensions, I'm thinking something like this:

class CropBox(models.Model):
    some_sort_of_unique_imagespec_value = models.CharField(max_length=255, help_text='')
    content_type = models.ForeignKey(ContentType)
    object_id = models.PositiveIntegerField()
    content_object = generic.GenericForeignKey('content_type', 'object_id')

    x = models.PositiveSmallIntegerField(help_text='')
    y = models.PositiveSmallIntegerField(help_text='')
    x2 = models.PositiveSmallIntegerField(help_text='')
    y2 = models.PositiveSmallIntegerField(help_text='')

    class Meta:
        unique_together = (('some_sort_of_unique_imagespec_value', 'content_type', 'object_id'), ) 

I'm not sure of the best way to easily get and save cropbox instances.
Do I handle getting the cropbox instance in def process() of a custom crop ImageProcessor, using its classname as the some_sort_of_unique_imagespec_value... But if I did that, I'm not seeing an easy way to "save" cropbox dimensions. Maybe it should be handled in a custom ImageSpec.... or maybe an Accessor attribute... or maybe I'm thinking about this all wrong...

Anything that could be done on the imagekit side, at a low level, that might make a path to something like this more clear?

Since @hzy- came up with a pretty straightforward use case for generating files manually (#51), I think we should probably make _create a non-underscore function. I'm thinking generate() would probably be a better name, but we can bikeshed on that. I'm also thinking that we should change the "lazy" kwarg to default to true so that people don't accidentally regenerate files all the time.

f570bd0 changed the way images were opened, and wraps the original copy method with one that tries to preserve extra imagekit related info.

however this results in the image data being retained somehow in memory - I'm not sure exactly how or where.

commenting out this line:

img.copy = types.MethodType(_wrap_copy(img.copy), img, img.__class__)

results in a substantially reduced memory footprint

I've created a quick sample app/project to demonstrate this - if you watch the process memory - you can see it climb.

https://github.com/ptone/imagefoobar (note this is using the new Django 1.4/trunk layout)

cc @lettertwo

With the old ImageKit API, I was able to create a single spec module at the site root that I could share across all of my image models within a project. This allowed me to have standard image sizes to use in various layouts. All I'd have to do is have that model inherit ImageModel and provide the path to the spec file.

Now, it seems that for each model that I want these ImageSpecs, I'll have to recreate each of the specs as fields on each individual model. And if I want to add or change specs, I'd need to edit each and every model.

Am I missing something with how this new API works? Is there a way to specify specs without having to spell out all the properties each time?

One thought that I had was to create an abstract model mixin that contains the site-wide specs have each of the models with images inherit that mixin. But this would require being very careful that the imageField be named the same field name.

The other problem is working with third party apps. Previously, I could create a proxy model that would give a third party model the ImageModel behavior and attach the spec file to it. I suppose I could do something similar now, but again the issue of field names remains.

If a new file is uploaded on an existing model, replacing the original photo, the preprocessor_spec is not run and the original image is saved.

I did a clean install of my Django dev environment, and got stuck here. I Google-d quite for sometime to figure this out, seems like it's a namespace thing.

Anyone else having this problem ? Or is it something I did wrong ? From what I've searched, this problem seems to have been solved ?

I have to remove the below codes in lib.py to get it working again.

I'm using latest clone of django, imagekit and PIL binaries from here

#try:
#   import Image
#   import ImageFile
#   import ImageFilter
#   import ImageEnhance
#   import ImageColor
# except ImportError:
try:
    from PIL import Image
    from PIL import ImageFile
    from PIL import ImageFilter
    from PIL import ImageEnhance
    from PIL import ImageColor
except ImportError:
    raise ImportError('ImageKit was unable to import the Python Imaging Library. Please confirm it`s installed and available on your current Python path.')

Hello,
is there any way how to define ImageSpec fields dynamically please? I have reusable application Photo and I would like define different photo variants per project (for example in project1 fields thumbnail, catalog and in project2 fields thumbnail and detail; in both cases there will be different crop sizes for ImageSpec fields).

I used ImageKit in past, and that time I should define effectivelly specs.py per project with own photo variants.

Is something like this possible in this new "Field API" approach please?

Thanks for any hint.

Specifying an explicit format="JPEG" does not work when the input file is in PNG format. Works fine when input is JPEG (haven't tested the usual suspects such as GIF, TIFF yet). Also works fine when format= is NOT specified.

class ZDPlainImage(EncryptedPKModel):

    original = ImageField(max_length=500, upload_to="uploads")

     # ImageKit generated image variations
     thumbnail = ImageSpec([resize.Fit(width=250)], image_field='original', format="JPEG", quality=90)
     display = ImageSpec([resize.Fit(width=650, upscale=False)], image_field='original', format="JPEG", quality=90)

Looks like this used to work well with the old 0.4 version. I updated to the latest django-imagekit today, made the necessary changes (got rid of separate "spec") and found that this doesn't work with the latest version.

here's the code in my model. The thumbnail works great. But the display isn't working. Suggestions welcome.

from imagekit.models import ImageSpec
from imagekit.processors import resize, Adjust

class Speaker(models.Model):
    first_name = models.CharField(max_length=100)
    last_name = models.CharField(max_length=100)
    bio = models.TextField(blank=True)
    original_image = models.ImageField(upload_to='speaker_photos')
    display = ImageSpec(image_field='original_image', options={'quality': 70})
    thumbnail = ImageSpec([Adjust(contrast=1.2, sharpness=1.1), resize.Crop(50, 50)], image_field='original_image', options={'quality': 50})

Is there any support planned for Progressive JPEGs? (of course, only in case the target format is JPEG)

This would be quite useful since progressive JPEGs seem to be the preferred way of displaying photos for online photo apps.

The test suite should include tests for format_to_extension and extension_to_format.

How can we admin a model contains a thumbnail with InlineModelAdmin or not?

For example, put these lines to admin.py;

class PhotoInline(admin.TabularInline):
model = Photo
admin_thumbnail = AdminThumbnail(image_field='thumbnail')
fields = ('admin_thumbnail', 'caption')

class ProductAdmin(admin.ModelAdmin):
inlines = [
PhotoInline,
]

This cause the following error;
Exception Type: ImproperlyConfigured
Exception Value: 'PhotoInline.fields' refers to field 'admin_thumbnail' that is missing from the form.

Add support for a pluggable backends system for image manipulations so that other image libraries other than PIL (like ImageMagick) could be used to manipulate images.

Also, I'd like a pony please. ;)

I want to assign the processors parameter for each instance.
For example, a ImageSpec is following;

display  = ImageSpec([Adjust(contrast=1.2, sharpness=1.1),Transpose(Transpose.ROTATE_90),
        resize.Fit(300, None)], image_field='original_image',
        format='JPEG', quality=90,cache_to=my_cache_to)

and I want to change contrast, sharpness, and the Transpose method at runtime for each instance.

I found the similar request at #11 issue.
#11 expects more general way, but I don't have a good idea for #11

My idea is;

1. change "processors" of ImageSpec() to check callable or not, like "cache_to".
2. the callable "processors" gets parameters from instance of Model and generates processors array

Is it an effective change or not?

With the addition of the AutoConvert processor, ImageKit can now handle different transparency formats a lot more intelligently than it used to; the Adjust processor needs to be updated to match.

The first problem is that the processor converts to 'RGB'. This is a pretty quick fix as we can just change this to RGBA. Unfortunately, that's not quite enough: PIL's color and contrast filters don't behave as expected with transparencies. That's because both of their implementations involve converting the image to L mode:

class Color(_Enhance):
    "Adjust image colour balance"
    def __init__(self, image):
        self.image = image
        self.degenerate = image.convert("L").convert(image.mode)

class Contrast(_Enhance):
    "Adjust image contrast"
    def __init__(self, image):
        self.image = image
        mean = int(ImageStat.Stat(image.convert("L")).mean[0] + 0.5)
        self.degenerate = Image.new("L", image.size, mean).convert(image.mode)

So basically, we have to write new versions of these two that don't.

Any takers? (:

Anyone else encountering an issue using the ImageModel in an Admin form - uploading a file causes the file pre_save signal to be called which instantiates an InMemoryUploadedFile and fails when it tries to test if the '_committed' Attribute exists.

Traceback of the error I get:

File "/var/www/lib/python2.7/site-packages/django/core/handlers/base.py" in get_response
  111.                         response = callback(request, *callback_args, **callback_kwargs)
File "/var/www/lib/python2.7/site-packages/django/contrib/admin/options.py" in wrapper
  307.                 return self.admin_site.admin_view(view)(*args, **kwargs)
File "/var/www/lib/python2.7/site-packages/django/utils/decorators.py" in _wrapped_view
  93.                     response = view_func(request, *args, **kwargs)
File "/var/www/lib/python2.7/site-packages/django/views/decorators/cache.py" in _wrapped_view_func
  79.         response = view_func(request, *args, **kwargs)
File "/var/www/lib/python2.7/site-packages/django/contrib/admin/sites.py" in inner
  197.             return view(request, *args, **kwargs)
File "/var/www/lib/python2.7/site-packages/django/utils/decorators.py" in _wrapper
  28.             return bound_func(*args, **kwargs)
File "/var/www/lib/python2.7/site-packages/django/utils/decorators.py" in _wrapped_view
  93.                     response = view_func(request, *args, **kwargs)
File "/var/www/lib/python2.7/site-packages/django/utils/decorators.py" in bound_func
  24.                 return func(self, *args2, **kwargs2)
File "/var/www/lib/python2.7/site-packages/django/db/transaction.py" in inner
  217.                 res = func(*args, **kwargs)
File "/var/www/lib/python2.7/site-packages/django/contrib/admin/options.py" in change_view
  982.                 self.save_model(request, new_object, form, change=True)
File "/var/www/lib/python2.7/site-packages/django/contrib/admin/options.py" in save_model
  665.         obj.save()
File "/var/www/yipit/apps/deal/models.py" in save
  198.         super(Deal, self).save(*args, **kwargs)
File "/var/www/src/django-imagekit/imagekit/models.py" in save
  124.         super(ImageModel, self).save(*args, **kwargs)
File "/var/www/lib/python2.7/site-packages/django/db/models/base.py" in save
  460.         self.save_base(using=using, force_insert=force_insert, force_update=force_update)
File "/var/www/lib/python2.7/site-packages/django/db/models/base.py" in save_base
  504.                 self.save_base(cls=parent, origin=org, using=using)
File "/var/www/lib/python2.7/site-packages/django/db/models/base.py" in save_base
  525.                         values = [(f, None, (raw and getattr(self, f.attname) or f.pre_save(self, False))) for f in non_pks]
File "/var/www/lib/python2.7/site-packages/django/db/models/fields/files.py" in pre_save
  253.         if file and not file._committed:

Exception Type: AttributeError
Exception Value: 'InMemoryUploadedFile' object has no attribute '_committed'

Would appreciate any input on this. Thanks,
Nitya

While poorly documented, Django's ImageField supports the 'default' option for field value.

on a related note, while the value of the default is made available as the path to the ProcessedImageField, any imagespec that references will run into trouble like this:

Traceback (most recent call last):
  File "<input>", line 1, in <module>
  File "/Users/preston/Projects/code/forks/django-imagekit/imagekit/models.py", line 178, in url
    self.generate()
  File "/Users/preston/Projects/code/forks/django-imagekit/imagekit/models.py", line 191, in generate
    if lazy and (getattr(self, '_file', None) or self.storage.exists(self.name)):
  File "/Users/preston/UNIX/src/django-trunk-svn/django/core/files/storage.py", line 230, in exists
    return os.path.exists(self.path(name))
  File "/Users/preston/UNIX/src/django-trunk-svn/django/core/files/storage.py", line 246, in path
    raise SuspiciousOperation("Attempted access to '%s' denied." % name)
SuspiciousOperation: Attempted access to '/Users/preston/Projects/code/fotw/project/fotw_project/static/image/missing_thumbnail.jpg' denied.

Just an idea, how about a processor that can rename the file on save, eg, renaming the file to the primary key of the object.

Hey, I have my Photo model as you can see below, and I would like to have the 4 different image sizes created when a new photo is uploaded, or there abouts. I was wondering if there's been any attempt to use something like celery to have a background queue for generating images?

thanks.

class Photo(models.Model):
    user = models.ForeignKey(User)
    title = models.CharField(max_length = 100, blank=True)
    description = models.TextField(blank=True)
    is_active = models.BooleanField(default=False)
    original_image = models.ImageField(upload_to = imagefield_upload_to)

    thumbnail = ImageSpec(
        processors = [
            resize.Crop(width=100, height=100),
            Adjust(sharpness=1.2),
        ],
        image_field = 'original_image',
        format = 'JPEG',
        quality = 85,
        pre_cache = True,
        cache_to = 'photos/t',
    )

    large_image = ImageSpec(
        processors = [
            resize.Fit(width=1024, height=1024),
            Adjust(sharpness=1.2),
        ],
        image_field = 'original_image',
        format = 'JPEG',
        quality = 85,
        pre_cache = True,
        cache_to = 'photos/l',
    )

    medium_image = ImageSpec(
        processors = [
            resize.Fit(width=640, height=640),
            Adjust(sharpness=1.2),
        ],
        image_field = 'original_image',
        format = 'JPEG',
        quality = 85,
        pre_cache = True,
        cache_to = 'photos/m',
    )

    small_image = ImageSpec(
        processors = [
            resize.Fit(width=240, height=240),
            Adjust(sharpness=1.2),
        ],
        image_field = 'original_image',
        format = 'JPEG',
        quality = 85,
        pre_cache = True,
        cache_to = 'photos/s',
    )

    def __unicode__(self):
        return u'%s by %s' % (self.title, self.user.username)

In the following locations:
django-imagekit/docs/tutorial.rst
django-imagekit/README.rst

class EnchanceThumb(processors.Adjustment):
contrast = 1.2
sharpness = 1.1

class Thumbnail(ImageSpec):
access_as = 'thumbnail_image'
pre_cache = True
processors = [ResizeThumb, EnchanceThumb]

Hello,

Documentation says 'pip install django-imagekit' is enough to install imagekit but if you don't have PIL/Pillow installed in your environment you can't use imagekit until you install these.

Is changing documentation the right way or setup.py?

Regards.

It would be great for imagekit to be able to accept and process a default image. (ie, if no image supplied, use the image specified in the parameter 'default' instead)

I did a clean install of my Django dev environment, and got stuck here. I Google-d quite for sometime to figure this out, seems like it's a namespace thing.

Anyone else having this problem ? Or is it something I did wrong ? From what I've searched, this problem seems to have been solved ?

I have to remove the below codes in lib.py to get it working again.

I'm using latest clone of django, imagekit and PIL binaries from here

#try:
#   import Image
#   import ImageFile
#   import ImageFilter
#   import ImageEnhance
#   import ImageColor
# except ImportError:
try:
    from PIL import Image
    from PIL import ImageFile
    from PIL import ImageFilter
    from PIL import ImageEnhance
    from PIL import ImageColor
except ImportError:
    raise ImportError('ImageKit was unable to import the Python Imaging Library. Please confirm it`s installed and available on your current Python path.')

I'm probably doing something stupid, but I've been trying to follow the documentation, and it appears files are missing from the downloads. ProcessorPipeline and AutoConvert are not in processors/resize.py, which is the only file in that directory; therefore attempting to generate any new image fails. I've tried cloning the git archive and downloading the tar files, neither contains anything except resize.py in the processor directory.

Hi. What is the right way to delete a ImageModel?
I tried this 3 ways:

1. Trying to delete using the ImageModel.delete() method:

   photo.delete()
   

   What happens: This cleans the cache, but "forget" the original image on the disk
2. Trying to delete the original image before calling the ImageModel.delete() method:

   photo.original_image.delete()
   photo.delete()
   

   What happens: This deletes the original image, but "forget" about the cached ones
3. Trying to delete the original image after calling ImageModel.delete()

   # photo.id == 1
   photo.delete()
   # photo.id == None
   photo.original_image.delete()
   # photo.id == 2
   

   What happens: This will clean the cache and delete the original_image, but then it will re-create the model with no images.

So, how can I delete the cache, the original image and the database entry?
Thanks!

I'm not sure whether this is by-design, but what are my options if i need two images associated with a model, for example? The IKOptions seems to only let me define a single image in image_field. Am i missing something?

Is there a complete tutorial on how to upload, update and delete images using ImageKit?

On my project I want for example to update avatar images if an avatar image exits for that user. So my script should be able to update the database table but also delete the existing image and upload the new one.

When I try to fire up ./manage.py celeryd after integrating the latest imagekit into my models, I get this error:

AttributeError: 'ImageSpecFile' object has no attribute 'source_file'

https://gist.github.com/1352084

Am I doing something wrong (hopefully) or is celery incompatible with imagekit?

Thanks,
Dave

If an imagespec references an imagefield that does not have a file, this will raise a ValueError exception.

This can cause entire gallery pages or the admin to err even if the issue is just with a single model

I propose a setting: IMAGEKIT_MISSING_IMAGE which would be a str URL, or a callable (not sure what if any args it would take, maybe the original imagespec - or the model instance?)

This would then be the URL used for ie, model_instance.thumbnail_imagespec.url

a callable would allow something like the use of http://placehold.it

I am getting an error when an thumbnail doesn't exist in the cache for an image used in a rss feed.
handler.addQuickElement(u"url", show.img_itunes_sm.url) from https://github.com/rizumu/django-podcasting/blob/master/podcasting/feeds.py

The error is triggerd in utils.py when calling img.save():

    ipdb> img
    <PIL.Image.Image image mode=RGB size=1000x1000 at 0x3A64290>
    ipdb> img.save(tmp, format, **kwargs)
    Suspension not allowed here
    *** IOError: encoder error -2 when writing image file

and the traceback

 Environment:

Request Method: GET
Request URL: http://snowprayers.net/feeds/podcasts/snowprayers/itunes/mp3/rss/


Traceback:
File "/home/deploy/.virtualenvs/snowprayers/lib/python2.7/site-packages/django/core/handlers/base.py" in get_response
  111.                         response = callback(request, *callback_args, **callback_kwargs)
File "/home/deploy/.virtualenvs/snowprayers/lib/python2.7/site-packages/newrelic-0.5.58.122/newrelic/hooks/framework_django.py" in __call__
  327.                     return self.__wrapped(*args, **kwargs)
File "/home/deploy/.virtualenvs/snowprayers/lib/python2.7/site-packages/django/contrib/syndication/views.py" in __call__
  39.         feedgen.write(response, 'utf-8')
File "/home/deploy/.virtualenvs/snowprayers/lib/python2.7/site-packages/django/utils/feedgenerator.py" in write
  214.         self.add_root_elements(handler)
File "/home/deploy/.virtualenvs/snowprayers/lib/python2.7/site-packages/podcasting/feeds.py" in add_root_elements
  39.         handler.addQuickElement(u"itunes:image", attrs={"href": show.img_itunes_lg.url})
File "/home/deploy/.virtualenvs/snowprayers/lib/python2.7/site-packages/imagekit/models.py" in url
  159.         self._create(True)
File "/home/deploy/.virtualenvs/snowprayers/lib/python2.7/site-packages/imagekit/models.py" in _create
  184.             img, content = self._process_content(self.name, fp)
File "/home/deploy/.virtualenvs/snowprayers/lib/python2.7/site-packages/imagekit/models.py" in _process_content
  135.                                   optimize=True)
File "/home/deploy/.virtualenvs/snowprayers/lib/python2.7/site-packages/imagekit/utils.py" in img_to_fobj
  18.     img.save(tmp, format, **kwargs)
File "/home/deploy/.virtualenvs/snowprayers/lib/python2.7/site-packages/PIL/Image.py" in save
  1439.             save_handler(self, fp, filename)
File "/home/deploy/.virtualenvs/snowprayers/lib/python2.7/site-packages/PIL/JpegImagePlugin.py" in _save
  471.     ImageFile._save(im, fp, [("jpeg", (0,0)+im.size, 0, rawmode)])
File "/home/deploy/.virtualenvs/snowprayers/lib/python2.7/site-packages/PIL/ImageFile.py" in _save
  501.                 raise IOError("encoder error %d when writing image file" % s)

Exception Type: IOError at /feeds/podcasts/snowprayers/itunes/mp3/rss/
Exception Value: encoder error -2 when writing image file

I'll attempt to figure it out over the weekend.

There are pieces of the old ImageKit wiki on Bitbucket that should be pulled into the up-and-coming documentation:

• https://bitbucket.org/jdriscoll/django-imagekit/wiki/Options
• https://bitbucket.org/jdriscoll/django-imagekit/wiki/Tips
• https://bitbucket.org/jdriscoll/django-imagekit/wiki/OldWatermark

This'll be helpful.

If I have a regular imagefield, I can call it's save method with a django File object as the content

but with a ProcessedImageField, this causes PIL to choke

instead the content needs to be a plain python file like

open('/some/path.jpg', 'rb')

Not sure if the solution is some form of isinstance check?

  File "/Users/preston/Projects/Python/virtualenvs/fotw/src/imagekit/imagekit/models.py", line 339, in save
    img, content = self._process_content(new_filename, content)
  File "/Users/preston/Projects/Python/virtualenvs/fotw/src/imagekit/imagekit/models.py", line 125, in _process_content
    img = open_image(content)
  File "/Users/preston/Projects/Python/virtualenvs/fotw/src/imagekit/imagekit/utils.py", line 34, in open_image
    img = Image.open(target)
  File "/Library/Python/2.6/site-packages/PIL/Image.py", line 1980, in open
    raise IOError("cannot identify image file")
IOError: cannot identify image file

The readthedocs site isn't building all of the documentation (specifically, the API reference). I think the autodoc extension needs to be enabled.

Hello!

I installed django-imagekit and created my ImovelPhoto class, but when running syncdb appears the error:

ImportError: cannot import name ImageModel

PrintScreen: http://postimage.org/image/1dulp0744/

I use Eclipse and Helios with PyDev and Django-1.3
You know what might be happening?
Thanks!

I want to make thumbnails for over 10000 JPEG files.
python2.6 manage.py ikflush app
makes some thumbnails and abnormally ends.

Traceback (most recent call last):
File "manage.py", line 11, in
File "/usr/lib/python2.6/site-packages/django/core/management/init.py", line 438, in execute_manager
File "/usr/lib/python2.6/site-packages/django/core/management/init.py", line 379, in execute
File "/usr/lib/python2.6/site-packages/django/core/management/base.py", line 191, in run_from_argv
File "/usr/lib/python2.6/site-packages/django/core/management/base.py", line 220, in execute
File "/usr/lib/python2.6/site-packages/imagekit/management/commands/ikflush.py", line 15, in handle
File "/usr/lib/python2.6/site-packages/imagekit/management/commands/ikflush.py", line 34, in flush_cache
File "/usr/lib/python2.6/site-packages/imagekit/specs.py", line 67, in _create
File "/usr/lib/python2.6/site-packages/imagekit/specs.py", line 50, in _get_imgfile
File "/usr/lib/python2.6/site-packages/imagekit/utils.py", line 6, in img_to_fobj
File "/usr/lib64/python2.6/tempfile.py", line 475, in TemporaryFile
File "/usr/lib64/python2.6/tempfile.py", line 228, in _mkstemp_inner
OSError: [Errno 24] Too many open files: '/tmp/tmp3zbmaD'

Following the docs.

Step 1. pip install django-imagekit:
Successfully installed django-imagekit

Step 2. add "from imagekit.models import ImageModel" to my model
ImportError: cannot import name ImageModel

When deleting a file image model (like the one in the example) neither file or thumbnail are removed from the FS.

Is anyone else getting failing unit tests? As a side effect, the test images are not getting destroyed at the end of the tests as well.