Performing a negative crop on an image previously just returned a (0, 0)
image. Now it will raise a ValueError
, to help reduce confusion if a user has unintentionally provided the wrong arguments.
Rather than returning a SystemError
, passing the incorrect types of coordinates into a path will now raise a more specific ValueError
, with the message "incorrect coordinate type".
Rather than installing all dependencies for docs and tests via requirements.txt
, extras_require
is used instead. This installs only those needed and at the same time as installing Pillow.
For example:
# Install with dependencies for tests:
python3 -m pip install .[tests]
# Or for building docs:
python3 -m pip install .[docs]
# Or for all:
python3 -m pip install .[docs,tests]
On macOS, the last argument may need to be wrapped in quotes, e.g. python3 -m pip install ".[tests]"
Therefore requirements.txt
has been removed along with the make install-req
command for installing its contents.
A number of constants have been deprecated and will be removed in Pillow 10.0.0 (2023-07-01). Instead, enum.IntEnum
classes have been added.
Deprecated | Use instead |
---|---|
Image.NONE |
Either Image.Dither.NONE or Image.Resampling.NEAREST |
Image.NEAREST |
Either Image.Dither.NONE or Image.Resampling.NEAREST |
Image.ORDERED |
Image.Dither.ORDERED |
Image.RASTERIZE |
Image.Dither.RASTERIZE |
Image.FLOYDSTEINBERG |
Image.Dither.FLOYDSTEINBERG |
Image.WEB |
Image.Palette.WEB |
Image.ADAPTIVE |
Image.Palette.ADAPTIVE |
Image.AFFINE |
Image.Transform.AFFINE |
Image.EXTENT |
Image.Transform.EXTENT |
Image.PERSPECTIVE |
Image.Transform.PERSPECTIVE |
Image.QUAD |
Image.Transform.QUAD |
Image.MESH |
Image.Transform.MESH |
Image.FLIP_LEFT_RIGHT |
Image.Transpose.FLIP_LEFT_RIGHT |
Image.FLIP_TOP_BOTTOM |
Image.Transpose.FLIP_TOP_BOTTOM |
Image.ROTATE_90 |
Image.Transpose.ROTATE_90 |
Image.ROTATE_180 |
Image.Transpose.ROTATE_180 |
Image.ROTATE_270 |
Image.Transpose.ROTATE_270 |
Image.TRANSPOSE |
Image.Transpose.TRANSPOSE |
Image.TRANSVERSE |
Image.Transpose.TRANSVERSE |
Image.BOX |
Image.Resampling.BOX |
Image.BILINEAR |
Image.Resampling.BILINEAR |
Image.LINEAR |
Image.Resampling.BILINEAR |
Image.HAMMING |
Image.Resampling.HAMMING |
Image.BICUBIC |
Image.Resampling.BICUBIC |
Image.CUBIC |
Image.Resampling.BICUBIC |
Image.LANCZOS |
Image.Resampling.LANCZOS |
Image.ANTIALIAS |
Image.Resampling.LANCZOS |
Image.MEDIANCUT |
Image.Quantize.MEDIANCUT |
Image.MAXCOVERAGE |
Image.Quantize.MAXCOVERAGE |
Image.FASTOCTREE |
Image.Quantize.FASTOCTREE |
Image.LIBIMAGEQUANT |
Image.Quantize.LIBIMAGEQUANT |
ImageCms.INTENT_PERCEPTUAL |
ImageCms.Intent.PERCEPTUAL |
ImageCms.INTENT_RELATIVE_COLORMETRIC |
ImageCms.Intent.RELATIVE_COLORMETRIC |
ImageCms.INTENT_SATURATION |
ImageCms.Intent.SATURATION |
ImageCms.INTENT_ABSOLUTE_COLORIMETRIC |
ImageCms.Intent.ABSOLUTE_COLORIMETRIC |
ImageCms.DIRECTION_INPUT |
ImageCms.Direction.INPUT |
ImageCms.DIRECTION_OUTPUT |
ImageCms.Direction.OUTPUT |
ImageCms.DIRECTION_PROOF |
ImageCms.Direction.PROOF |
ImageFont.LAYOUT_BASIC |
ImageFont.Layout.BASIC |
ImageFont.LAYOUT_RAQM |
ImageFont.Layout.RAQM |
BlpImagePlugin.BLP_FORMAT_JPEG |
BlpImagePlugin.Format.JPEG |
BlpImagePlugin.BLP_ENCODING_UNCOMPRESSED |
BlpImagePlugin.Encoding.UNCOMPRESSED |
BlpImagePlugin.BLP_ENCODING_DXT |
BlpImagePlugin.Encoding.DXT |
BlpImagePlugin.BLP_ENCODING_UNCOMPRESSED_RAW_RGBA |
BlpImagePlugin.Encoding.UNCOMPRESSED_RAW_RGBA |
BlpImagePlugin.BLP_ALPHA_ENCODING_DXT1 |
BlpImagePlugin.AlphaEncoding.DXT1 |
BlpImagePlugin.BLP_ALPHA_ENCODING_DXT3 |
BlpImagePlugin.AlphaEncoding.DXT3 |
BlpImagePlugin.BLP_ALPHA_ENCODING_DXT5 |
BlpImagePlugin.AlphaEncoding.DXT5 |
FtexImagePlugin.FORMAT_DXT1 |
FtexImagePlugin.Format.DXT1 |
FtexImagePlugin.FORMAT_UNCOMPRESSED |
FtexImagePlugin.Format.UNCOMPRESSED |
PngImagePlugin.APNG_DISPOSE_OP_NONE |
PngImagePlugin.Disposal.OP_NONE |
PngImagePlugin.APNG_DISPOSE_OP_BACKGROUND |
PngImagePlugin.Disposal.OP_BACKGROUND |
PngImagePlugin.APNG_DISPOSE_OP_PREVIOUS |
PngImagePlugin.Disposal.OP_PREVIOUS |
PngImagePlugin.APNG_BLEND_OP_SOURCE |
PngImagePlugin.Blend.OP_SOURCE |
PngImagePlugin.APNG_BLEND_OP_OVER |
PngImagePlugin.Blend.OP_OVER |
The file
argument in :py~PIL.ImageShow.Viewer.show_file()
has been deprecated and will be removed in Pillow 10.0.0 (2023-07-01). It has been replaced by path
.
In effect, viewer.show_file("test.jpg")
will continue to work unchanged. viewer.show_file(file="test.jpg")
will raise a deprecation warning, and suggest viewer.show_file(path="test.jpg")
instead.
9.1.0
The stub image plugin FitsStubImagePlugin
has been deprecated and will be removed in Pillow 10.0.0 (2023-07-01). FITS images can be read without a handler through ~PIL.FitsImagePlugin
instead.
:py~PIL.TiffImagePlugin.TiffImageFile.get_photoshop_blocks
has been added, to allow users to determine what Photoshop "Image Resource Blocks" are contained within an image. The keys of the returned dictionary are the image resource IDs.
At present, the information within each block is merely returned as a dictionary with a "data" entry. This will allow more useful information to be added in the future without breaking backwards compatibility.
The :pyPIL.Image.Image.save
method now supports the following options for JPEG 2000:
- mct
If
1
then enable multiple component transformation when encoding, otherwise use0
for no component transformation (default). If MCT is enabled andirreversible
isTrue
then the Irreversible Color Transformation will be applied, otherwise encoding will use the Reversible Color Transformation. MCT works best with amode
ofRGB
and is only applicable when the image data has 3 components.- no_jp2
If
True
then don't wrap the raw codestream in the JP2 file format when saving, otherwise the extension of the filename will be used to determine the format (default).
:py~PIL.ImageFile.PyEncoder
has been added, allowing for file encoders to be written in Python. See Writing Your Own File Codec in Python<file-codecs-py>
for more information.
Pillow 9.0.0 introduced the conversion of subsequent GIF frames to RGB
or RGBA
. This behaviour can now be changed so that the first P
frame is converted to RGB
as well.
from PIL import GifImagePlugin
GifImagePlugin.LOADING_STRATEGY = GifImagePlugin.LoadingStrategy.RGB_ALWAYS
Or subsequent frames can be kept in P
mode as long as there is only a single palette.
from PIL import GifImagePlugin
GifImagePlugin.LOADING_STRATEGY = GifImagePlugin.LoadingStrategy.RGB_AFTER_DIFFERENT_PALETTE_ONLY
Pillow now builds binary wheels for musllinux, suitable for Linux distributions based on the musl C standard library such as Alpine (rather than the glibc library used by manylinux wheels). See 656
.
When calling :py~PIL.Image.Image.show
or using :py~PIL.ImageShow
, a temporary file is created from the image. On Unix, Pillow will no longer delete these files, and instead leave it to the operating system to do so.
Image._repr_pretty
im._repr_pretty_
has been added to provide a representation of an image without the identity of the object. This allows Jupyter to describe an image and have that description stay the same on subsequent executions of the same code.
Support has been added for reading BigTIFF images.
Support has been added for saving BLP images. blp_version
can be used to specify whether the image should be saved as BLP1 or BLP2, e.g. im.save("out.blp", blp_version="BLP1")
. By default, BLP2 will be used.