kitepdf-svg/io.github.yuroyami.kitepdf.svg/SvgImage

SvgImage

An SVG image, painted as vectors into the shared KiteCanvas rather than rasterised, so it stays crisp at any scale. This is what draws the illustrations, cover art and diagrams that ship inside EPUBs and comics; SvgDocument wraps it when the .svg file IS the document.

val svg = SvgImage.parse(bytes) ?: return
svg.render(canvas, KiteMatrix.scaling(2.0, 2.0))

Drawn: <svg> (width/height/viewBox), <g>, <use>, <path> (every d command including elliptical arcs), <rect> (+ rx/ry), <circle>, <ellipse>, <line>, <polyline>, <polygon>, <text> and <image>; fill, stroke, stroke-width, opacity, fill-opacity, stroke-opacity, fill-rule, display and visibility with inheritance; transform (translate/scale/rotate/skewX/skewY/matrix); linear and radial gradients and <pattern> tiles as paint; clip-path; and mask, by luminance or by alpha. Embedded <style> rules support type, universal, class and ID selectors, their compounds and comma lists, with specificity, source order and !important (SVG 1.1, section 6). Combinators, attribute/pseudo selectors, CSS escapes, at-rules and external stylesheets are not interpreted.

Text is measured against standard-font metrics and drawn through a host typeface, because SVG ships no font file of its own.

filter draws through KiteCanvas.rasterStep: a <filter> with any of the seventeen primitives of Filter Effects 1, in linearRGB or sRGB, and the filter functions such as blur() and drop-shadow(). A canvas without raster steps draws the element unfiltered.

Not drawn: animation.

Types

Companion

object Companion

Properties

hasIntrinsicSize

val hasIntrinsicSize: Boolean

False when the root gives neither a width nor a height, only a viewBox: the image then has an intrinsic ratio and no intrinsic size, and a layout gives it the width of its box (CSS 2.1, 10.3.2, #569). True when width and height are a size of its own.

height

val height: Double

width

val width: Double

Intrinsic size in px: from width and height, a missing one from the other and the viewBox's ratio, else the viewBox extent, else 300x150.

Functions

render

fun render(canvas: KiteCanvas, ctm: KiteMatrix, loadResource: (String) -> ByteArray?? = null)

Paint the SVG into canvas; ctm maps the (0,0)-(width,height) viewport to device space.

textContent

fun textContent(ctm: KiteMatrix): SvgTextContent

The text this image draws, where render with ctm draws it (#523): its lines for search and selection, and where the text of each element with an id is. Text inside a pattern tile, a mask or a clip path is not the image's text, and text that a clip hides is left out.