Docs: migrate to Astro Starlight

This commit is contained in:
Lovell Fuller
2025-01-16 12:55:45 +00:00
parent 14c83e1f4c
commit eeac8d4656
39 changed files with 279 additions and 505 deletions

View File

@@ -0,0 +1,5 @@
<svg xmlns="http://www.w3.org/2000/svg" viewBox="86 86 550 550">
<!-- Creative Commons CC0 1.0 Universal Public Domain Dedication -->
<path fill="none" stroke="#9c0" stroke-width="80" d="M258.411 285.777l200.176-26.8M244.113 466.413L451.44 438.66M451.441 438.66V238.484M451.441 88.363v171.572l178.725-23.917M270.323 255.602V477.22M272.71 634.17V462.591L93.984 486.515"/>
<path fill="none" stroke="#090" stroke-width="80" d="M451.441 610.246V438.66l178.725-23.91M269.688 112.59v171.58L90.964 308.093"/>
</svg>

After

Width:  |  Height:  |  Size: 508 B

View File

@@ -0,0 +1,7 @@
import { defineCollection } from 'astro:content';
import { docsLoader } from '@astrojs/starlight/loaders';
import { docsSchema } from '@astrojs/starlight/schema';
export const collections = {
docs: defineCollection({ loader: docsLoader(), schema: docsSchema() }),
};

View File

@@ -0,0 +1,140 @@
---
title: Channel manipulation
---
## removeAlpha
> removeAlpha() ⇒ <code>Sharp</code>
Remove alpha channel, if any. This is a no-op if the image does not have an alpha channel.
See also [flatten](/api-operation#flatten).
**Example**
```js
sharp('rgba.png')
.removeAlpha()
.toFile('rgb.png', function(err, info) {
// rgb.png is a 3 channel image without an alpha channel
});
```
## ensureAlpha
> ensureAlpha([alpha]) ⇒ <code>Sharp</code>
Ensure the output image has an alpha transparency channel.
If missing, the added alpha channel will have the specified
transparency level, defaulting to fully-opaque (1).
This is a no-op if the image already has an alpha channel.
**Throws**:
- <code>Error</code> Invalid alpha transparency level
**Since**: 0.21.2
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [alpha] | <code>number</code> | <code>1</code> | alpha transparency level (0=fully-transparent, 1=fully-opaque) |
**Example**
```js
// rgba.png will be a 4 channel image with a fully-opaque alpha channel
await sharp('rgb.jpg')
.ensureAlpha()
.toFile('rgba.png')
```
**Example**
```js
// rgba is a 4 channel image with a fully-transparent alpha channel
const rgba = await sharp(rgb)
.ensureAlpha(0)
.toBuffer();
```
## extractChannel
> extractChannel(channel) ⇒ <code>Sharp</code>
Extract a single channel from a multi-channel image.
**Throws**:
- <code>Error</code> Invalid channel
| Param | Type | Description |
| --- | --- | --- |
| channel | <code>number</code> \| <code>string</code> | zero-indexed channel/band number to extract, or `red`, `green`, `blue` or `alpha`. |
**Example**
```js
// green.jpg is a greyscale image containing the green channel of the input
await sharp(input)
.extractChannel('green')
.toFile('green.jpg');
```
**Example**
```js
// red1 is the red value of the first pixel, red2 the second pixel etc.
const [red1, red2, ...] = await sharp(input)
.extractChannel(0)
.raw()
.toBuffer();
```
## joinChannel
> joinChannel(images, options) ⇒ <code>Sharp</code>
Join one or more channels to the image.
The meaning of the added channels depends on the output colourspace, set with `toColourspace()`.
By default the output image will be web-friendly sRGB, with additional channels interpreted as alpha channels.
Channel ordering follows vips convention:
- sRGB: 0: Red, 1: Green, 2: Blue, 3: Alpha.
- CMYK: 0: Magenta, 1: Cyan, 2: Yellow, 3: Black, 4: Alpha.
Buffers may be any of the image formats supported by sharp.
For raw pixel input, the `options` object should contain a `raw` attribute, which follows the format of the attribute of the same name in the `sharp()` constructor.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Description |
| --- | --- | --- |
| images | <code>Array.&lt;(string\|Buffer)&gt;</code> \| <code>string</code> \| <code>Buffer</code> | one or more images (file paths, Buffers). |
| options | <code>Object</code> | image options, see `sharp()` constructor. |
## bandbool
> bandbool(boolOp) ⇒ <code>Sharp</code>
Perform a bitwise boolean operation on all input image channels (bands) to produce a single channel output image.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Description |
| --- | --- | --- |
| boolOp | <code>string</code> | one of `and`, `or` or `eor` to perform that bitwise operation, like the C logic operators `&`, `|` and `^` respectively. |
**Example**
```js
sharp('3-channel-rgb-input.png')
.bandbool(sharp.bool.and)
.toFile('1-channel-output.png', function (err, info) {
// The output will be a single channel image where each pixel `P = R & G & B`.
// If `I(1,1) = [247, 170, 14] = [0b11110111, 0b10101010, 0b00001111]`
// then `O(1,1) = 0b11110111 & 0b10101010 & 0b00001111 = 0b00000010 = 2`.
});
```

View File

@@ -0,0 +1,149 @@
---
title: Colour manipulation
---
## tint
> tint(tint) ⇒ <code>Sharp</code>
Tint the image using the provided colour.
An alpha channel may be present and will be unchanged by the operation.
**Throws**:
- <code>Error</code> Invalid parameter
| Param | Type | Description |
| --- | --- | --- |
| tint | <code>string</code> \| <code>Object</code> | Parsed by the [color](https://www.npmjs.org/package/color) module. |
**Example**
```js
const output = await sharp(input)
.tint({ r: 255, g: 240, b: 16 })
.toBuffer();
```
## greyscale
> greyscale([greyscale]) ⇒ <code>Sharp</code>
Convert to 8-bit greyscale; 256 shades of grey.
This is a linear operation. If the input image is in a non-linear colour space such as sRGB, use `gamma()` with `greyscale()` for the best results.
By default the output image will be web-friendly sRGB and contain three (identical) colour channels.
This may be overridden by other sharp operations such as `toColourspace('b-w')`,
which will produce an output image containing one colour channel.
An alpha channel may be present, and will be unchanged by the operation.
| Param | Type | Default |
| --- | --- | --- |
| [greyscale] | <code>Boolean</code> | <code>true</code> |
**Example**
```js
const output = await sharp(input).greyscale().toBuffer();
```
## grayscale
> grayscale([grayscale]) ⇒ <code>Sharp</code>
Alternative spelling of `greyscale`.
| Param | Type | Default |
| --- | --- | --- |
| [grayscale] | <code>Boolean</code> | <code>true</code> |
## pipelineColourspace
> pipelineColourspace([colourspace]) ⇒ <code>Sharp</code>
Set the pipeline colourspace.
The input image will be converted to the provided colourspace at the start of the pipeline.
All operations will use this colourspace before converting to the output colourspace,
as defined by [toColourspace](#tocolourspace).
**Throws**:
- <code>Error</code> Invalid parameters
**Since**: 0.29.0
| Param | Type | Description |
| --- | --- | --- |
| [colourspace] | <code>string</code> | pipeline colourspace e.g. `rgb16`, `scrgb`, `lab`, `grey16` [...](https://github.com/libvips/libvips/blob/41cff4e9d0838498487a00623462204eb10ee5b8/libvips/iofuncs/enumtypes.c#L774) |
**Example**
```js
// Run pipeline in 16 bits per channel RGB while converting final result to 8 bits per channel sRGB.
await sharp(input)
.pipelineColourspace('rgb16')
.toColourspace('srgb')
.toFile('16bpc-pipeline-to-8bpc-output.png')
```
## pipelineColorspace
> pipelineColorspace([colorspace]) ⇒ <code>Sharp</code>
Alternative spelling of `pipelineColourspace`.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Description |
| --- | --- | --- |
| [colorspace] | <code>string</code> | pipeline colorspace. |
## toColourspace
> toColourspace([colourspace]) ⇒ <code>Sharp</code>
Set the output colourspace.
By default output image will be web-friendly sRGB, with additional channels interpreted as alpha channels.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Description |
| --- | --- | --- |
| [colourspace] | <code>string</code> | output colourspace e.g. `srgb`, `rgb`, `cmyk`, `lab`, `b-w` [...](https://github.com/libvips/libvips/blob/3c0bfdf74ce1dc37a6429bed47fa76f16e2cd70a/libvips/iofuncs/enumtypes.c#L777-L794) |
**Example**
```js
// Output 16 bits per pixel RGB
await sharp(input)
.toColourspace('rgb16')
.toFile('16-bpp.png')
```
## toColorspace
> toColorspace([colorspace]) ⇒ <code>Sharp</code>
Alternative spelling of `toColourspace`.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Description |
| --- | --- | --- |
| [colorspace] | <code>string</code> | output colorspace. |

View File

@@ -0,0 +1,102 @@
---
title: Compositing images
---
## composite
> composite(images) ⇒ <code>Sharp</code>
Composite image(s) over the processed (resized, extracted etc.) image.
The images to composite must be the same size or smaller than the processed image.
If both `top` and `left` options are provided, they take precedence over `gravity`.
Other operations in the same processing pipeline (e.g. resize, rotate, flip,
flop, extract) will always be applied to the input image before composition.
The `blend` option can be one of `clear`, `source`, `over`, `in`, `out`, `atop`,
`dest`, `dest-over`, `dest-in`, `dest-out`, `dest-atop`,
`xor`, `add`, `saturate`, `multiply`, `screen`, `overlay`, `darken`, `lighten`,
`colour-dodge`, `color-dodge`, `colour-burn`,`color-burn`,
`hard-light`, `soft-light`, `difference`, `exclusion`.
More information about blend modes can be found at
https://www.libvips.org/API/current/libvips-conversion.html#VipsBlendMode
and https://www.cairographics.org/operators/
**Throws**:
- <code>Error</code> Invalid parameters
**Since**: 0.22.0
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| images | <code>Array.&lt;Object&gt;</code> | | Ordered list of images to composite |
| [images[].input] | <code>Buffer</code> \| <code>String</code> | | Buffer containing image data, String containing the path to an image file, or Create object (see below) |
| [images[].input.create] | <code>Object</code> | | describes a blank overlay to be created. |
| [images[].input.create.width] | <code>Number</code> | | |
| [images[].input.create.height] | <code>Number</code> | | |
| [images[].input.create.channels] | <code>Number</code> | | 3-4 |
| [images[].input.create.background] | <code>String</code> \| <code>Object</code> | | parsed by the [color](https://www.npmjs.org/package/color) module to extract values for red, green, blue and alpha. |
| [images[].input.text] | <code>Object</code> | | describes a new text image to be created. |
| [images[].input.text.text] | <code>string</code> | | text to render as a UTF-8 string. It can contain Pango markup, for example `<i>Le</i>Monde`. |
| [images[].input.text.font] | <code>string</code> | | font name to render with. |
| [images[].input.text.fontfile] | <code>string</code> | | absolute filesystem path to a font file that can be used by `font`. |
| [images[].input.text.width] | <code>number</code> | <code>0</code> | integral number of pixels to word-wrap at. Lines of text wider than this will be broken at word boundaries. |
| [images[].input.text.height] | <code>number</code> | <code>0</code> | integral number of pixels high. When defined, `dpi` will be ignored and the text will automatically fit the pixel resolution defined by `width` and `height`. Will be ignored if `width` is not specified or set to 0. |
| [images[].input.text.align] | <code>string</code> | <code>&quot;&#x27;left&#x27;&quot;</code> | text alignment (`'left'`, `'centre'`, `'center'`, `'right'`). |
| [images[].input.text.justify] | <code>boolean</code> | <code>false</code> | set this to true to apply justification to the text. |
| [images[].input.text.dpi] | <code>number</code> | <code>72</code> | the resolution (size) at which to render the text. Does not take effect if `height` is specified. |
| [images[].input.text.rgba] | <code>boolean</code> | <code>false</code> | set this to true to enable RGBA output. This is useful for colour emoji rendering, or support for Pango markup features like `<span foreground="red">Red!</span>`. |
| [images[].input.text.spacing] | <code>number</code> | <code>0</code> | text line height in points. Will use the font line height if none is specified. |
| [images[].autoOrient] | <code>Boolean</code> | <code>false</code> | set to true to use EXIF orientation data, if present, to orient the image. |
| [images[].blend] | <code>String</code> | <code>&#x27;over&#x27;</code> | how to blend this image with the image below. |
| [images[].gravity] | <code>String</code> | <code>&#x27;centre&#x27;</code> | gravity at which to place the overlay. |
| [images[].top] | <code>Number</code> | | the pixel offset from the top edge. |
| [images[].left] | <code>Number</code> | | the pixel offset from the left edge. |
| [images[].tile] | <code>Boolean</code> | <code>false</code> | set to true to repeat the overlay image across the entire image with the given `gravity`. |
| [images[].premultiplied] | <code>Boolean</code> | <code>false</code> | set to true to avoid premultiplying the image below. Equivalent to the `--premultiplied` vips option. |
| [images[].density] | <code>Number</code> | <code>72</code> | number representing the DPI for vector overlay image. |
| [images[].raw] | <code>Object</code> | | describes overlay when using raw pixel data. |
| [images[].raw.width] | <code>Number</code> | | |
| [images[].raw.height] | <code>Number</code> | | |
| [images[].raw.channels] | <code>Number</code> | | |
| [images[].animated] | <code>boolean</code> | <code>false</code> | Set to `true` to read all frames/pages of an animated image. |
| [images[].failOn] | <code>string</code> | <code>&quot;&#x27;warning&#x27;&quot;</code> | @see [constructor parameters](/api-constructor#parameters) |
| [images[].limitInputPixels] | <code>number</code> \| <code>boolean</code> | <code>268402689</code> | @see [constructor parameters](/api-constructor#parameters) |
**Example**
```js
await sharp(background)
.composite([
{ input: layer1, gravity: 'northwest' },
{ input: layer2, gravity: 'southeast' },
])
.toFile('combined.png');
```
**Example**
```js
const output = await sharp('input.gif', { animated: true })
.composite([
{ input: 'overlay.png', tile: true, blend: 'saturate' }
])
.toBuffer();
```
**Example**
```js
sharp('input.png')
.rotate(180)
.resize(300)
.flatten( { background: '#ff6600' } )
.composite([{ input: 'overlay.png', gravity: 'southeast' }])
.sharpen()
.withMetadata()
.webp( { quality: 90 } )
.toBuffer()
.then(function(outputBuffer) {
// outputBuffer contains upside down, 300px wide, alpha channel flattened
// onto orange background, composited with overlay.png with SE gravity,
// sharpened, with metadata, 90% quality WebP image data. Phew!
});
```

View File

@@ -0,0 +1,240 @@
---
title: Constructor
---
## Sharp
> Sharp
**Emits**: <code>Sharp#event:info</code>, <code>Sharp#event:warning</code>
<a name="new_Sharp_new"></a>
### new
> new Sharp([input], [options])
Constructor factory to create an instance of `sharp`, to which further methods are chained.
JPEG, PNG, WebP, GIF, AVIF or TIFF format image data can be streamed out from this object.
When using Stream based output, derived attributes are available from the `info` event.
Non-critical problems encountered during processing are emitted as `warning` events.
Implements the [stream.Duplex](http://nodejs.org/api/stream.html#stream_class_stream_duplex) class.
When loading more than one page/frame of an animated image,
these are combined as a vertically-stacked "toilet roll" image
where the overall height is the `pageHeight` multiplied by the number of `pages`.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [input] | <code>Buffer</code> \| <code>ArrayBuffer</code> \| <code>Uint8Array</code> \| <code>Uint8ClampedArray</code> \| <code>Int8Array</code> \| <code>Uint16Array</code> \| <code>Int16Array</code> \| <code>Uint32Array</code> \| <code>Int32Array</code> \| <code>Float32Array</code> \| <code>Float64Array</code> \| <code>string</code> | | if present, can be a Buffer / ArrayBuffer / Uint8Array / Uint8ClampedArray containing JPEG, PNG, WebP, AVIF, GIF, SVG or TIFF image data, or a TypedArray containing raw pixel image data, or a String containing the filesystem path to an JPEG, PNG, WebP, AVIF, GIF, SVG or TIFF image file. JPEG, PNG, WebP, AVIF, GIF, SVG, TIFF or raw pixel image data can be streamed into the object when not present. |
| [options] | <code>Object</code> | | if present, is an Object with optional attributes. |
| [options.failOn] | <code>string</code> | <code>&quot;&#x27;warning&#x27;&quot;</code> | When to abort processing of invalid pixel data, one of (in order of sensitivity, least to most): 'none', 'truncated', 'error', 'warning'. Higher levels imply lower levels. Invalid metadata will always abort. |
| [options.limitInputPixels] | <code>number</code> \| <code>boolean</code> | <code>268402689</code> | Do not process input images where the number of pixels (width x height) exceeds this limit. Assumes image dimensions contained in the input metadata can be trusted. An integral Number of pixels, zero or false to remove limit, true to use default limit of 268402689 (0x3FFF x 0x3FFF). |
| [options.unlimited] | <code>boolean</code> | <code>false</code> | Set this to `true` to remove safety features that help prevent memory exhaustion (JPEG, PNG, SVG, HEIF). |
| [options.autoOrient] | <code>boolean</code> | <code>false</code> | Set this to `true` to rotate/flip the image to match EXIF `Orientation`, if any. |
| [options.sequentialRead] | <code>boolean</code> | <code>true</code> | Set this to `false` to use random access rather than sequential read. Some operations will do this automatically. |
| [options.density] | <code>number</code> | <code>72</code> | number representing the DPI for vector images in the range 1 to 100000. |
| [options.ignoreIcc] | <code>number</code> | <code>false</code> | should the embedded ICC profile, if any, be ignored. |
| [options.pages] | <code>number</code> | <code>1</code> | Number of pages to extract for multi-page input (GIF, WebP, TIFF), use -1 for all pages. |
| [options.page] | <code>number</code> | <code>0</code> | Page number to start extracting from for multi-page input (GIF, WebP, TIFF), zero based. |
| [options.subifd] | <code>number</code> | <code>-1</code> | subIFD (Sub Image File Directory) to extract for OME-TIFF, defaults to main image. |
| [options.level] | <code>number</code> | <code>0</code> | level to extract from a multi-level input (OpenSlide), zero based. |
| [options.pdfBackground] | <code>string</code> \| <code>Object</code> | | Background colour to use when PDF is partially transparent. Parsed by the [color](https://www.npmjs.org/package/color) module to extract values for red, green, blue and alpha. Requires the use of a globally-installed libvips compiled with support for PDFium, Poppler, ImageMagick or GraphicsMagick. |
| [options.animated] | <code>boolean</code> | <code>false</code> | Set to `true` to read all frames/pages of an animated image (GIF, WebP, TIFF), equivalent of setting `pages` to `-1`. |
| [options.raw] | <code>Object</code> | | describes raw pixel input image data. See `raw()` for pixel ordering. |
| [options.raw.width] | <code>number</code> | | integral number of pixels wide. |
| [options.raw.height] | <code>number</code> | | integral number of pixels high. |
| [options.raw.channels] | <code>number</code> | | integral number of channels, between 1 and 4. |
| [options.raw.premultiplied] | <code>boolean</code> | | specifies that the raw input has already been premultiplied, set to `true` to avoid sharp premultiplying the image. (optional, default `false`) |
| [options.create] | <code>Object</code> | | describes a new image to be created. |
| [options.create.width] | <code>number</code> | | integral number of pixels wide. |
| [options.create.height] | <code>number</code> | | integral number of pixels high. |
| [options.create.channels] | <code>number</code> | | integral number of channels, either 3 (RGB) or 4 (RGBA). |
| [options.create.background] | <code>string</code> \| <code>Object</code> | | parsed by the [color](https://www.npmjs.org/package/color) module to extract values for red, green, blue and alpha. |
| [options.create.noise] | <code>Object</code> | | describes a noise to be created. |
| [options.create.noise.type] | <code>string</code> | | type of generated noise, currently only `gaussian` is supported. |
| [options.create.noise.mean] | <code>number</code> | | mean of pixels in generated noise. |
| [options.create.noise.sigma] | <code>number</code> | | standard deviation of pixels in generated noise. |
| [options.text] | <code>Object</code> | | describes a new text image to be created. |
| [options.text.text] | <code>string</code> | | text to render as a UTF-8 string. It can contain Pango markup, for example `<i>Le</i>Monde`. |
| [options.text.font] | <code>string</code> | | font name to render with. |
| [options.text.fontfile] | <code>string</code> | | absolute filesystem path to a font file that can be used by `font`. |
| [options.text.width] | <code>number</code> | <code>0</code> | Integral number of pixels to word-wrap at. Lines of text wider than this will be broken at word boundaries. |
| [options.text.height] | <code>number</code> | <code>0</code> | Maximum integral number of pixels high. When defined, `dpi` will be ignored and the text will automatically fit the pixel resolution defined by `width` and `height`. Will be ignored if `width` is not specified or set to 0. |
| [options.text.align] | <code>string</code> | <code>&quot;&#x27;left&#x27;&quot;</code> | Alignment style for multi-line text (`'left'`, `'centre'`, `'center'`, `'right'`). |
| [options.text.justify] | <code>boolean</code> | <code>false</code> | set this to true to apply justification to the text. |
| [options.text.dpi] | <code>number</code> | <code>72</code> | the resolution (size) at which to render the text. Does not take effect if `height` is specified. |
| [options.text.rgba] | <code>boolean</code> | <code>false</code> | set this to true to enable RGBA output. This is useful for colour emoji rendering, or support for pango markup features like `<span foreground="red">Red!</span>`. |
| [options.text.spacing] | <code>number</code> | <code>0</code> | text line height in points. Will use the font line height if none is specified. |
| [options.text.wrap] | <code>string</code> | <code>&quot;&#x27;word&#x27;&quot;</code> | word wrapping style when width is provided, one of: 'word', 'char', 'word-char' (prefer word, fallback to char) or 'none'. |
**Example**
```js
sharp('input.jpg')
.resize(300, 200)
.toFile('output.jpg', function(err) {
// output.jpg is a 300 pixels wide and 200 pixels high image
// containing a scaled and cropped version of input.jpg
});
```
**Example**
```js
// Read image data from remote URL,
// resize to 300 pixels wide,
// emit an 'info' event with calculated dimensions
// and finally write image data to writableStream
const { body } = fetch('https://...');
const readableStream = Readable.fromWeb(body);
const transformer = sharp()
.resize(300)
.on('info', ({ height }) => {
console.log(`Image height is ${height}`);
});
readableStream.pipe(transformer).pipe(writableStream);
```
**Example**
```js
// Create a blank 300x200 PNG image of semi-translucent red pixels
sharp({
create: {
width: 300,
height: 200,
channels: 4,
background: { r: 255, g: 0, b: 0, alpha: 0.5 }
}
})
.png()
.toBuffer()
.then( ... );
```
**Example**
```js
// Convert an animated GIF to an animated WebP
await sharp('in.gif', { animated: true }).toFile('out.webp');
```
**Example**
```js
// Read a raw array of pixels and save it to a png
const input = Uint8Array.from([255, 255, 255, 0, 0, 0]); // or Uint8ClampedArray
const image = sharp(input, {
// because the input does not contain its dimensions or how many channels it has
// we need to specify it in the constructor options
raw: {
width: 2,
height: 1,
channels: 3
}
});
await image.toFile('my-two-pixels.png');
```
**Example**
```js
// Generate RGB Gaussian noise
await sharp({
create: {
width: 300,
height: 200,
channels: 3,
noise: {
type: 'gaussian',
mean: 128,
sigma: 30
}
}
}).toFile('noise.png');
```
**Example**
```js
// Generate an image from text
await sharp({
text: {
text: 'Hello, world!',
width: 400, // max width
height: 300 // max height
}
}).toFile('text_bw.png');
```
**Example**
```js
// Generate an rgba image from text using pango markup and font
await sharp({
text: {
text: '<span foreground="red">Red!</span><span background="cyan">blue</span>',
font: 'sans',
rgba: true,
dpi: 300
}
}).toFile('text_rgba.png');
```
## clone
> clone() ⇒ [<code>Sharp</code>](#Sharp)
Take a "snapshot" of the Sharp instance, returning a new instance.
Cloned instances inherit the input of their parent instance.
This allows multiple output Streams and therefore multiple processing pipelines to share a single input Stream.
**Example**
```js
const pipeline = sharp().rotate();
pipeline.clone().resize(800, 600).pipe(firstWritableStream);
pipeline.clone().extract({ left: 20, top: 20, width: 100, height: 100 }).pipe(secondWritableStream);
readableStream.pipe(pipeline);
// firstWritableStream receives auto-rotated, resized readableStream
// secondWritableStream receives auto-rotated, extracted region of readableStream
```
**Example**
```js
// Create a pipeline that will download an image, resize it and format it to different files
// Using Promises to know when the pipeline is complete
const fs = require("fs");
const got = require("got");
const sharpStream = sharp({ failOn: 'none' });
const promises = [];
promises.push(
sharpStream
.clone()
.jpeg({ quality: 100 })
.toFile("originalFile.jpg")
);
promises.push(
sharpStream
.clone()
.resize({ width: 500 })
.jpeg({ quality: 80 })
.toFile("optimized-500.jpg")
);
promises.push(
sharpStream
.clone()
.resize({ width: 500 })
.webp({ quality: 80 })
.toFile("optimized-500.webp")
);
// https://github.com/sindresorhus/got/blob/main/documentation/3-streams.md
got.stream("https://www.example.com/some-file.jpg").pipe(sharpStream);
Promise.all(promises)
.then(res => { console.log("Done!", res); })
.catch(err => {
console.error("Error processing files, let's clean it up", err);
try {
fs.unlinkSync("originalFile.jpg");
fs.unlinkSync("optimized-500.jpg");
fs.unlinkSync("optimized-500.webp");
} catch (e) {}
});
```

View File

@@ -0,0 +1,137 @@
---
title: Input metadata
---
## metadata
> metadata([callback]) ⇒ <code>Promise.&lt;Object&gt;</code> \| <code>Sharp</code>
Fast access to (uncached) image metadata without decoding any compressed pixel data.
This is read from the header of the input image.
It does not take into consideration any operations to be applied to the output image,
such as resize or rotate.
Dimensions in the response will respect the `page` and `pages` properties of the
[constructor parameters](/api-constructor#parameters).
A `Promise` is returned when `callback` is not provided.
- `format`: Name of decoder used to decompress image data e.g. `jpeg`, `png`, `webp`, `gif`, `svg`
- `size`: Total size of image in bytes, for Stream and Buffer input only
- `width`: Number of pixels wide (EXIF orientation is not taken into consideration, see example below)
- `height`: Number of pixels high (EXIF orientation is not taken into consideration, see example below)
- `space`: Name of colour space interpretation e.g. `srgb`, `rgb`, `cmyk`, `lab`, `b-w` [...](https://www.libvips.org/API/current/VipsImage.html#VipsInterpretation)
- `channels`: Number of bands e.g. `3` for sRGB, `4` for CMYK
- `depth`: Name of pixel depth format e.g. `uchar`, `char`, `ushort`, `float` [...](https://www.libvips.org/API/current/VipsImage.html#VipsBandFormat)
- `density`: Number of pixels per inch (DPI), if present
- `chromaSubsampling`: String containing JPEG chroma subsampling, `4:2:0` or `4:4:4` for RGB, `4:2:0:4` or `4:4:4:4` for CMYK
- `isProgressive`: Boolean indicating whether the image is interlaced using a progressive scan
- `isPalette`: Boolean indicating whether the image is palette-based (GIF, PNG).
- `bitsPerSample`: Number of bits per sample for each channel (GIF, PNG, HEIF).
- `pages`: Number of pages/frames contained within the image, with support for TIFF, HEIF, PDF, animated GIF and animated WebP
- `pageHeight`: Number of pixels high each page in a multi-page image will be.
- `loop`: Number of times to loop an animated image, zero refers to a continuous loop.
- `delay`: Delay in ms between each page in an animated image, provided as an array of integers.
- `pagePrimary`: Number of the primary page in a HEIF image
- `levels`: Details of each level in a multi-level image provided as an array of objects, requires libvips compiled with support for OpenSlide
- `subifds`: Number of Sub Image File Directories in an OME-TIFF image
- `background`: Default background colour, if present, for PNG (bKGD) and GIF images
- `compression`: The encoder used to compress an HEIF file, `av1` (AVIF) or `hevc` (HEIC)
- `resolutionUnit`: The unit of resolution (density), either `inch` or `cm`, if present
- `hasProfile`: Boolean indicating the presence of an embedded ICC profile
- `hasAlpha`: Boolean indicating the presence of an alpha transparency channel
- `orientation`: Number value of the EXIF Orientation header, if present
- `exif`: Buffer containing raw EXIF data, if present
- `icc`: Buffer containing raw [ICC](https://www.npmjs.com/package/icc) profile data, if present
- `iptc`: Buffer containing raw IPTC data, if present
- `xmp`: Buffer containing raw XMP data, if present
- `tifftagPhotoshop`: Buffer containing raw TIFFTAG_PHOTOSHOP data, if present
- `formatMagick`: String containing format for images loaded via *magick
- `comments`: Array of keyword/text pairs representing PNG text blocks, if present.
| Param | Type | Description |
| --- | --- | --- |
| [callback] | <code>function</code> | called with the arguments `(err, metadata)` |
**Example**
```js
const metadata = await sharp(input).metadata();
```
**Example**
```js
const image = sharp(inputJpg);
image
.metadata()
.then(function(metadata) {
return image
.resize(Math.round(metadata.width / 2))
.webp()
.toBuffer();
})
.then(function(data) {
// data contains a WebP image half the width and height of the original JPEG
});
```
**Example**
```js
// Get dimensions taking EXIF Orientation into account.
const { autoOrient } = await sharp(input).metadata();
const { width, height } = autoOrient;
```
## stats
> stats([callback]) ⇒ <code>Promise.&lt;Object&gt;</code>
Access to pixel-derived image statistics for every channel in the image.
A `Promise` is returned when `callback` is not provided.
- `channels`: Array of channel statistics for each channel in the image. Each channel statistic contains
- `min` (minimum value in the channel)
- `max` (maximum value in the channel)
- `sum` (sum of all values in a channel)
- `squaresSum` (sum of squared values in a channel)
- `mean` (mean of the values in a channel)
- `stdev` (standard deviation for the values in a channel)
- `minX` (x-coordinate of one of the pixel where the minimum lies)
- `minY` (y-coordinate of one of the pixel where the minimum lies)
- `maxX` (x-coordinate of one of the pixel where the maximum lies)
- `maxY` (y-coordinate of one of the pixel where the maximum lies)
- `isOpaque`: Is the image fully opaque? Will be `true` if the image has no alpha channel or if every pixel is fully opaque.
- `entropy`: Histogram-based estimation of greyscale entropy, discarding alpha channel if any.
- `sharpness`: Estimation of greyscale sharpness based on the standard deviation of a Laplacian convolution, discarding alpha channel if any.
- `dominant`: Object containing most dominant sRGB colour based on a 4096-bin 3D histogram.
**Note**: Statistics are derived from the original input image. Any operations performed on the image must first be
written to a buffer in order to run `stats` on the result (see third example).
| Param | Type | Description |
| --- | --- | --- |
| [callback] | <code>function</code> | called with the arguments `(err, stats)` |
**Example**
```js
const image = sharp(inputJpg);
image
.stats()
.then(function(stats) {
// stats contains the channel-wise statistics array and the isOpaque value
});
```
**Example**
```js
const { entropy, sharpness, dominant } = await sharp(input).stats();
const { r, g, b } = dominant;
```
**Example**
```js
const image = sharp(input);
// store intermediate result
const part = await image.extract(region).toBuffer();
// create new instance to obtain statistics of extracted region
const stats = await sharp(part).stats();
```

View File

@@ -0,0 +1,676 @@
---
title: Image operations
---
## rotate
> rotate([angle], [options]) ⇒ <code>Sharp</code>
Rotate the output image.
The provided angle is converted to a valid positive degree rotation.
For example, `-450` will produce a 270 degree rotation.
When rotating by an angle other than a multiple of 90,
the background colour can be provided with the `background` option.
For backwards compatibility, if no angle is provided, `.autoOrient()` will be called.
Only one rotation can occur per pipeline (aside from an initial call without
arguments to orient via EXIF data). Previous calls to `rotate` in the same
pipeline will be ignored.
Multi-page images can only be rotated by 180 degrees.
Method order is important when rotating, resizing and/or extracting regions,
for example `.rotate(x).extract(y)` will produce a different result to `.extract(y).rotate(x)`.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [angle] | <code>number</code> | <code>auto</code> | angle of rotation. |
| [options] | <code>Object</code> | | if present, is an Object with optional attributes. |
| [options.background] | <code>string</code> \| <code>Object</code> | <code>&quot;\&quot;#000000\&quot;&quot;</code> | parsed by the [color](https://www.npmjs.org/package/color) module to extract values for red, green, blue and alpha. |
**Example**
```js
const rotateThenResize = await sharp(input)
.rotate(90)
.resize({ width: 16, height: 8, fit: 'fill' })
.toBuffer();
const resizeThenRotate = await sharp(input)
.resize({ width: 16, height: 8, fit: 'fill' })
.rotate(90)
.toBuffer();
```
## autoOrient
> autoOrient() ⇒ <code>Sharp</code>
Auto-orient based on the EXIF `Orientation` tag, then remove the tag.
Mirroring is supported and may infer the use of a flip operation.
Previous or subsequent use of `rotate(angle)` and either `flip()` or `flop()`
will logically occur after auto-orientation, regardless of call order.
**Example**
```js
const output = await sharp(input).autoOrient().toBuffer();
```
**Example**
```js
const pipeline = sharp()
.autoOrient()
.resize(null, 200)
.toBuffer(function (err, outputBuffer, info) {
// outputBuffer contains 200px high JPEG image data,
// auto-oriented using EXIF Orientation tag
// info.width and info.height contain the dimensions of the resized image
});
readableStream.pipe(pipeline);
```
## flip
> flip([flip]) ⇒ <code>Sharp</code>
Mirror the image vertically (up-down) about the x-axis.
This always occurs before rotation, if any.
This operation does not work correctly with multi-page images.
| Param | Type | Default |
| --- | --- | --- |
| [flip] | <code>Boolean</code> | <code>true</code> |
**Example**
```js
const output = await sharp(input).flip().toBuffer();
```
## flop
> flop([flop]) ⇒ <code>Sharp</code>
Mirror the image horizontally (left-right) about the y-axis.
This always occurs before rotation, if any.
| Param | Type | Default |
| --- | --- | --- |
| [flop] | <code>Boolean</code> | <code>true</code> |
**Example**
```js
const output = await sharp(input).flop().toBuffer();
```
## affine
> affine(matrix, [options]) ⇒ <code>Sharp</code>
Perform an affine transform on an image. This operation will always occur after resizing, extraction and rotation, if any.
You must provide an array of length 4 or a 2x2 affine transformation matrix.
By default, new pixels are filled with a black background. You can provide a background colour with the `background` option.
A particular interpolator may also be specified. Set the `interpolator` option to an attribute of the `sharp.interpolators` Object e.g. `sharp.interpolators.nohalo`.
In the case of a 2x2 matrix, the transform is:
- X = `matrix[0, 0]` \* (x + `idx`) + `matrix[0, 1]` \* (y + `idy`) + `odx`
- Y = `matrix[1, 0]` \* (x + `idx`) + `matrix[1, 1]` \* (y + `idy`) + `ody`
where:
- x and y are the coordinates in input image.
- X and Y are the coordinates in output image.
- (0,0) is the upper left corner.
**Throws**:
- <code>Error</code> Invalid parameters
**Since**: 0.27.0
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| matrix | <code>Array.&lt;Array.&lt;number&gt;&gt;</code> \| <code>Array.&lt;number&gt;</code> | | affine transformation matrix |
| [options] | <code>Object</code> | | if present, is an Object with optional attributes. |
| [options.background] | <code>String</code> \| <code>Object</code> | <code>&quot;#000000&quot;</code> | parsed by the [color](https://www.npmjs.org/package/color) module to extract values for red, green, blue and alpha. |
| [options.idx] | <code>Number</code> | <code>0</code> | input horizontal offset |
| [options.idy] | <code>Number</code> | <code>0</code> | input vertical offset |
| [options.odx] | <code>Number</code> | <code>0</code> | output horizontal offset |
| [options.ody] | <code>Number</code> | <code>0</code> | output vertical offset |
| [options.interpolator] | <code>String</code> | <code>sharp.interpolators.bicubic</code> | interpolator |
**Example**
```js
const pipeline = sharp()
.affine([[1, 0.3], [0.1, 0.7]], {
background: 'white',
interpolator: sharp.interpolators.nohalo
})
.toBuffer((err, outputBuffer, info) => {
// outputBuffer contains the transformed image
// info.width and info.height contain the new dimensions
});
inputStream
.pipe(pipeline);
```
## sharpen
> sharpen([options], [flat], [jagged]) ⇒ <code>Sharp</code>
Sharpen the image.
When used without parameters, performs a fast, mild sharpen of the output image.
When a `sigma` is provided, performs a slower, more accurate sharpen of the L channel in the LAB colour space.
Fine-grained control over the level of sharpening in "flat" (m1) and "jagged" (m2) areas is available.
See [libvips sharpen](https://www.libvips.org/API/current/libvips-convolution.html#vips-sharpen) operation.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> \| <code>number</code> | | if present, is an Object with attributes |
| [options.sigma] | <code>number</code> | | the sigma of the Gaussian mask, where `sigma = 1 + radius / 2`, between 0.000001 and 10 |
| [options.m1] | <code>number</code> | <code>1.0</code> | the level of sharpening to apply to "flat" areas, between 0 and 1000000 |
| [options.m2] | <code>number</code> | <code>2.0</code> | the level of sharpening to apply to "jagged" areas, between 0 and 1000000 |
| [options.x1] | <code>number</code> | <code>2.0</code> | threshold between "flat" and "jagged", between 0 and 1000000 |
| [options.y2] | <code>number</code> | <code>10.0</code> | maximum amount of brightening, between 0 and 1000000 |
| [options.y3] | <code>number</code> | <code>20.0</code> | maximum amount of darkening, between 0 and 1000000 |
| [flat] | <code>number</code> | | (deprecated) see `options.m1`. |
| [jagged] | <code>number</code> | | (deprecated) see `options.m2`. |
**Example**
```js
const data = await sharp(input).sharpen().toBuffer();
```
**Example**
```js
const data = await sharp(input).sharpen({ sigma: 2 }).toBuffer();
```
**Example**
```js
const data = await sharp(input)
.sharpen({
sigma: 2,
m1: 0,
m2: 3,
x1: 3,
y2: 15,
y3: 15,
})
.toBuffer();
```
## median
> median([size]) ⇒ <code>Sharp</code>
Apply median filter.
When used without parameters the default window is 3x3.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [size] | <code>number</code> | <code>3</code> | square mask size: size x size |
**Example**
```js
const output = await sharp(input).median().toBuffer();
```
**Example**
```js
const output = await sharp(input).median(5).toBuffer();
```
## blur
> blur([options]) ⇒ <code>Sharp</code>
Blur the image.
When used without parameters, performs a fast 3x3 box blur (equivalent to a box linear filter).
When a `sigma` is provided, performs a slower, more accurate Gaussian blur.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> \| <code>number</code> \| <code>Boolean</code> | | |
| [options.sigma] | <code>number</code> | | a value between 0.3 and 1000 representing the sigma of the Gaussian mask, where `sigma = 1 + radius / 2`. |
| [options.precision] | <code>string</code> | <code>&quot;&#x27;integer&#x27;&quot;</code> | How accurate the operation should be, one of: integer, float, approximate. |
| [options.minAmplitude] | <code>number</code> | <code>0.2</code> | A value between 0.001 and 1. A smaller value will generate a larger, more accurate mask. |
**Example**
```js
const boxBlurred = await sharp(input)
.blur()
.toBuffer();
```
**Example**
```js
const gaussianBlurred = await sharp(input)
.blur(5)
.toBuffer();
```
## flatten
> flatten([options]) ⇒ <code>Sharp</code>
Merge alpha transparency channel, if any, with a background, then remove the alpha channel.
See also [removeAlpha](/api-channel#removealpha).
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> | | |
| [options.background] | <code>string</code> \| <code>Object</code> | <code>&quot;{r: 0, g: 0, b: 0}&quot;</code> | background colour, parsed by the [color](https://www.npmjs.org/package/color) module, defaults to black. |
**Example**
```js
await sharp(rgbaInput)
.flatten({ background: '#F0A703' })
.toBuffer();
```
## unflatten
> unflatten()
Ensure the image has an alpha channel
with all white pixel values made fully transparent.
Existing alpha channel values for non-white pixels remain unchanged.
This feature is experimental and the API may change.
**Since**: 0.32.1
**Example**
```js
await sharp(rgbInput)
.unflatten()
.toBuffer();
```
**Example**
```js
await sharp(rgbInput)
.threshold(128, { grayscale: false }) // converter bright pixels to white
.unflatten()
.toBuffer();
```
## gamma
> gamma([gamma], [gammaOut]) ⇒ <code>Sharp</code>
Apply a gamma correction by reducing the encoding (darken) pre-resize at a factor of `1/gamma`
then increasing the encoding (brighten) post-resize at a factor of `gamma`.
This can improve the perceived brightness of a resized image in non-linear colour spaces.
JPEG and WebP input images will not take advantage of the shrink-on-load performance optimisation
when applying a gamma correction.
Supply a second argument to use a different output gamma value, otherwise the first value is used in both cases.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [gamma] | <code>number</code> | <code>2.2</code> | value between 1.0 and 3.0. |
| [gammaOut] | <code>number</code> | | value between 1.0 and 3.0. (optional, defaults to same as `gamma`) |
## negate
> negate([options]) ⇒ <code>Sharp</code>
Produce the "negative" of the image.
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> | | |
| [options.alpha] | <code>Boolean</code> | <code>true</code> | Whether or not to negate any alpha channel |
**Example**
```js
const output = await sharp(input)
.negate()
.toBuffer();
```
**Example**
```js
const output = await sharp(input)
.negate({ alpha: false })
.toBuffer();
```
## normalise
> normalise([options]) ⇒ <code>Sharp</code>
Enhance output image contrast by stretching its luminance to cover a full dynamic range.
Uses a histogram-based approach, taking a default range of 1% to 99% to reduce sensitivity to noise at the extremes.
Luminance values below the `lower` percentile will be underexposed by clipping to zero.
Luminance values above the `upper` percentile will be overexposed by clipping to the max pixel value.
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> | | |
| [options.lower] | <code>number</code> | <code>1</code> | Percentile below which luminance values will be underexposed. |
| [options.upper] | <code>number</code> | <code>99</code> | Percentile above which luminance values will be overexposed. |
**Example**
```js
const output = await sharp(input)
.normalise()
.toBuffer();
```
**Example**
```js
const output = await sharp(input)
.normalise({ lower: 0, upper: 100 })
.toBuffer();
```
## normalize
> normalize([options]) ⇒ <code>Sharp</code>
Alternative spelling of normalise.
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> | | |
| [options.lower] | <code>number</code> | <code>1</code> | Percentile below which luminance values will be underexposed. |
| [options.upper] | <code>number</code> | <code>99</code> | Percentile above which luminance values will be overexposed. |
**Example**
```js
const output = await sharp(input)
.normalize()
.toBuffer();
```
## clahe
> clahe(options) ⇒ <code>Sharp</code>
Perform contrast limiting adaptive histogram equalization
[CLAHE](https://en.wikipedia.org/wiki/Adaptive_histogram_equalization#Contrast_Limited_AHE).
This will, in general, enhance the clarity of the image by bringing out darker details.
**Throws**:
- <code>Error</code> Invalid parameters
**Since**: 0.28.3
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| options | <code>Object</code> | | |
| options.width | <code>number</code> | | Integral width of the search window, in pixels. |
| options.height | <code>number</code> | | Integral height of the search window, in pixels. |
| [options.maxSlope] | <code>number</code> | <code>3</code> | Integral level of brightening, between 0 and 100, where 0 disables contrast limiting. |
**Example**
```js
const output = await sharp(input)
.clahe({
width: 3,
height: 3,
})
.toBuffer();
```
## convolve
> convolve(kernel) ⇒ <code>Sharp</code>
Convolve the image with the specified kernel.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| kernel | <code>Object</code> | | |
| kernel.width | <code>number</code> | | width of the kernel in pixels. |
| kernel.height | <code>number</code> | | height of the kernel in pixels. |
| kernel.kernel | <code>Array.&lt;number&gt;</code> | | Array of length `width*height` containing the kernel values. |
| [kernel.scale] | <code>number</code> | <code>sum</code> | the scale of the kernel in pixels. |
| [kernel.offset] | <code>number</code> | <code>0</code> | the offset of the kernel in pixels. |
**Example**
```js
sharp(input)
.convolve({
width: 3,
height: 3,
kernel: [-1, 0, 1, -2, 0, 2, -1, 0, 1]
})
.raw()
.toBuffer(function(err, data, info) {
// data contains the raw pixel data representing the convolution
// of the input image with the horizontal Sobel operator
});
```
## threshold
> threshold([threshold], [options]) ⇒ <code>Sharp</code>
Any pixel value greater than or equal to the threshold value will be set to 255, otherwise it will be set to 0.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [threshold] | <code>number</code> | <code>128</code> | a value in the range 0-255 representing the level at which the threshold will be applied. |
| [options] | <code>Object</code> | | |
| [options.greyscale] | <code>Boolean</code> | <code>true</code> | convert to single channel greyscale. |
| [options.grayscale] | <code>Boolean</code> | <code>true</code> | alternative spelling for greyscale. |
## boolean
> boolean(operand, operator, [options]) ⇒ <code>Sharp</code>
Perform a bitwise boolean operation with operand image.
This operation creates an output image where each pixel is the result of
the selected bitwise boolean `operation` between the corresponding pixels of the input images.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Description |
| --- | --- | --- |
| operand | <code>Buffer</code> \| <code>string</code> | Buffer containing image data or string containing the path to an image file. |
| operator | <code>string</code> | one of `and`, `or` or `eor` to perform that bitwise operation, like the C logic operators `&`, `|` and `^` respectively. |
| [options] | <code>Object</code> | |
| [options.raw] | <code>Object</code> | describes operand when using raw pixel data. |
| [options.raw.width] | <code>number</code> | |
| [options.raw.height] | <code>number</code> | |
| [options.raw.channels] | <code>number</code> | |
## linear
> linear([a], [b]) ⇒ <code>Sharp</code>
Apply the linear formula `a` * input + `b` to the image to adjust image levels.
When a single number is provided, it will be used for all image channels.
When an array of numbers is provided, the array length must match the number of channels.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [a] | <code>number</code> \| <code>Array.&lt;number&gt;</code> | <code>[]</code> | multiplier |
| [b] | <code>number</code> \| <code>Array.&lt;number&gt;</code> | <code>[]</code> | offset |
**Example**
```js
await sharp(input)
.linear(0.5, 2)
.toBuffer();
```
**Example**
```js
await sharp(rgbInput)
.linear(
[0.25, 0.5, 0.75],
[150, 100, 50]
)
.toBuffer();
```
## recomb
> recomb(inputMatrix) ⇒ <code>Sharp</code>
Recombine the image with the specified matrix.
**Throws**:
- <code>Error</code> Invalid parameters
**Since**: 0.21.1
| Param | Type | Description |
| --- | --- | --- |
| inputMatrix | <code>Array.&lt;Array.&lt;number&gt;&gt;</code> | 3x3 or 4x4 Recombination matrix |
**Example**
```js
sharp(input)
.recomb([
[0.3588, 0.7044, 0.1368],
[0.2990, 0.5870, 0.1140],
[0.2392, 0.4696, 0.0912],
])
.raw()
.toBuffer(function(err, data, info) {
// data contains the raw pixel data after applying the matrix
// With this example input, a sepia filter has been applied
});
```
## modulate
> modulate([options]) ⇒ <code>Sharp</code>
Transforms the image using brightness, saturation, hue rotation, and lightness.
Brightness and lightness both operate on luminance, with the difference being that
brightness is multiplicative whereas lightness is additive.
**Since**: 0.22.1
| Param | Type | Description |
| --- | --- | --- |
| [options] | <code>Object</code> | |
| [options.brightness] | <code>number</code> | Brightness multiplier |
| [options.saturation] | <code>number</code> | Saturation multiplier |
| [options.hue] | <code>number</code> | Degrees for hue rotation |
| [options.lightness] | <code>number</code> | Lightness addend |
**Example**
```js
// increase brightness by a factor of 2
const output = await sharp(input)
.modulate({
brightness: 2
})
.toBuffer();
```
**Example**
```js
// hue-rotate by 180 degrees
const output = await sharp(input)
.modulate({
hue: 180
})
.toBuffer();
```
**Example**
```js
// increase lightness by +50
const output = await sharp(input)
.modulate({
lightness: 50
})
.toBuffer();
```
**Example**
```js
// decrease brightness and saturation while also hue-rotating by 90 degrees
const output = await sharp(input)
.modulate({
brightness: 0.5,
saturation: 0.5,
hue: 90,
})
.toBuffer();
```

View File

@@ -0,0 +1,846 @@
---
title: Output options
---
## toFile
> toFile(fileOut, [callback]) ⇒ <code>Promise.&lt;Object&gt;</code>
Write output image data to a file.
If an explicit output format is not selected, it will be inferred from the extension,
with JPEG, PNG, WebP, AVIF, TIFF, GIF, DZI, and libvips' V format supported.
Note that raw pixel data is only supported for buffer output.
By default all metadata will be removed, which includes EXIF-based orientation.
See [withMetadata](#withmetadata) for control over this.
The caller is responsible for ensuring directory structures and permissions exist.
A `Promise` is returned when `callback` is not provided.
**Returns**: <code>Promise.&lt;Object&gt;</code> - - when no callback is provided
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Description |
| --- | --- | --- |
| fileOut | <code>string</code> | the path to write the image data to. |
| [callback] | <code>function</code> | called on completion with two arguments `(err, info)`. `info` contains the output image `format`, `size` (bytes), `width`, `height`, `channels` and `premultiplied` (indicating if premultiplication was used). When using a crop strategy also contains `cropOffsetLeft` and `cropOffsetTop`. When using the attention crop strategy also contains `attentionX` and `attentionY`, the focal point of the cropped region. Animated output will also contain `pageHeight` and `pages`. May also contain `textAutofitDpi` (dpi the font was rendered at) if image was created from text. |
**Example**
```js
sharp(input)
.toFile('output.png', (err, info) => { ... });
```
**Example**
```js
sharp(input)
.toFile('output.png')
.then(info => { ... })
.catch(err => { ... });
```
## toBuffer
> toBuffer([options], [callback]) ⇒ <code>Promise.&lt;Buffer&gt;</code>
Write output to a Buffer.
JPEG, PNG, WebP, AVIF, TIFF, GIF and raw pixel data output are supported.
Use [toFormat](#toformat) or one of the format-specific functions such as [jpeg](#jpeg), [png](#png) etc. to set the output format.
If no explicit format is set, the output format will match the input image, except SVG input which becomes PNG output.
By default all metadata will be removed, which includes EXIF-based orientation.
See [withMetadata](#withmetadata) for control over this.
`callback`, if present, gets three arguments `(err, data, info)` where:
- `err` is an error, if any.
- `data` is the output image data.
- `info` contains the output image `format`, `size` (bytes), `width`, `height`,
`channels` and `premultiplied` (indicating if premultiplication was used).
When using a crop strategy also contains `cropOffsetLeft` and `cropOffsetTop`.
Animated output will also contain `pageHeight` and `pages`.
May also contain `textAutofitDpi` (dpi the font was rendered at) if image was created from text.
A `Promise` is returned when `callback` is not provided.
**Returns**: <code>Promise.&lt;Buffer&gt;</code> - - when no callback is provided
| Param | Type | Description |
| --- | --- | --- |
| [options] | <code>Object</code> | |
| [options.resolveWithObject] | <code>boolean</code> | Resolve the Promise with an Object containing `data` and `info` properties instead of resolving only with `data`. |
| [callback] | <code>function</code> | |
**Example**
```js
sharp(input)
.toBuffer((err, data, info) => { ... });
```
**Example**
```js
sharp(input)
.toBuffer()
.then(data => { ... })
.catch(err => { ... });
```
**Example**
```js
sharp(input)
.png()
.toBuffer({ resolveWithObject: true })
.then(({ data, info }) => { ... })
.catch(err => { ... });
```
**Example**
```js
const { data, info } = await sharp('my-image.jpg')
// output the raw pixels
.raw()
.toBuffer({ resolveWithObject: true });
// create a more type safe way to work with the raw pixel data
// this will not copy the data, instead it will change `data`s underlying ArrayBuffer
// so `data` and `pixelArray` point to the same memory location
const pixelArray = new Uint8ClampedArray(data.buffer);
// When you are done changing the pixelArray, sharp takes the `pixelArray` as an input
const { width, height, channels } = info;
await sharp(pixelArray, { raw: { width, height, channels } })
.toFile('my-changed-image.jpg');
```
## keepExif
> keepExif() ⇒ <code>Sharp</code>
Keep all EXIF metadata from the input image in the output image.
EXIF metadata is unsupported for TIFF output.
**Since**: 0.33.0
**Example**
```js
const outputWithExif = await sharp(inputWithExif)
.keepExif()
.toBuffer();
```
## withExif
> withExif(exif) ⇒ <code>Sharp</code>
Set EXIF metadata in the output image, ignoring any EXIF in the input image.
**Throws**:
- <code>Error</code> Invalid parameters
**Since**: 0.33.0
| Param | Type | Description |
| --- | --- | --- |
| exif | <code>Object.&lt;string, Object.&lt;string, string&gt;&gt;</code> | Object keyed by IFD0, IFD1 etc. of key/value string pairs to write as EXIF data. |
**Example**
```js
const dataWithExif = await sharp(input)
.withExif({
IFD0: {
Copyright: 'The National Gallery'
},
IFD3: {
GPSLatitudeRef: 'N',
GPSLatitude: '51/1 30/1 3230/100',
GPSLongitudeRef: 'W',
GPSLongitude: '0/1 7/1 4366/100'
}
})
.toBuffer();
```
## withExifMerge
> withExifMerge(exif) ⇒ <code>Sharp</code>
Update EXIF metadata from the input image in the output image.
**Throws**:
- <code>Error</code> Invalid parameters
**Since**: 0.33.0
| Param | Type | Description |
| --- | --- | --- |
| exif | <code>Object.&lt;string, Object.&lt;string, string&gt;&gt;</code> | Object keyed by IFD0, IFD1 etc. of key/value string pairs to write as EXIF data. |
**Example**
```js
const dataWithMergedExif = await sharp(inputWithExif)
.withExifMerge({
IFD0: {
Copyright: 'The National Gallery'
}
})
.toBuffer();
```
## keepIccProfile
> keepIccProfile() ⇒ <code>Sharp</code>
Keep ICC profile from the input image in the output image.
Where necessary, will attempt to convert the output colour space to match the profile.
**Since**: 0.33.0
**Example**
```js
const outputWithIccProfile = await sharp(inputWithIccProfile)
.keepIccProfile()
.toBuffer();
```
## withIccProfile
> withIccProfile(icc, [options]) ⇒ <code>Sharp</code>
Transform using an ICC profile and attach to the output image.
This can either be an absolute filesystem path or
built-in profile name (`srgb`, `p3`, `cmyk`).
**Throws**:
- <code>Error</code> Invalid parameters
**Since**: 0.33.0
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| icc | <code>string</code> | | Absolute filesystem path to output ICC profile or built-in profile name (srgb, p3, cmyk). |
| [options] | <code>Object</code> | | |
| [options.attach] | <code>number</code> | <code>true</code> | Should the ICC profile be included in the output image metadata? |
**Example**
```js
const outputWithP3 = await sharp(input)
.withIccProfile('p3')
.toBuffer();
```
## keepMetadata
> keepMetadata() ⇒ <code>Sharp</code>
Keep all metadata (EXIF, ICC, XMP, IPTC) from the input image in the output image.
The default behaviour, when `keepMetadata` is not used, is to convert to the device-independent
sRGB colour space and strip all metadata, including the removal of any ICC profile.
**Since**: 0.33.0
**Example**
```js
const outputWithMetadata = await sharp(inputWithMetadata)
.keepMetadata()
.toBuffer();
```
## withMetadata
> withMetadata([options]) ⇒ <code>Sharp</code>
Keep most metadata (EXIF, XMP, IPTC) from the input image in the output image.
This will also convert to and add a web-friendly sRGB ICC profile if appropriate.
Allows orientation and density to be set or updated.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Description |
| --- | --- | --- |
| [options] | <code>Object</code> | |
| [options.orientation] | <code>number</code> | Used to update the EXIF `Orientation` tag, integer between 1 and 8. |
| [options.density] | <code>number</code> | Number of pixels per inch (DPI). |
**Example**
```js
const outputSrgbWithMetadata = await sharp(inputRgbWithMetadata)
.withMetadata()
.toBuffer();
```
**Example**
```js
// Set output metadata to 96 DPI
const data = await sharp(input)
.withMetadata({ density: 96 })
.toBuffer();
```
## toFormat
> toFormat(format, options) ⇒ <code>Sharp</code>
Force output to a given format.
**Throws**:
- <code>Error</code> unsupported format or options
| Param | Type | Description |
| --- | --- | --- |
| format | <code>string</code> \| <code>Object</code> | as a string or an Object with an 'id' attribute |
| options | <code>Object</code> | output options |
**Example**
```js
// Convert any input to PNG output
const data = await sharp(input)
.toFormat('png')
.toBuffer();
```
## jpeg
> jpeg([options]) ⇒ <code>Sharp</code>
Use these JPEG options for output image.
**Throws**:
- <code>Error</code> Invalid options
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> | | output options |
| [options.quality] | <code>number</code> | <code>80</code> | quality, integer 1-100 |
| [options.progressive] | <code>boolean</code> | <code>false</code> | use progressive (interlace) scan |
| [options.chromaSubsampling] | <code>string</code> | <code>&quot;&#x27;4:2:0&#x27;&quot;</code> | set to '4:4:4' to prevent chroma subsampling otherwise defaults to '4:2:0' chroma subsampling |
| [options.optimiseCoding] | <code>boolean</code> | <code>true</code> | optimise Huffman coding tables |
| [options.optimizeCoding] | <code>boolean</code> | <code>true</code> | alternative spelling of optimiseCoding |
| [options.mozjpeg] | <code>boolean</code> | <code>false</code> | use mozjpeg defaults, equivalent to `{ trellisQuantisation: true, overshootDeringing: true, optimiseScans: true, quantisationTable: 3 }` |
| [options.trellisQuantisation] | <code>boolean</code> | <code>false</code> | apply trellis quantisation |
| [options.overshootDeringing] | <code>boolean</code> | <code>false</code> | apply overshoot deringing |
| [options.optimiseScans] | <code>boolean</code> | <code>false</code> | optimise progressive scans, forces progressive |
| [options.optimizeScans] | <code>boolean</code> | <code>false</code> | alternative spelling of optimiseScans |
| [options.quantisationTable] | <code>number</code> | <code>0</code> | quantization table to use, integer 0-8 |
| [options.quantizationTable] | <code>number</code> | <code>0</code> | alternative spelling of quantisationTable |
| [options.force] | <code>boolean</code> | <code>true</code> | force JPEG output, otherwise attempt to use input format |
**Example**
```js
// Convert any input to very high quality JPEG output
const data = await sharp(input)
.jpeg({
quality: 100,
chromaSubsampling: '4:4:4'
})
.toBuffer();
```
**Example**
```js
// Use mozjpeg to reduce output JPEG file size (slower)
const data = await sharp(input)
.jpeg({ mozjpeg: true })
.toBuffer();
```
## png
> png([options]) ⇒ <code>Sharp</code>
Use these PNG options for output image.
By default, PNG output is full colour at 8 bits per pixel.
Indexed PNG input at 1, 2 or 4 bits per pixel is converted to 8 bits per pixel.
Set `palette` to `true` for slower, indexed PNG output.
For 16 bits per pixel output, convert to `rgb16` via
[toColourspace](/api-colour#tocolourspace).
**Throws**:
- <code>Error</code> Invalid options
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> | | |
| [options.progressive] | <code>boolean</code> | <code>false</code> | use progressive (interlace) scan |
| [options.compressionLevel] | <code>number</code> | <code>6</code> | zlib compression level, 0 (fastest, largest) to 9 (slowest, smallest) |
| [options.adaptiveFiltering] | <code>boolean</code> | <code>false</code> | use adaptive row filtering |
| [options.palette] | <code>boolean</code> | <code>false</code> | quantise to a palette-based image with alpha transparency support |
| [options.quality] | <code>number</code> | <code>100</code> | use the lowest number of colours needed to achieve given quality, sets `palette` to `true` |
| [options.effort] | <code>number</code> | <code>7</code> | CPU effort, between 1 (fastest) and 10 (slowest), sets `palette` to `true` |
| [options.colours] | <code>number</code> | <code>256</code> | maximum number of palette entries, sets `palette` to `true` |
| [options.colors] | <code>number</code> | <code>256</code> | alternative spelling of `options.colours`, sets `palette` to `true` |
| [options.dither] | <code>number</code> | <code>1.0</code> | level of Floyd-Steinberg error diffusion, sets `palette` to `true` |
| [options.force] | <code>boolean</code> | <code>true</code> | force PNG output, otherwise attempt to use input format |
**Example**
```js
// Convert any input to full colour PNG output
const data = await sharp(input)
.png()
.toBuffer();
```
**Example**
```js
// Convert any input to indexed PNG output (slower)
const data = await sharp(input)
.png({ palette: true })
.toBuffer();
```
**Example**
```js
// Output 16 bits per pixel RGB(A)
const data = await sharp(input)
.toColourspace('rgb16')
.png()
.toBuffer();
```
## webp
> webp([options]) ⇒ <code>Sharp</code>
Use these WebP options for output image.
**Throws**:
- <code>Error</code> Invalid options
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> | | output options |
| [options.quality] | <code>number</code> | <code>80</code> | quality, integer 1-100 |
| [options.alphaQuality] | <code>number</code> | <code>100</code> | quality of alpha layer, integer 0-100 |
| [options.lossless] | <code>boolean</code> | <code>false</code> | use lossless compression mode |
| [options.nearLossless] | <code>boolean</code> | <code>false</code> | use near_lossless compression mode |
| [options.smartSubsample] | <code>boolean</code> | <code>false</code> | use high quality chroma subsampling |
| [options.smartDeblock] | <code>boolean</code> | <code>false</code> | auto-adjust the deblocking filter, can improve low contrast edges (slow) |
| [options.preset] | <code>string</code> | <code>&quot;&#x27;default&#x27;&quot;</code> | named preset for preprocessing/filtering, one of: default, photo, picture, drawing, icon, text |
| [options.effort] | <code>number</code> | <code>4</code> | CPU effort, between 0 (fastest) and 6 (slowest) |
| [options.loop] | <code>number</code> | <code>0</code> | number of animation iterations, use 0 for infinite animation |
| [options.delay] | <code>number</code> \| <code>Array.&lt;number&gt;</code> | | delay(s) between animation frames (in milliseconds) |
| [options.minSize] | <code>boolean</code> | <code>false</code> | prevent use of animation key frames to minimise file size (slow) |
| [options.mixed] | <code>boolean</code> | <code>false</code> | allow mixture of lossy and lossless animation frames (slow) |
| [options.force] | <code>boolean</code> | <code>true</code> | force WebP output, otherwise attempt to use input format |
**Example**
```js
// Convert any input to lossless WebP output
const data = await sharp(input)
.webp({ lossless: true })
.toBuffer();
```
**Example**
```js
// Optimise the file size of an animated WebP
const outputWebp = await sharp(inputWebp, { animated: true })
.webp({ effort: 6 })
.toBuffer();
```
## gif
> gif([options]) ⇒ <code>Sharp</code>
Use these GIF options for the output image.
The first entry in the palette is reserved for transparency.
The palette of the input image will be re-used if possible.
**Throws**:
- <code>Error</code> Invalid options
**Since**: 0.30.0
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> | | output options |
| [options.reuse] | <code>boolean</code> | <code>true</code> | re-use existing palette, otherwise generate new (slow) |
| [options.progressive] | <code>boolean</code> | <code>false</code> | use progressive (interlace) scan |
| [options.colours] | <code>number</code> | <code>256</code> | maximum number of palette entries, including transparency, between 2 and 256 |
| [options.colors] | <code>number</code> | <code>256</code> | alternative spelling of `options.colours` |
| [options.effort] | <code>number</code> | <code>7</code> | CPU effort, between 1 (fastest) and 10 (slowest) |
| [options.dither] | <code>number</code> | <code>1.0</code> | level of Floyd-Steinberg error diffusion, between 0 (least) and 1 (most) |
| [options.interFrameMaxError] | <code>number</code> | <code>0</code> | maximum inter-frame error for transparency, between 0 (lossless) and 32 |
| [options.interPaletteMaxError] | <code>number</code> | <code>3</code> | maximum inter-palette error for palette reuse, between 0 and 256 |
| [options.loop] | <code>number</code> | <code>0</code> | number of animation iterations, use 0 for infinite animation |
| [options.delay] | <code>number</code> \| <code>Array.&lt;number&gt;</code> | | delay(s) between animation frames (in milliseconds) |
| [options.force] | <code>boolean</code> | <code>true</code> | force GIF output, otherwise attempt to use input format |
**Example**
```js
// Convert PNG to GIF
await sharp(pngBuffer)
.gif()
.toBuffer();
```
**Example**
```js
// Convert animated WebP to animated GIF
await sharp('animated.webp', { animated: true })
.toFile('animated.gif');
```
**Example**
```js
// Create a 128x128, cropped, non-dithered, animated thumbnail of an animated GIF
const out = await sharp('in.gif', { animated: true })
.resize({ width: 128, height: 128 })
.gif({ dither: 0 })
.toBuffer();
```
**Example**
```js
// Lossy file size reduction of animated GIF
await sharp('in.gif', { animated: true })
.gif({ interFrameMaxError: 8 })
.toFile('optim.gif');
```
## jp2
> jp2([options]) ⇒ <code>Sharp</code>
Use these JP2 options for output image.
Requires libvips compiled with support for OpenJPEG.
The prebuilt binaries do not include this - see
[installing a custom libvips](https://sharp.pixelplumbing.com/install#custom-libvips).
**Throws**:
- <code>Error</code> Invalid options
**Since**: 0.29.1
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> | | output options |
| [options.quality] | <code>number</code> | <code>80</code> | quality, integer 1-100 |
| [options.lossless] | <code>boolean</code> | <code>false</code> | use lossless compression mode |
| [options.tileWidth] | <code>number</code> | <code>512</code> | horizontal tile size |
| [options.tileHeight] | <code>number</code> | <code>512</code> | vertical tile size |
| [options.chromaSubsampling] | <code>string</code> | <code>&quot;&#x27;4:4:4&#x27;&quot;</code> | set to '4:2:0' to use chroma subsampling |
**Example**
```js
// Convert any input to lossless JP2 output
const data = await sharp(input)
.jp2({ lossless: true })
.toBuffer();
```
**Example**
```js
// Convert any input to very high quality JP2 output
const data = await sharp(input)
.jp2({
quality: 100,
chromaSubsampling: '4:4:4'
})
.toBuffer();
```
## tiff
> tiff([options]) ⇒ <code>Sharp</code>
Use these TIFF options for output image.
The `density` can be set in pixels/inch via [withMetadata](#withmetadata)
instead of providing `xres` and `yres` in pixels/mm.
**Throws**:
- <code>Error</code> Invalid options
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> | | output options |
| [options.quality] | <code>number</code> | <code>80</code> | quality, integer 1-100 |
| [options.force] | <code>boolean</code> | <code>true</code> | force TIFF output, otherwise attempt to use input format |
| [options.compression] | <code>string</code> | <code>&quot;&#x27;jpeg&#x27;&quot;</code> | compression options: none, jpeg, deflate, packbits, ccittfax4, lzw, webp, zstd, jp2k |
| [options.predictor] | <code>string</code> | <code>&quot;&#x27;horizontal&#x27;&quot;</code> | compression predictor options: none, horizontal, float |
| [options.pyramid] | <code>boolean</code> | <code>false</code> | write an image pyramid |
| [options.tile] | <code>boolean</code> | <code>false</code> | write a tiled tiff |
| [options.tileWidth] | <code>number</code> | <code>256</code> | horizontal tile size |
| [options.tileHeight] | <code>number</code> | <code>256</code> | vertical tile size |
| [options.xres] | <code>number</code> | <code>1.0</code> | horizontal resolution in pixels/mm |
| [options.yres] | <code>number</code> | <code>1.0</code> | vertical resolution in pixels/mm |
| [options.resolutionUnit] | <code>string</code> | <code>&quot;&#x27;inch&#x27;&quot;</code> | resolution unit options: inch, cm |
| [options.bitdepth] | <code>number</code> | <code>8</code> | reduce bitdepth to 1, 2 or 4 bit |
| [options.miniswhite] | <code>boolean</code> | <code>false</code> | write 1-bit images as miniswhite |
**Example**
```js
// Convert SVG input to LZW-compressed, 1 bit per pixel TIFF output
sharp('input.svg')
.tiff({
compression: 'lzw',
bitdepth: 1
})
.toFile('1-bpp-output.tiff')
.then(info => { ... });
```
## avif
> avif([options]) ⇒ <code>Sharp</code>
Use these AVIF options for output image.
AVIF image sequences are not supported.
Prebuilt binaries support a bitdepth of 8 only.
**Throws**:
- <code>Error</code> Invalid options
**Since**: 0.27.0
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> | | output options |
| [options.quality] | <code>number</code> | <code>50</code> | quality, integer 1-100 |
| [options.lossless] | <code>boolean</code> | <code>false</code> | use lossless compression |
| [options.effort] | <code>number</code> | <code>4</code> | CPU effort, between 0 (fastest) and 9 (slowest) |
| [options.chromaSubsampling] | <code>string</code> | <code>&quot;&#x27;4:4:4&#x27;&quot;</code> | set to '4:2:0' to use chroma subsampling |
| [options.bitdepth] | <code>number</code> | <code>8</code> | set bitdepth to 8, 10 or 12 bit |
**Example**
```js
const data = await sharp(input)
.avif({ effort: 2 })
.toBuffer();
```
**Example**
```js
const data = await sharp(input)
.avif({ lossless: true })
.toBuffer();
```
## heif
> heif(options) ⇒ <code>Sharp</code>
Use these HEIF options for output image.
Support for patent-encumbered HEIC images using `hevc` compression requires the use of a
globally-installed libvips compiled with support for libheif, libde265 and x265.
**Throws**:
- <code>Error</code> Invalid options
**Since**: 0.23.0
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| options | <code>Object</code> | | output options |
| options.compression | <code>string</code> | | compression format: av1, hevc |
| [options.quality] | <code>number</code> | <code>50</code> | quality, integer 1-100 |
| [options.lossless] | <code>boolean</code> | <code>false</code> | use lossless compression |
| [options.effort] | <code>number</code> | <code>4</code> | CPU effort, between 0 (fastest) and 9 (slowest) |
| [options.chromaSubsampling] | <code>string</code> | <code>&quot;&#x27;4:4:4&#x27;&quot;</code> | set to '4:2:0' to use chroma subsampling |
| [options.bitdepth] | <code>number</code> | <code>8</code> | set bitdepth to 8, 10 or 12 bit |
**Example**
```js
const data = await sharp(input)
.heif({ compression: 'hevc' })
.toBuffer();
```
## jxl
> jxl([options]) ⇒ <code>Sharp</code>
Use these JPEG-XL (JXL) options for output image.
This feature is experimental, please do not use in production systems.
Requires libvips compiled with support for libjxl.
The prebuilt binaries do not include this - see
[installing a custom libvips](https://sharp.pixelplumbing.com/install#custom-libvips).
**Throws**:
- <code>Error</code> Invalid options
**Since**: 0.31.3
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> | | output options |
| [options.distance] | <code>number</code> | <code>1.0</code> | maximum encoding error, between 0 (highest quality) and 15 (lowest quality) |
| [options.quality] | <code>number</code> | | calculate `distance` based on JPEG-like quality, between 1 and 100, overrides distance if specified |
| [options.decodingTier] | <code>number</code> | <code>0</code> | target decode speed tier, between 0 (highest quality) and 4 (lowest quality) |
| [options.lossless] | <code>boolean</code> | <code>false</code> | use lossless compression |
| [options.effort] | <code>number</code> | <code>7</code> | CPU effort, between 1 (fastest) and 9 (slowest) |
| [options.loop] | <code>number</code> | <code>0</code> | number of animation iterations, use 0 for infinite animation |
| [options.delay] | <code>number</code> \| <code>Array.&lt;number&gt;</code> | | delay(s) between animation frames (in milliseconds) |
## raw
> raw([options]) ⇒ <code>Sharp</code>
Force output to be raw, uncompressed pixel data.
Pixel ordering is left-to-right, top-to-bottom, without padding.
Channel ordering will be RGB or RGBA for non-greyscale colourspaces.
**Throws**:
- <code>Error</code> Invalid options
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> | | output options |
| [options.depth] | <code>string</code> | <code>&quot;&#x27;uchar&#x27;&quot;</code> | bit depth, one of: char, uchar (default), short, ushort, int, uint, float, complex, double, dpcomplex |
**Example**
```js
// Extract raw, unsigned 8-bit RGB pixel data from JPEG input
const { data, info } = await sharp('input.jpg')
.raw()
.toBuffer({ resolveWithObject: true });
```
**Example**
```js
// Extract alpha channel as raw, unsigned 16-bit pixel data from PNG input
const data = await sharp('input.png')
.ensureAlpha()
.extractChannel(3)
.toColourspace('b-w')
.raw({ depth: 'ushort' })
.toBuffer();
```
## tile
> tile([options]) ⇒ <code>Sharp</code>
Use tile-based deep zoom (image pyramid) output.
Set the format and options for tile images via the `toFormat`, `jpeg`, `png` or `webp` functions.
Use a `.zip` or `.szi` file extension with `toFile` to write to a compressed archive file format.
The container will be set to `zip` when the output is a Buffer or Stream, otherwise it will default to `fs`.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> | | |
| [options.size] | <code>number</code> | <code>256</code> | tile size in pixels, a value between 1 and 8192. |
| [options.overlap] | <code>number</code> | <code>0</code> | tile overlap in pixels, a value between 0 and 8192. |
| [options.angle] | <code>number</code> | <code>0</code> | tile angle of rotation, must be a multiple of 90. |
| [options.background] | <code>string</code> \| <code>Object</code> | <code>&quot;{r: 255, g: 255, b: 255, alpha: 1}&quot;</code> | background colour, parsed by the [color](https://www.npmjs.org/package/color) module, defaults to white without transparency. |
| [options.depth] | <code>string</code> | | how deep to make the pyramid, possible values are `onepixel`, `onetile` or `one`, default based on layout. |
| [options.skipBlanks] | <code>number</code> | <code>-1</code> | Threshold to skip tile generation. Range is 0-255 for 8-bit images, 0-65535 for 16-bit images. Default is 5 for `google` layout, -1 (no skip) otherwise. |
| [options.container] | <code>string</code> | <code>&quot;&#x27;fs&#x27;&quot;</code> | tile container, with value `fs` (filesystem) or `zip` (compressed file). |
| [options.layout] | <code>string</code> | <code>&quot;&#x27;dz&#x27;&quot;</code> | filesystem layout, possible values are `dz`, `iiif`, `iiif3`, `zoomify` or `google`. |
| [options.centre] | <code>boolean</code> | <code>false</code> | centre image in tile. |
| [options.center] | <code>boolean</code> | <code>false</code> | alternative spelling of centre. |
| [options.id] | <code>string</code> | <code>&quot;&#x27;https://example.com/iiif&#x27;&quot;</code> | when `layout` is `iiif`/`iiif3`, sets the `@id`/`id` attribute of `info.json` |
| [options.basename] | <code>string</code> | | the name of the directory within the zip file when container is `zip`. |
**Example**
```js
sharp('input.tiff')
.png()
.tile({
size: 512
})
.toFile('output.dz', function(err, info) {
// output.dzi is the Deep Zoom XML definition
// output_files contains 512x512 tiles grouped by zoom level
});
```
**Example**
```js
const zipFileWithTiles = await sharp(input)
.tile({ basename: "tiles" })
.toBuffer();
```
**Example**
```js
const iiififier = sharp().tile({ layout: "iiif" });
readableStream
.pipe(iiififier)
.pipe(writeableStream);
```
## timeout
> timeout(options) ⇒ <code>Sharp</code>
Set a timeout for processing, in seconds.
Use a value of zero to continue processing indefinitely, the default behaviour.
The clock starts when libvips opens an input image for processing.
Time spent waiting for a libuv thread to become available is not included.
**Since**: 0.29.2
| Param | Type | Description |
| --- | --- | --- |
| options | <code>Object</code> | |
| options.seconds | <code>number</code> | Number of seconds after which processing will be stopped |
**Example**
```js
// Ensure processing takes no longer than 3 seconds
try {
const data = await sharp(input)
.blur(1000)
.timeout({ seconds: 3 })
.toBuffer();
} catch (err) {
if (err.message.includes('timeout')) { ... }
}
```

View File

@@ -0,0 +1,320 @@
---
title: Resizing images
---
## resize
> resize([width], [height], [options]) ⇒ <code>Sharp</code>
Resize image to `width`, `height` or `width x height`.
When both a `width` and `height` are provided, the possible methods by which the image should **fit** these are:
- `cover`: (default) Preserving aspect ratio, attempt to ensure the image covers both provided dimensions by cropping/clipping to fit.
- `contain`: Preserving aspect ratio, contain within both provided dimensions using "letterboxing" where necessary.
- `fill`: Ignore the aspect ratio of the input and stretch to both provided dimensions.
- `inside`: Preserving aspect ratio, resize the image to be as large as possible while ensuring its dimensions are less than or equal to both those specified.
- `outside`: Preserving aspect ratio, resize the image to be as small as possible while ensuring its dimensions are greater than or equal to both those specified.
Some of these values are based on the [object-fit](https://developer.mozilla.org/en-US/docs/Web/CSS/object-fit) CSS property.
<img alt="Examples of various values for the fit property when resizing" width="100%" style="aspect-ratio: 998/243" src="https://cdn.jsdelivr.net/gh/lovell/sharp@main/docs/image/api-resize-fit.svg">
When using a **fit** of `cover` or `contain`, the default **position** is `centre`. Other options are:
- `sharp.position`: `top`, `right top`, `right`, `right bottom`, `bottom`, `left bottom`, `left`, `left top`.
- `sharp.gravity`: `north`, `northeast`, `east`, `southeast`, `south`, `southwest`, `west`, `northwest`, `center` or `centre`.
- `sharp.strategy`: `cover` only, dynamically crop using either the `entropy` or `attention` strategy.
Some of these values are based on the [object-position](https://developer.mozilla.org/en-US/docs/Web/CSS/object-position) CSS property.
The strategy-based approach initially resizes so one dimension is at its target length
then repeatedly ranks edge regions, discarding the edge with the lowest score based on the selected strategy.
- `entropy`: focus on the region with the highest [Shannon entropy](https://en.wikipedia.org/wiki/Entropy_%28information_theory%29).
- `attention`: focus on the region with the highest luminance frequency, colour saturation and presence of skin tones.
Possible downsizing kernels are:
- `nearest`: Use [nearest neighbour interpolation](http://en.wikipedia.org/wiki/Nearest-neighbor_interpolation).
- `linear`: Use a [triangle filter](https://en.wikipedia.org/wiki/Triangular_function).
- `cubic`: Use a [Catmull-Rom spline](https://en.wikipedia.org/wiki/Centripetal_Catmull%E2%80%93Rom_spline).
- `mitchell`: Use a [Mitchell-Netravali spline](https://www.cs.utexas.edu/~fussell/courses/cs384g-fall2013/lectures/mitchell/Mitchell.pdf).
- `lanczos2`: Use a [Lanczos kernel](https://en.wikipedia.org/wiki/Lanczos_resampling#Lanczos_kernel) with `a=2`.
- `lanczos3`: Use a Lanczos kernel with `a=3` (the default).
When upsampling, these kernels map to `nearest`, `linear` and `cubic` interpolators.
Downsampling kernels without a matching upsampling interpolator map to `cubic`.
Only one resize can occur per pipeline.
Previous calls to `resize` in the same pipeline will be ignored.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [width] | <code>number</code> | | How many pixels wide the resultant image should be. Use `null` or `undefined` to auto-scale the width to match the height. |
| [height] | <code>number</code> | | How many pixels high the resultant image should be. Use `null` or `undefined` to auto-scale the height to match the width. |
| [options] | <code>Object</code> | | |
| [options.width] | <code>number</code> | | An alternative means of specifying `width`. If both are present this takes priority. |
| [options.height] | <code>number</code> | | An alternative means of specifying `height`. If both are present this takes priority. |
| [options.fit] | <code>String</code> | <code>&#x27;cover&#x27;</code> | How the image should be resized/cropped to fit the target dimension(s), one of `cover`, `contain`, `fill`, `inside` or `outside`. |
| [options.position] | <code>String</code> | <code>&#x27;centre&#x27;</code> | A position, gravity or strategy to use when `fit` is `cover` or `contain`. |
| [options.background] | <code>String</code> \| <code>Object</code> | <code>{r: 0, g: 0, b: 0, alpha: 1}</code> | background colour when `fit` is `contain`, parsed by the [color](https://www.npmjs.org/package/color) module, defaults to black without transparency. |
| [options.kernel] | <code>String</code> | <code>&#x27;lanczos3&#x27;</code> | The kernel to use for image reduction and the inferred interpolator to use for upsampling. Use the `fastShrinkOnLoad` option to control kernel vs shrink-on-load. |
| [options.withoutEnlargement] | <code>Boolean</code> | <code>false</code> | Do not scale up if the width *or* height are already less than the target dimensions, equivalent to GraphicsMagick's `>` geometry option. This may result in output dimensions smaller than the target dimensions. |
| [options.withoutReduction] | <code>Boolean</code> | <code>false</code> | Do not scale down if the width *or* height are already greater than the target dimensions, equivalent to GraphicsMagick's `<` geometry option. This may still result in a crop to reach the target dimensions. |
| [options.fastShrinkOnLoad] | <code>Boolean</code> | <code>true</code> | Take greater advantage of the JPEG and WebP shrink-on-load feature, which can lead to a slight moiré pattern or round-down of an auto-scaled dimension. |
**Example**
```js
sharp(input)
.resize({ width: 100 })
.toBuffer()
.then(data => {
// 100 pixels wide, auto-scaled height
});
```
**Example**
```js
sharp(input)
.resize({ height: 100 })
.toBuffer()
.then(data => {
// 100 pixels high, auto-scaled width
});
```
**Example**
```js
sharp(input)
.resize(200, 300, {
kernel: sharp.kernel.nearest,
fit: 'contain',
position: 'right top',
background: { r: 255, g: 255, b: 255, alpha: 0.5 }
})
.toFile('output.png')
.then(() => {
// output.png is a 200 pixels wide and 300 pixels high image
// containing a nearest-neighbour scaled version
// contained within the north-east corner of a semi-transparent white canvas
});
```
**Example**
```js
const transformer = sharp()
.resize({
width: 200,
height: 200,
fit: sharp.fit.cover,
position: sharp.strategy.entropy
});
// Read image data from readableStream
// Write 200px square auto-cropped image data to writableStream
readableStream
.pipe(transformer)
.pipe(writableStream);
```
**Example**
```js
sharp(input)
.resize(200, 200, {
fit: sharp.fit.inside,
withoutEnlargement: true
})
.toFormat('jpeg')
.toBuffer()
.then(function(outputBuffer) {
// outputBuffer contains JPEG image data
// no wider and no higher than 200 pixels
// and no larger than the input image
});
```
**Example**
```js
sharp(input)
.resize(200, 200, {
fit: sharp.fit.outside,
withoutReduction: true
})
.toFormat('jpeg')
.toBuffer()
.then(function(outputBuffer) {
// outputBuffer contains JPEG image data
// of at least 200 pixels wide and 200 pixels high while maintaining aspect ratio
// and no smaller than the input image
});
```
**Example**
```js
const scaleByHalf = await sharp(input)
.metadata()
.then(({ width }) => sharp(input)
.resize(Math.round(width * 0.5))
.toBuffer()
);
```
## extend
> extend(extend) ⇒ <code>Sharp</code>
Extend / pad / extrude one or more edges of the image with either
the provided background colour or pixels derived from the image.
This operation will always occur after resizing and extraction, if any.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| extend | <code>number</code> \| <code>Object</code> | | single pixel count to add to all edges or an Object with per-edge counts |
| [extend.top] | <code>number</code> | <code>0</code> | |
| [extend.left] | <code>number</code> | <code>0</code> | |
| [extend.bottom] | <code>number</code> | <code>0</code> | |
| [extend.right] | <code>number</code> | <code>0</code> | |
| [extend.extendWith] | <code>String</code> | <code>&#x27;background&#x27;</code> | populate new pixels using this method, one of: background, copy, repeat, mirror. |
| [extend.background] | <code>String</code> \| <code>Object</code> | <code>{r: 0, g: 0, b: 0, alpha: 1}</code> | background colour, parsed by the [color](https://www.npmjs.org/package/color) module, defaults to black without transparency. |
**Example**
```js
// Resize to 140 pixels wide, then add 10 transparent pixels
// to the top, left and right edges and 20 to the bottom edge
sharp(input)
.resize(140)
.extend({
top: 10,
bottom: 20,
left: 10,
right: 10,
background: { r: 0, g: 0, b: 0, alpha: 0 }
})
...
```
**Example**
```js
// Add a row of 10 red pixels to the bottom
sharp(input)
.extend({
bottom: 10,
background: 'red'
})
...
```
**Example**
```js
// Extrude image by 8 pixels to the right, mirroring existing right hand edge
sharp(input)
.extend({
right: 8,
background: 'mirror'
})
...
```
## extract
> extract(options) ⇒ <code>Sharp</code>
Extract/crop a region of the image.
- Use `extract` before `resize` for pre-resize extraction.
- Use `extract` after `resize` for post-resize extraction.
- Use `extract` twice and `resize` once for extract-then-resize-then-extract in a fixed operation order.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Description |
| --- | --- | --- |
| options | <code>Object</code> | describes the region to extract using integral pixel values |
| options.left | <code>number</code> | zero-indexed offset from left edge |
| options.top | <code>number</code> | zero-indexed offset from top edge |
| options.width | <code>number</code> | width of region to extract |
| options.height | <code>number</code> | height of region to extract |
**Example**
```js
sharp(input)
.extract({ left: left, top: top, width: width, height: height })
.toFile(output, function(err) {
// Extract a region of the input image, saving in the same format.
});
```
**Example**
```js
sharp(input)
.extract({ left: leftOffsetPre, top: topOffsetPre, width: widthPre, height: heightPre })
.resize(width, height)
.extract({ left: leftOffsetPost, top: topOffsetPost, width: widthPost, height: heightPost })
.toFile(output, function(err) {
// Extract a region, resize, then extract from the resized image
});
```
## trim
> trim([options]) ⇒ <code>Sharp</code>
Trim pixels from all edges that contain values similar to the given background colour, which defaults to that of the top-left pixel.
Images with an alpha channel will use the combined bounding box of alpha and non-alpha channels.
If the result of this operation would trim an image to nothing then no change is made.
The `info` response Object will contain `trimOffsetLeft` and `trimOffsetTop` properties.
**Throws**:
- <code>Error</code> Invalid parameters
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> | | |
| [options.background] | <code>string</code> \| <code>Object</code> | <code>&quot;&#x27;top-left pixel&#x27;&quot;</code> | Background colour, parsed by the [color](https://www.npmjs.org/package/color) module, defaults to that of the top-left pixel. |
| [options.threshold] | <code>number</code> | <code>10</code> | Allowed difference from the above colour, a positive number. |
| [options.lineArt] | <code>boolean</code> | <code>false</code> | Does the input more closely resemble line art (e.g. vector) rather than being photographic? |
**Example**
```js
// Trim pixels with a colour similar to that of the top-left pixel.
await sharp(input)
.trim()
.toFile(output);
```
**Example**
```js
// Trim pixels with the exact same colour as that of the top-left pixel.
await sharp(input)
.trim({
threshold: 0
})
.toFile(output);
```
**Example**
```js
// Assume input is line art and trim only pixels with a similar colour to red.
const output = await sharp(input)
.trim({
background: "#FF0000",
lineArt: true
})
.toBuffer();
```
**Example**
```js
// Trim all "yellow-ish" pixels, being more lenient with the higher threshold.
const output = await sharp(input)
.trim({
background: "yellow",
threshold: 42,
})
.toBuffer();
```

View File

@@ -0,0 +1,237 @@
---
title: Global properties
---
## versions
> versions
An Object containing the version numbers of sharp, libvips
and (when using prebuilt binaries) its dependencies.
**Example**
```js
console.log(sharp.versions);
```
## interpolators
> interpolators : <code>enum</code>
An Object containing the available interpolators and their proper values
**Read only**: true
**Properties**
| Name | Type | Default | Description |
| --- | --- | --- | --- |
| nearest | <code>string</code> | <code>&quot;nearest&quot;</code> | [Nearest neighbour interpolation](http://en.wikipedia.org/wiki/Nearest-neighbor_interpolation). Suitable for image enlargement only. |
| bilinear | <code>string</code> | <code>&quot;bilinear&quot;</code> | [Bilinear interpolation](http://en.wikipedia.org/wiki/Bilinear_interpolation). Faster than bicubic but with less smooth results. |
| bicubic | <code>string</code> | <code>&quot;bicubic&quot;</code> | [Bicubic interpolation](http://en.wikipedia.org/wiki/Bicubic_interpolation) (the default). |
| locallyBoundedBicubic | <code>string</code> | <code>&quot;lbb&quot;</code> | [LBB interpolation](https://github.com/libvips/libvips/blob/master/libvips/resample/lbb.cpp#L100). Prevents some "[acutance](http://en.wikipedia.org/wiki/Acutance)" but typically reduces performance by a factor of 2. |
| nohalo | <code>string</code> | <code>&quot;nohalo&quot;</code> | [Nohalo interpolation](http://eprints.soton.ac.uk/268086/). Prevents acutance but typically reduces performance by a factor of 3. |
| vertexSplitQuadraticBasisSpline | <code>string</code> | <code>&quot;vsqbs&quot;</code> | [VSQBS interpolation](https://github.com/libvips/libvips/blob/master/libvips/resample/vsqbs.cpp#L48). Prevents "staircasing" when enlarging. |
## format
> format ⇒ <code>Object</code>
An Object containing nested boolean values representing the available input and output formats/methods.
**Example**
```js
console.log(sharp.format);
```
## queue
> queue
An EventEmitter that emits a `change` event when a task is either:
- queued, waiting for _libuv_ to provide a worker thread
- complete
**Example**
```js
sharp.queue.on('change', function(queueLength) {
console.log('Queue contains ' + queueLength + ' task(s)');
});
```
## cache
> cache([options]) ⇒ <code>Object</code>
Gets or, when options are provided, sets the limits of _libvips'_ operation cache.
Existing entries in the cache will be trimmed after any change in limits.
This method always returns cache statistics,
useful for determining how much working memory is required for a particular task.
| Param | Type | Default | Description |
| --- | --- | --- | --- |
| [options] | <code>Object</code> \| <code>boolean</code> | <code>true</code> | Object with the following attributes, or boolean where true uses default cache settings and false removes all caching |
| [options.memory] | <code>number</code> | <code>50</code> | is the maximum memory in MB to use for this cache |
| [options.files] | <code>number</code> | <code>20</code> | is the maximum number of files to hold open |
| [options.items] | <code>number</code> | <code>100</code> | is the maximum number of operations to cache |
**Example**
```js
const stats = sharp.cache();
```
**Example**
```js
sharp.cache( { items: 200 } );
sharp.cache( { files: 0 } );
sharp.cache(false);
```
## concurrency
> concurrency([concurrency]) ⇒ <code>number</code>
Gets or, when a concurrency is provided, sets
the maximum number of threads _libvips_ should use to process _each image_.
These are from a thread pool managed by glib,
which helps avoid the overhead of creating new threads.
This method always returns the current concurrency.
The default value is the number of CPU cores,
except when using glibc-based Linux without jemalloc,
where the default is `1` to help reduce memory fragmentation.
A value of `0` will reset this to the number of CPU cores.
Some image format libraries spawn additional threads,
e.g. libaom manages its own 4 threads when encoding AVIF images,
and these are independent of the value set here.
The maximum number of images that sharp can process in parallel
is controlled by libuv's `UV_THREADPOOL_SIZE` environment variable,
which defaults to 4.
https://nodejs.org/api/cli.html#uv_threadpool_sizesize
For example, by default, a machine with 8 CPU cores will process
4 images in parallel and use up to 8 threads per image,
so there will be up to 32 concurrent threads.
**Returns**: <code>number</code> - concurrency
| Param | Type |
| --- | --- |
| [concurrency] | <code>number</code> |
**Example**
```js
const threads = sharp.concurrency(); // 4
sharp.concurrency(2); // 2
sharp.concurrency(0); // 4
```
## counters
> counters() ⇒ <code>Object</code>
Provides access to internal task counters.
- queue is the number of tasks this module has queued waiting for _libuv_ to provide a worker thread from its pool.
- process is the number of resize tasks currently being processed.
**Example**
```js
const counters = sharp.counters(); // { queue: 2, process: 4 }
```
## simd
> simd([simd]) ⇒ <code>boolean</code>
Get and set use of SIMD vector unit instructions.
Requires libvips to have been compiled with highway support.
Improves the performance of `resize`, `blur` and `sharpen` operations
by taking advantage of the SIMD vector unit of the CPU, e.g. Intel SSE and ARM NEON.
| Param | Type | Default |
| --- | --- | --- |
| [simd] | <code>boolean</code> | <code>true</code> |
**Example**
```js
const simd = sharp.simd();
// simd is `true` if the runtime use of highway is currently enabled
```
**Example**
```js
const simd = sharp.simd(false);
// prevent libvips from using highway at runtime
```
## block
> block(options)
Block libvips operations at runtime.
This is in addition to the `VIPS_BLOCK_UNTRUSTED` environment variable,
which when set will block all "untrusted" operations.
**Since**: 0.32.4
| Param | Type | Description |
| --- | --- | --- |
| options | <code>Object</code> | |
| options.operation | <code>Array.&lt;string&gt;</code> | List of libvips low-level operation names to block. |
**Example** *(Block all TIFF input.)*
```js
sharp.block({
operation: ['VipsForeignLoadTiff']
});
```
## unblock
> unblock(options)
Unblock libvips operations at runtime.
This is useful for defining a list of allowed operations.
**Since**: 0.32.4
| Param | Type | Description |
| --- | --- | --- |
| options | <code>Object</code> | |
| options.operation | <code>Array.&lt;string&gt;</code> | List of libvips low-level operation names to unblock. |
**Example** *(Block all input except WebP from the filesystem.)*
```js
sharp.block({
operation: ['VipsForeignLoad']
});
sharp.unblock({
operation: ['VipsForeignLoadWebpFile']
});
```
**Example** *(Block all input except JPEG and PNG from a Buffer or Stream.)*
```js
sharp.block({
operation: ['VipsForeignLoad']
});
sharp.unblock({
operation: ['VipsForeignLoadJpegBuffer', 'VipsForeignLoadPngBuffer']
});
```

File diff suppressed because it is too large Load Diff

View File

@@ -0,0 +1,98 @@
---
title: "High performance Node.js image processing"
---
<img src="https://cdn.jsdelivr.net/gh/lovell/sharp@main/docs/image/sharp-logo.svg" width="160" height="160" alt="sharp logo" align="right">
The typical use case for this high speed Node-API module
is to convert large images in common formats to
smaller, web-friendly JPEG, PNG, WebP, GIF and AVIF images of varying dimensions.
It can be used with all JavaScript runtimes
that provide support for Node-API v9, including
Node.js >= 18.17.0, Deno and Bun.
Resizing an image is typically 4x-5x faster than using the
quickest ImageMagick and GraphicsMagick settings
due to its use of [libvips](https://github.com/libvips/libvips).
Colour spaces, embedded ICC profiles and alpha transparency channels are all handled correctly.
Lanczos resampling ensures quality is not sacrificed for speed.
As well as image resizing, operations such as
rotation, extraction, compositing and gamma correction are available.
Most modern macOS, Windows and Linux systems
do not require any additional install or runtime dependencies.
```sh
npm install sharp
```
## Formats
This module supports reading JPEG, PNG, WebP, GIF, AVIF, TIFF and SVG images.
Output images can be in JPEG, PNG, WebP, GIF, AVIF and TIFF formats as well as uncompressed raw pixel data.
Streams, Buffer objects and the filesystem can be used for input and output.
A single input Stream can be split into multiple processing pipelines and output Streams.
Deep Zoom image pyramids can be generated,
suitable for use with "slippy map" tile viewers like
[OpenSeadragon](https://github.com/openseadragon/openseadragon).
## Fast
This module is powered by the blazingly fast
[libvips](https://github.com/libvips/libvips) image processing library,
originally created in 1989 at Birkbeck College
and currently maintained by a small team led by
[John Cupitt](https://github.com/jcupitt).
Only small regions of uncompressed image data
are held in memory and processed at a time,
taking full advantage of multiple CPU cores and L1/L2/L3 cache.
Everything remains non-blocking thanks to _libuv_,
no child processes are spawned and Promises/async/await are supported.
## Optimal
The features of `mozjpeg` and `pngquant` can be used
to optimise the file size of JPEG and PNG images respectively,
without having to invoke separate `imagemin` processes.
Huffman tables are optimised when generating JPEG output images
without having to use separate command line tools like
[jpegoptim](https://github.com/tjko/jpegoptim) and
[jpegtran](http://jpegclub.org/jpegtran/).
PNG filtering is disabled by default,
which for diagrams and line art often produces the same result
as [pngcrush](https://pmt.sourceforge.io/pngcrush/).
The file size of animated GIF output is optimised
without having to use separate command line tools such as
[gifsicle](https://www.lcdf.org/gifsicle/).
## Contributing
A [guide for contributors](https://github.com/lovell/sharp/blob/main/.github/CONTRIBUTING.md)
covers reporting bugs, requesting features and submitting code changes.
## Licensing
Copyright 2013 Lovell Fuller and others.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
[https://www.apache.org/licenses/LICENSE-2.0](https://www.apache.org/licenses/LICENSE-2.0)
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

View File

@@ -0,0 +1,344 @@
---
title: Installation
---
Works with your choice of JavaScript package manager.
:::caution
Please ensure your package manager is configured to install optional dependencies
:::
If a package manager lockfile must support multiple platforms,
please see the [cross-platform](#cross-platform) section
to help decide which package manager is appropriate.
```sh
npm install sharp
```
```sh
pnpm add sharp
```
```sh
yarn add sharp
```
```sh
bun add sharp
```
```sh
deno run --allow-ffi ...
```
## Prerequisites
* Node-API v9 compatible runtime e.g. Node.js ^18.17.0 or >=20.3.0.
## Prebuilt binaries
Ready-compiled sharp and libvips binaries are provided for use on the most common platforms:
* macOS x64 (>= 10.15)
* macOS ARM64
* Linux ARM (glibc >= 2.31)
* Linux ARM64 (glibc >= 2.26, musl >= 1.2.2)
* Linux ppc64 (glibc >= 2.31)
* Linux s390x (glibc >= 2.31)
* Linux x64 (glibc >= 2.26, musl >= 1.2.2, CPU with SSE4.2)
* Windows x64
* Windows x86
This provides support for the
JPEG, PNG, WebP, AVIF (limited to 8-bit depth), TIFF, GIF and SVG (input) image formats.
## Cross-platform
At install time, package managers will automatically select prebuilt binaries
for the current OS platform and CPU architecture, where available.
Some package managers support multiple platforms and architectures
within the same installation tree and/or using the same lockfile.
### npm v10+
:::caution
npm `package-lock.json` files shared by multiple platforms can cause installation problems due to [npm bug #4828](https://github.com/npm/cli/issues/4828)
:::
Provides limited support via `--os`, `--cpu` and `--libc` flags.
To support macOS with Intel x64 and ARM64 CPUs:
```sh
npm install --cpu=x64 --os=darwin sharp
npm install --cpu=arm64 --os=darwin sharp
```
When the cross-target is Linux, the C standard library must be specified.
To support glibc (e.g. Debian) and musl (e.g. Alpine) Linux with Intel x64 CPUs:
```sh
npm install --cpu=x64 --os=linux --libc=glibc sharp
npm install --cpu=x64 --os=linux --libc=musl sharp
```
### yarn v3+
Use the [supportedArchitectures](https://yarnpkg.com/configuration/yarnrc#supportedArchitectures) configuration.
### pnpm v8+
Use the [supportedArchitectures](https://pnpm.io/package_json#pnpmsupportedarchitectures) configuration.
## Custom libvips
To use a custom, globally-installed version of libvips instead of the provided binaries,
make sure it is at least the version listed under `config.libvips` in the `package.json` file
and that it can be located using `pkg-config --modversion vips-cpp`.
For help compiling libvips and its dependencies, please see
[building libvips from source](https://www.libvips.org/install.html#building-libvips-from-source).
The use of a globally-installed libvips is unsupported on Windows
and on macOS when running Node.js under Rosetta.
## Building from source
This module will be compiled from source at `npm install` time when:
* a globally-installed libvips is detected, or
* when the `npm install --build-from-source` flag is used.
The logic to detect a globally-installed libvips can be skipped by setting the
`SHARP_IGNORE_GLOBAL_LIBVIPS` (never try to use it) or
`SHARP_FORCE_GLOBAL_LIBVIPS` (always try to use it, even when missing or outdated)
environment variables.
Building from source requires:
* C++17 compiler
* [node-addon-api](https://www.npmjs.com/package/node-addon-api) version 7+
* [node-gyp](https://github.com/nodejs/node-gyp#installation) version 9+ and its dependencies
There is an install-time check for these dependencies.
If `node-addon-api` or `node-gyp` cannot be found, try adding them via:
```sh
npm install --save node-addon-api node-gyp
```
For cross-compiling, the `--platform`, `--arch` and `--libc` npm flags
(or the `npm_config_platform`, `npm_config_arch` and `npm_config_libc` environment variables)
can be used to configure the target environment.
## WebAssembly
Experimental support is provided for runtime environments that provide
multi-threaded Wasm via Workers.
Use in web browsers is unsupported.
Native text rendering is unsupported.
[Tile-based output](/api-output#tile) is unsupported.
```sh
npm install --cpu=wasm32 sharp
```
## FreeBSD
The `vips` package must be installed before `npm install` is run.
```sh
pkg install -y pkgconf vips
```
```sh
cd /usr/ports/graphics/vips/ && make install clean
```
## Linux memory allocator
The default memory allocator on most glibc-based Linux systems
(e.g. Debian, Red Hat) is unsuitable for long-running, multi-threaded
processes that involve lots of small memory allocations.
For this reason, by default, sharp will limit the use of thread-based
[concurrency](api-utility#concurrency) when the glibc allocator is
detected at runtime.
To help avoid fragmentation and improve performance on these systems,
the use of an alternative memory allocator such as
[jemalloc](https://github.com/jemalloc/jemalloc) is recommended.
Those using musl-based Linux (e.g. Alpine) and non-Linux systems are
unaffected.
## AWS Lambda
The `node_modules` directory of the
[deployment package](https://docs.aws.amazon.com/lambda/latest/dg/nodejs-package.html)
must include binaries for either the linux-x64 or linux-arm64 platforms
depending on the chosen architecture.
When building your deployment package on a machine that differs from the target architecture,
see the [cross-platform](#cross-platform) section to help decide which package manager is appropriate
and how to configure it.
Some package managers use symbolic links
but AWS Lambda does not support these within deployment packages.
To get the best performance select the largest memory available.
A 1536 MB function provides ~12x more CPU time than a 128 MB function.
When integrating with AWS API Gateway, ensure it is configured with the relevant
[binary media types](https://docs.aws.amazon.com/apigateway/latest/developerguide/api-gateway-payload-encodings.html).
## Bundlers
### webpack
Ensure sharp is excluded from bundling via the
[externals](https://webpack.js.org/configuration/externals/)
configuration.
```js
externals: {
'sharp': 'commonjs sharp'
}
```
### esbuild
Ensure sharp is excluded from bundling via the
[external](https://esbuild.github.io/api/#external)
configuration.
```js
buildSync({
entryPoints: ['app.js'],
bundle: true,
platform: 'node',
external: ['sharp'],
})
```
```sh
esbuild app.js --bundle --platform=node --external:sharp
```
For `serverless-esbuild`, ensure platform-specific binaries are installed
via the `serverless.yml` configuration.
```yaml
custom:
esbuild:
external:
- sharp
packagerOptions:
scripts:
- npm install --os=linux --cpu=x64 sharp
```
### electron
#### electron-builder
Ensure `sharp` is unpacked from the ASAR archive file using the
[asarUnpack](https://www.electron.build/app-builder-lib.interface.platformspecificbuildoptions#asarunpack)
option.
```json
{
"build": {
"asar": true,
"asarUnpack": [
"**/node_modules/sharp/**/*",
"**/node_modules/@img/**/*"
]
}
}
```
#### electron-forge
Ensure `sharp` is unpacked from the ASAR archive file using the
[unpack](https://js.electronforge.io/interfaces/_electron_forge_maker_squirrel.InternalOptions.Options.html#asar)
option.
```json
{
"packagerConfig": {
"asar": {
"unpack": "**/node_modules/{sharp,@img}/**/*"
}
}
}
```
### vite
Ensure `sharp` is excluded from bundling via the
[build.rollupOptions](https://vitejs.dev/config/build-options.html)
configuration.
```js
import { defineConfig } from 'vite';
export default defineConfig({
build: {
rollupOptions: {
external: [
"sharp"
]
}
}
});
```
## TypeScript
TypeScript definitions are published as part of
the `sharp` package from v0.32.0.
Previously these were available via the `@types/sharp` package,
which is now deprecated.
When using Typescript, please ensure `devDependencies` includes
the `@types/node` package.
## Fonts
When creating text images or rendering SVG images that contain text elements,
`fontconfig` is used to find the relevant fonts.
On Windows and macOS systems, all system fonts are available for use.
On macOS systems using Homebrew, you may need to set the
`PANGOCAIRO_BACKEND` environment variable to a value of `fontconfig`
to ensure it is used for font discovery instead of Core Text.
On Linux systems, fonts that include the relevant
[`fontconfig` configuration](https://www.freedesktop.org/software/fontconfig/fontconfig-user.html)
when installed via package manager are available for use.
If `fontconfig` configuration is not found, the following error will occur:
```
Fontconfig error: Cannot load default config file
```
In serverless environments where there is no control over font packages,
use the `FONTCONFIG_PATH` environment variable to point to a custom location.
Embedded SVG fonts are unsupported.
## Known conflicts
### Canvas and Windows
If both `canvas` and `sharp` modules are used in the same Windows process, the following error may occur:
```
The specified procedure could not be found.
```

View File

@@ -0,0 +1,113 @@
---
title: Performance
---
A test to benchmark the performance of this module relative to alternatives.
Greater libvips performance can be expected with caching enabled (default)
and using 8+ core machines, especially those with larger L1/L2 CPU caches.
The I/O limits of the relevant (de)compression library will generally determine maximum throughput.
## Contenders
* [jimp](https://www.npmjs.com/package/jimp) v0.22.10 - Image processing in pure JavaScript.
* [imagemagick](https://www.npmjs.com/package/imagemagick) v0.1.3 - Supports filesystem only and "*has been unmaintained for a long time*".
* [gm](https://www.npmjs.com/package/gm) v1.25.0 - Fully featured wrapper around GraphicsMagick's `gm` command line utility.
* [@squoosh/lib](https://www.npmjs.com/package/@squoosh/lib) v0.5.3 - Image libraries transpiled to WebAssembly, includes GPLv3 code, but "*Project no longer maintained*".
* [@squoosh/cli](https://www.npmjs.com/package/@squoosh/cli) v0.7.3 - Command line wrapper around `@squoosh/lib`, avoids GPLv3 by spawning process, but "*Project no longer maintained*".
* sharp v0.33.0 / libvips v8.15.0 - Caching within libvips disabled to ensure a fair comparison.
## Environment
### AMD64
* AWS EC2 us-east-2 [c7a.xlarge](https://aws.amazon.com/ec2/instance-types/c7a/) (4x AMD EPYC 9R14)
* Ubuntu 23.10 [13f233a16be2](https://hub.docker.com/layers/library/ubuntu/23.10/images/sha256-13f233a16be210b57907b98b0d927ceff7571df390701e14fe1f3901b2c4a4d7)
* Node.js 20.10.0
### ARM64
* AWS EC2 us-east-2 [c7g.xlarge](https://aws.amazon.com/ec2/instance-types/c7g/) (4x ARM Graviton3)
* Ubuntu 23.10 [7708743264cb](https://hub.docker.com/layers/library/ubuntu/23.10/images/sha256-7708743264cbb7f6cf7fc13e915faece45a6cdda455748bc55e58e8de3d27b63)
* Node.js 20.10.0
## Task: JPEG
Decompress a 2725x2225 JPEG image,
resize to 720x588 using Lanczos 3 resampling (where available),
then compress to JPEG at a "quality" setting of 80.
Note: jimp does not support Lanczos 3, bicubic resampling used instead.
#### Results: JPEG (AMD64)
| Module | Input | Output | Ops/sec | Speed-up |
| :----------------- | :----- | :----- | ------: | -------: |
| jimp | buffer | buffer | 0.84 | 1.0 |
| squoosh-cli | file | file | 1.54 | 1.8 |
| squoosh-lib | buffer | buffer | 2.24 | 2.7 |
| imagemagick | file | file | 11.75 | 14.0 |
| gm | buffer | buffer | 12.66 | 15.1 |
| gm | file | file | 12.72 | 15.1 |
| sharp | stream | stream | 48.31 | 57.5 |
| sharp | file | file | 51.42 | 61.2 |
| sharp | buffer | buffer | 52.41 | 62.4 |
#### Results: JPEG (ARM64)
| Module | Input | Output | Ops/sec | Speed-up |
| :----------------- | :----- | :----- | ------: | -------: |
| jimp | buffer | buffer | 0.88 | 1.0 |
| squoosh-cli | file | file | 1.18 | 1.3 |
| squoosh-lib | buffer | buffer | 1.99 | 2.3 |
| gm | buffer | buffer | 6.06 | 6.9 |
| gm | file | file | 10.81 | 12.3 |
| imagemagick | file | file | 10.95 | 12.4 |
| sharp | stream | stream | 33.15 | 37.7 |
| sharp | file | file | 34.99 | 39.8 |
| sharp | buffer | buffer | 36.05 | 41.0 |
## Task: PNG
Decompress a 2048x1536 RGBA PNG image,
premultiply the alpha channel,
resize to 720x540 using Lanczos 3 resampling (where available),
unpremultiply then compress as PNG with a "default" zlib compression level of 6
and without adaptive filtering.
Note: jimp does not support premultiply/unpremultiply.
### Results: PNG (AMD64)
| Module | Input | Output | Ops/sec | Speed-up |
| :----------------- | :----- | :----- | ------: | -------: |
| squoosh-cli | file | file | 0.34 | 1.0 |
| squoosh-lib | buffer | buffer | 0.51 | 1.5 |
| jimp | buffer | buffer | 3.59 | 10.6 |
| gm | file | file | 8.54 | 25.1 |
| imagemagick | file | file | 9.23 | 27.1 |
| sharp | file | file | 25.43 | 74.8 |
| sharp | buffer | buffer | 25.70 | 75.6 |
### Results: PNG (ARM64)
| Module | Input | Output | Ops/sec | Speed-up |
| :----------------- | :----- | :----- | ------: | -------: |
| squoosh-cli | file | file | 0.33 | 1.0 |
| squoosh-lib | buffer | buffer | 0.46 | 1.4 |
| jimp | buffer | buffer | 3.51 | 10.6 |
| gm | file | file | 7.47 | 22.6 |
| imagemagick | file | file | 8.06 | 24.4 |
| sharp | file | file | 17.31 | 52.5 |
| sharp | buffer | buffer | 17.66 | 53.5 |
## Running the benchmark test
Requires Docker.
```sh
git clone https://github.com/lovell/sharp.git
cd sharp/test/bench
./run-with-docker.sh
```

View File

@@ -0,0 +1,45 @@
@view-transition {
navigation: auto;
}
:root {
--sl-content-width: 60rem;
--sl-color-accent-low: #072d00;
--sl-color-accent: #247f00;
--sl-color-accent-high: #aad7a0;
--sl-color-white: #ffffff;
--sl-color-gray-1: #eaf0e8;
--sl-color-gray-2: #c5cdc3;
--sl-color-gray-3: #99a796;
--sl-color-gray-4: #4f5c4d;
--sl-color-gray-5: #303c2d;
--sl-color-gray-6: #1f2a1c;
--sl-color-black: #151a13;
}
:root[data-theme="light"] {
--sl-color-accent-low: #c0e2b8;
--sl-color-accent: #165800;
--sl-color-accent-high: #0d3e00;
--sl-color-white: #151a13;
--sl-color-gray-1: #1f2a1c;
--sl-color-gray-2: #303c2d;
--sl-color-gray-3: #4f5c4d;
--sl-color-gray-4: #82907f;
--sl-color-gray-5: #bdc4bb;
--sl-color-gray-6: #eaf0e8;
--sl-color-gray-7: #f4f7f3;
--sl-color-black: #ffffff;
}
blockquote {
background-color: var(--sl-color-gray-6);
padding: 1rem;
}
.site-title::after {
content: "High performance Node.js image processing";
color: var(--sl-color-text);
font-size: var(--sl-text-sm);
padding-top: 0.3rem;
}