meo-canvas - v9.0.2
    Preparing search index...

    Interface BoxNode

    Base node class for rendering rectangular boxes with layout, styling, and children. It uses the Yoga layout engine for positioning and sizing.

    interface BoxNode {
        backgroundBitmap: Image | null;
        children: BoxNode[];
        initialProps: Partial<BoxProps>;
        key?: string;
        name?: string;
        node: Node;
        placedByLayout: boolean;
        props: BoxProps & BaseProps;
        _renderContent(
            ctx: CanvasRenderingContext2D,
            x: number,
            y: number,
            width: number,
            height: number,
        ): Promise<void>;
        appendChild(child: BoxNode, index: number): void;
        applyDefaults(): void;
        blendMode(): string;
        filterChain(): string;
        finalizeLayout(): boolean;
        inkOverflow(): number;
        loadBackgroundImage(
            cache?: RenderImageCache,
            diskCacheKeys?: Set<string>,
        ): Promise<void>;
        processInitialChildren(): void;
        render(
            ctx: CanvasRenderingContext2D,
            offsetX?: number,
            offsetY?: number,
            suppressStacking?: boolean,
        ): Promise<void>;
        resolvedDirection(): Direction;
        resolveInheritedStyles(parentProps: BoxProps & BaseProps): void;
        setLayout(props: BoxProps): void;
        updateLayoutBasedOnComputedSize(): void;
    }

    Hierarchy (View Summary)

    Index
    backgroundBitmap: Image | null = null

    The decoded background picture, once the render's load pass has fetched it.

    children: BoxNode[]

    Child nodes.

    initialProps: Partial<BoxProps>

    Original props passed to the constructor before any modifications.

    key?: string

    Unique node identifier.

    name?: string

    Node type name.

    node: Node

    The Yoga layout engine node.

    placedByLayout: boolean = false

    Set where a layout places this node absolutely for its own reasons rather than because the caller asked. Grid does it to put items on their tracks, and a node placed that way becomes a containing block Yoga resolves against though nothing named a positionType -- CSS keeps a grid and its items static, so resolveContainingBlocks places an absolute descendant against the box above it that CSS does name instead.

    Current props including defaults and inherited values.

    • Renders the node's visual content including background fills, shadows, and borders. This is an internal method used by the render() pipeline.

      Parameters

      • ctx: CanvasRenderingContext2D

        The skia-canvas 2D rendering context to draw into

      • x: number

        The absolute x-coordinate where drawing should begin

      • y: number

        The absolute y-coordinate where drawing should begin

      • width: number

        The width of the content area to render

      • height: number

        The height of the content area to render

      Returns Promise<void>

    • Appends a child node at the specified index.

      Parameters

      • child: BoxNode

        Child node to append.

      • index: number

        Index to insert child at

      Returns void

    • The blend mode this node composites with, or an empty string for the ordinary source-over.

      normal is the default and means exactly source-over, so it is not worth an offscreen.

      Returns string

    • The CSS filter chain this node draws through, or an empty string for none.

      saturate came first and stays a shorthand for the same machinery, so it leads the chain and filter follows — the order they would appear in if the shorthand were written out.

      Returns string

    • Performs final layout adjustments recursively after the main layout calculation.

      Returns boolean

      Whether any node was marked as dirty during finalization.

    • How far this node's own drawing reaches past its box, beyond what its box model accounts for.

      Zero for a box, which paints within its bounds; TextNode overrides it, since a line too long for its box still draws unless something clips it. Used to size a group's offscreen, where anything unaccounted for is cut off at a hard edge.

      Returns number

    • Fetches this node's background picture, if it has one.

      Runs in the same pass as the images, before layout, and through the same cache — so a picture used as one node's background and another's image is fetched once. A failure leaves the node without a picture rather than failing the render, which is what a missing background should do.

      Parameters

      Returns Promise<void>

    • Draws this node, through its mask when it has one.

      Every component's drawing arrives here — Text, Image, Chart and Grid override _renderContent rather than this — so masking one node type means masking all of them.

      The two kinds of mask are applied differently because they are different operations. A shape or path clips, which is a yes-or-no test per pixel and costs nothing but a save/restore. A gradient cannot: its whole purpose is the answers in between, so the node is composited through one instead. Both are applied to the node's layout box, before its own transform, which is what keeps the two consistent with each other.

      dither is held on the context around all of it, so the node's mask, background, content and children are drawn with one answer and whatever draws next is left with its own.

      Parameters

      Returns Promise<void>

    • The direction this node ends up laid out in.

      Yoga resolves Inherit while it lays out but keeps no record of the answer: getDirection hands back what was set, and there is no computed counterpart. So the chain is walked to the first ancestor that named one, which is what Yoga itself would have found.

      Returns Direction

    • Hook for subclasses to update layout based on computed size.

      Returns void