Stitch & collage

await painter.image.stitchImages(images, options?) concatenates images horizontally, vertically, or on a square-ish grid. await painter.image.createCollage(images, layout) lays out sized cells as a grid, shortest-column masonry, or carousel with backgrounds and optional rounded clips.

Implementation: lib-next/output/stitch.ts (PNG output through canvas.toBuffer('image/png')). Image sources use the shared Apexify image loading/preflight path and may be a local/remote string source or Buffer.

Both helpers bound source loading by the central batch-concurrency limit and validate final canvas dimensions before allocation.

#stitchImages(images, options?)

##StitchOptions (lib-next/types/batch.ts)

##Layout rules

• Horizontal / vertical: canvas spans the sum of tile extents minus overlap plus spacing; the orthogonal extent is the maximum input edge.
• Grid: cols = ceil(sqrt(n)), rows = ceil(n / cols); every cell uses the maximum input width/height. Inputs keep their intrinsic dimensions and are anchored to the top-left of each cell.
• A combination of overlap/spacing that would produce a non-positive output dimension is rejected before canvas allocation.

1const strip = await painter.image.stitchImages(['./a.png', './b.png'], {2  direction: 'horizontal',3  spacing: 8,4});5 6const sheet = await painter.image.stitchImages(urls, { direction: 'grid' });

#createCollage(images, layout)

##Images argument

1Array<{ source: string | Buffer; width?: number; height?: number }>;

Optional width / height control the draw size and default to the decoded raster dimensions. Positive integer dimensions are required when supplied.

##CollageLayout

##Layout behaviour

1const board = await painter.image.createCollage(2  [3    { source: './photo-a.jpg', width: 320, height: 240 },4    { source: './photo-b.jpg' },5  ],6  {7    type: 'grid',8    columns: 2,9    rows: 2,10    spacing: 12,11    background: '#0f172a',12    borderRadius: 8,13  }14);