Various Pillow utilities. These are monkey-patched on, allowing continued use of the Image.new factory function. Most have only been tested in RGB/RGBA modes and may not work in other modes.
Required: pillow, toolz, utils.
Optional: pyphen (for text hyphenation), requests (for HTTP request headers).
Padding: a class representing box padding, initialized from one, two or four integers. Any function below that has a padding parameter can be passed any of these.
>> Padding(10)
Padding(l=10, u=10, r=10, d=10)
>> Padding((10,0))
Padding(l=10, u=0, r=10, d=0)
>> Padding((10,0,0,5))
Padding(l=10, u=0, r=0, d=5)Alignment: a class representing element alignment, initialized from one or two floats between 0 and 1. Any function below that has an element alignment parameter can be passed one of these.
>> Alignment(0.5)
Alignment(x=50%, y=50%)
>> Alignment((0,1))
Alignment(x=0%, y=100%)BoundingBox: a class representing a bounding box, initialized from the 4 coordinates or from a collection of points, with optional padding. Not really used at the moment.
>> BoundingBox([(10,10), (15,8), (12,15)])
Box(l=10, u=8, r=15, d=15)
>> _.width
25
>> BoundingBox((5,5,10,10), padding=2)
Box(l=3, u=3, r=12, d=12)font: shorthand function for generating a truetype font object with standard variant naming (e.g. arialbd for bold). Also, arial is defined explicitly for the Arial family.
>> font("times", 24, italics=True)
<PIL.ImageFont.FreeTypeFont at 0x48e9ad0>
>> [_.font.family, _.font.style, _.size]
['Times New Roman', 'Italic', 24]
>> arial(24, bold=True)
<PIL.ImageFont.FreeTypeFont at 0x4a335f0>ImageColor.getrgba: converts a named color or RGB triple to an RGBA named tuple.
>> ImageColor.getrgba("yellow")
RGBA(red=255, green=255, blue=0, alpha=255)
>> ImageColor.getrgba("yellow")._replace(alpha=0)
RGBA(red=255, green=255, blue=0, alpha=0)
>> ImageColor.getrgba("#fafafa")
RGBA(red=250, green=250, blue=250, alpha=255)
>> ImageColor.getrgba((100,50,50))
RGBA(red=100, green=50, blue=50, alpha=255)
>> ImageColor.getrgba((100,50,50,5))
RGBA(red=100, green=50, blue=50, alpha=5)ImageDraw.textsize: same as ImageDraw.Draw.textsize but doesn't require a drawable object. Similarly, ImageDraw.multiline_textsize.
>> ImageDraw.textsize("The rain", arial(24))
(88, 22)Image.from_url: create an image from a URL. Image.from_url_with_cache does the same but uses a file cache (with a default filename structure of cache_directory/hostname/hash_of_path.extension).
>> Image.from_url('https://assets-cdn.github.com/images/modules/logos_page/Octocat.png')
[10:05:03] pillar:DEBUG - Reading image from https://assets-cdn.github.com/images/modules/logos_page/Octocat.png
>> Image.from_url_with_cache('https://assets-cdn.github.com/images/modules/logos_page/Octocat.png')
[10:05:27] pillar:DEBUG - Reading image from https://assets-cdn.github.com/images/modules/logos_page/Octocat.png
[10:05:28] pillar:DEBUG - Saving cached image to cache\assets-cdn.github.com\d4b25973fbbe302e1b486000f43aa25b9c61f9bf.png
>> Image.from_url_with_cache('https://assets-cdn.github.com/images/modules/logos_page/Octocat.png')
[10:05:38] pillar:DEBUG - Loading cached image at cache\assets-cdn.github.com\d4b25973fbbe302e1b486000f43aa25b9c61f9bf.pngImage.from_text: creates an image from text and font. Supports word-wrapping, alignment, padding and hyphenation. Wrapping can use an arbitrary tokenizer (it uses whitespace splitting by default) and hyphenation requires a hyphenator (the phyphen module is a good source of these).
>> Image.from_text("The rain", arial(16, bold=True), fg="white", bg="#1f5774", padding=10).show()
>> Image.from_text("dedicated to the proposition that all men are created equal", arial(16),
fg="white", bg="black", max_width=100, align="center", padding=5).show()
>> Image.from_text("dedicated to the proposition that all men are created equal", arial(16),
fg="white", bg="black", max_width=100, align="left", padding=5,
hyphenator=language_hyphenator("en_EN")).show()
Image.from_array: create an image from an array of images. Similarly, Image.from_row and Image.from_column create images form a list of images.
>> Image.from_array([[Image.from_text("{}×{}={}".format(x,y,x*y), arial(16)) for x in (2,4,8)] for y in range(5)],
padding=(5,2), xalign=(0,0.5,1), bg="#1f5774").show()
Image.from_pattern: create an image from a background pattern, either scaled or tiled. Similarly, Image.from_vertical_pattern and Image.from_horiztonal_pattern automatically scale to the image width or height.
>> flag = Image.from_url("https://upload.wikimedia.org/wikipedia/en/thumb/a/ae/Flag_of_the_United_Kingdom.svg/800px-Flag_of_the_United_Kingdom.svg.png")
>> Image.from_pattern(flag.resize_fixed_aspect(width=50), (150,150)).show()
>> Image.from_vertical_pattern(flag, (150,150)).show()
>> Image.from_horizontal_pattern(flag, (150,150), align=0.5).show()
Image.EMPTY_IMAGE: an empty "RGBA" image.
>> Image.EMPTY_IMAGE
<PIL.Image.Image image mode=RGBA size=0x0 at 0x497C310>Image.Image.overlay: like Image.Image.paste, but by default uses the pasted image's alpha channel as a mask and supports a copy parameter.
>> base = Image.new("RGB", (100,60), "blue")
>> base.overlay(Image.from_text("red", arial(24)), (0,0), copy=True).show()
Image.Image.place: overlay an image at the given alignment and padding.
>> img = Image.from_text("red", arial(24), bg="grey")
>> base.place(img).show()
>> base.place(img, align=(0,1), padding=5).show()
Image.Image.pin: pin an image on another, expanding the base image if necessary.
>> base.pin(img, (0,30), bg="black").show()
>> base.pin(img, (0,30), align=(1,0.5), bg="black").show()
Image.Image.pad: pad an image.
>> base.pad(5, "grey").show()
>> base.pad((10,10,0,0), "grey").show()
Image.Image.resize: monkey-patched to handle zero-width/height sizes.
>> flag.resize((0,100))
<PIL.Image.Image image mode=RGBA size=0x100 at 0x4DEDEF0>Image.Image.resize_fixed_aspect: resize an image, preserving its aspect ratio.
>> flag.resize_fixed_aspect(scale=0.25).show()
>> smallflag = flag.resize_fixed_aspect(width=100)
>> smallflag.show()
Image.Image.pad_to_aspect: pad an image so that it has the given aspect ratio.
>> smallflag.pad_to_aspect(1, bg="grey").show()
>> smallflag.pad_to_aspect(800, 600, align=0).show()
Image.Image.crop_to_aspect: crop an image so that it has the given aspect ratio.
>> smallflag.crop_to_aspect(1).show()
>> smallflag.crop_to_aspect(800, 600, align=0).show()
Image.Image.replace_color: return an image with one color replaced by another. Requires numpy.
>> base = Image.new("RGB", (80,40), "blue").pad(5)
>> img = Image.from_text("red", arial(24), "blue")
>> base.replace_color("blue", "grey").place(img.replace_color("blue", "red", ignore_alpha=True)).show()
Image.Image.select_color: return an image mask selection of a color. Requires numpy.
>> pattern = Image.from_pattern(flag.resize_fixed_aspect(width=30), base.size)
>> mask = base.select_color("blue")
>> base.place(pattern, mask=mask).show()