Use pilkit's process_image utility

imagekit/specs/__init__.py

@@ -4,8 +4,7 @@ from ..cachefiles.backends import get_default_cachefile_backend
 from ..cachefiles.strategies import StrategyWrapper
 from .. import hashers
 from ..exceptions import AlreadyRegistered, MissingSource
-from ..processors import ProcessorPipeline
-from ..utils import open_image, img_to_fobj, get_by_qname
+from ..utils import open_image, get_by_qname, process_image
 from ..registry import generator_registry, register
 
 
@@ -137,16 +136,9 @@ class ImageSpec(BaseImageSpec):
             self.source.open()
             img = open_image(self.source)
 
-        original_format = img.format
-
-        # Run the processors
-        processors = self.processors
-        img = ProcessorPipeline(processors or []).process(img)
-
-        options = dict(self.options or {})
-        format = self.format or img.format or original_format or 'JPEG'
-        content = img_to_fobj(img, format, **options)
-        return content
+        return process_image(img, processors=self.processors,
+                             format=self.format, autoconvert=self.autoconvert,
+                             options=self.options)
 
 
 def create_spec_class(class_attrs):

setup.py

@@ -50,7 +50,7 @@ setup(
     test_suite='testrunner.run_tests',
     install_requires=[
         'django-appconf>=0.5',
-        'pilkit',
+        'pilkit>=0.2.0',
     ],
     extras_require={
         'async': ['django-celery>=3.0'],