@alexkirsz/gatsby-plugin-sharp
v2.0.0-rc.1
Published
Wrapper of the Sharp image manipulation library for Gatsby plugins
Downloads
6
Maintainers
Readme
gatsby-plugin-sharp
Exposes several image processing functions built on the Sharp image processing library. This is a low-level helper plugin generally used by other Gatsby plugins. You generally shouldn't be using this directly but might find it helpful if doing very custom image processing.
It aims to provide excellent out-of-the box settings for processing common web image formats.
For JPEGs it generates progressive images with a default quality level of 50.
For PNGs it uses pngquant to compress images. By default it uses a quality setting of [50-75].
Install
npm install --save gatsby-plugin-sharp
How to use
// In your gatsby-config.js
plugins: [`gatsby-plugin-sharp`]
Methods
resize
Parameters
width
(int, default: 400)height
(int)quality
(int, default: 50)jpegProgressive
(bool, default: true)pngCompressionLevel
(int, default: 9)base64
(bool, default: false)
Returns
src
(string)width
(int)height
(int)aspectRatio
(float)
fixed
Automatically create sizes for different resolutions — we do 1x, 1.5x, 2x, and 3x.
Parameters
width
(int, default: 400)height
(int)quality
(int, default: 50)
Returns
base64
(string)aspectRatio
(float)width
(float)height
(float)src
(string)srcSet
(string)
fluid
Create fluid sizes (in width) for the image. If the max width of the container for the rendered markdown file is 800px, the sizes would then be: 200, 400, 800, 1200, 1600, 2400 – enough to provide close to the optimal image size for every device size / screen resolution.
On top of that, fluid
returns everything else (namely aspectRatio and
a base64 image to use as a placeholder) you need to implement the "blur up"
technique popularized by Medium and Facebook (and also available as a Gatsby
plugin for Markdown content as gatsby-remark-images).
Parameters
maxWidth
(int, default: 800)maxHeight
(int)quality
(int, default: 50)sizeByPixelDensity
(bool, default: false)
Returns
base64
(string)aspectRatio
(float)src
(string)srcSet
(string)srcSetType
(string)sizes
(string)originalImg
(string)
Shared Options
In addition to their individual parameters, all methods above share the following:
grayscale
(bool, default: false)duotone
(bool|obj, default: false)toFormat
(string, default: '')cropFocus
(string, default: 'sharp.strategy.attention')
toFormat
Convert the source image to one of the following available options: NO_CHANGE
,
JPG
, PNG
, WEBP
.
cropFocus
Change the cropping focus. Available options: CENTER
, NORTH
, NORTHEAST
,
EAST
, SOUTHEAST
, SOUTH
, SOUTHWEST
, WEST
, NORTHWEST
, ENTROPY
,
ATTENTION
. See Sharp's crop.
rotate
Rotate the image (after cropping). See Sharp's rotate.
grayscale
Uses Sharp's greyscale to convert the source image to 8-bit greyscale, 256 shades of grey, e.g.
allImageSharp {
edges {
node {
... on ImageSharp {
resize(width: 150, height: 150, grayscale: true) {
src
}
}
}
}
}
duotone
Applys a "duotone" effect (see I, II, III) to the source image if
given two hex colors shadow
and highlight
defining start and end color of
the duotone gradient, e.g.
fixed(
width: 800,
duotone: {
highlight: "#f00e2e",
shadow: "#192550"
}
) {
src
srcSet
base64
}
the source image colors will be converted to match a gradient color chosen based
on each pixel's relative luminance.
Logic is borrowed from react-duotone.
You can pass a third optional parameter, opacity
:
fluid(
width: 800,
duotone: {
highlight: "#f00e2e",
shadow: "#192550",
opacity: 50
}
) {
src
srcSet
base64
}
If set, a semi-transparent version of duotone'd image will be composited over the original image, allowing the original image and its colors to partially "shine through". Heads up: If the original image contains an alpha channel it will be flattened before creating the composite.
This works by adding an alpha channel to the duotone'd image - then we let Sharp
do its magic via
overlayWith
;
quoting the Sharp documentation:
If the overlay image contains an alpha channel then composition with premultiplication will occur.
tracedSVG
Generates a traced SVG of the image (see the original GitHub issue) and
returns the SVG as "optimized URL-encoded" data:
URI. It it used in
gatsby-image to provide an
alternative to the default inline base64 placeholder image.
Uses node-potrace and SVGO under the hood. Default settings for node-potrace:
{
color: `lightgray`,
optTolerance: 0.4,
turdSize: 100,
turnPolicy: TURNPOLICY_MAJORITY,
}
All node-potrace Potrace
parameters are exposed and can be set via the
traceSVG
argument:
fixed(
traceSVG: {
color: "#f00e2e"
turnPolicy: TURNPOLICY_MINORITY
blackOnWhite: false
}
) {
src
srcSet
tracedSVG
}