Comment

zgoda

At least on my machine these two are not the same:

In [14]: import PIL

In [15]: PIL.__file__
Out[15]: '/opt/local/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/PIL/__init__.pyc'

In [16]: import Image

In [17]: Image.__file__
Out[17]: '/opt/local/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/PIL/Image.pyc'

In [18]: dir(PIL)
Out[18]: ['__builtins__', '__doc__', '__file__', '__name__', '__path__']

In [19]: dir(Image)
Out[19]:
['ADAPTIVE',
'AFFINE',
'ANTIALIAS',
'BICUBIC',
'BILINEAR',
'CONTAINER',
'CUBIC',
'DEBUG',
'EXTENSION',
'EXTENT',
'FLIP_LEFT_RIGHT',
'FLIP_TOP_BOTTOM',
'FLOYDSTEINBERG',
'ID',
'Image',
'ImageMode',
'ImagePalette',
'IntType',
'LINEAR',
'MESH',
'MIME',
'MODES',
'NEAREST',
'NONE',
'NORMAL',
'OPEN',
'ORDERED',
'PERSPECTIVE',
'QUAD',
'RASTERIZE',
'ROTATE_180',
'ROTATE_270',
'ROTATE_90',
'SAVE',
'SEQUENCE',
'StringType',
'TupleType',
'UnicodeStringType',
'VERSION',
'WEB',
'_E',
'_ENDIAN',
'_ImageCrop',
'_MAPMODES',
'_MODEINFO',
'_MODE_CONV',
'__builtins__',
'__doc__',
'__file__',
'__name__',
'_conv_type_shape',
'_getdecoder',
'_getencoder',
'_getscaleoffset',
'_imaging_not_installed',
'_initialized',
'_showxv',
'_wedge',
'blend',
'composite',
'core',
'eval',
'fromarray',
'frombuffer',
'fromstring',
'getmodebandnames',
'getmodebands',
'getmodebase',
'getmodetype',
'init',
'isDirectory',
'isImageType',
'isNumberType',
'isSequenceType',
'isStringType',
'isTupleType',
'merge',
'new',
'open',
'os',
'preinit',
'register_extension',
'register_mime',
'register_open',
'register_save',
'string',
'sys',
'warnings']

Replies

Tomasz Elendt

I believe this is the real problem:

>>> from PIL import Image as A
>>> import Image as B
>>> B.__file__ == A.__file__
True