mirror of
https://github.com/DorkSquadRR/DorkNet.git
synced 2026-10-26 15:38:45 -07:00
Alexa
4edf25b2d5
Targets Rec Room 2020.12.18 client.
Re-import after adding:
- DorkNet.ClientMod/dorknet-clientmod.json.template — config schema
the launcher's manifest-driven patcher renders into UserData/.
- tools/build-patcher-zip.ps1 — produces the dorknet-clientpatch-{key}.zip
Release asset the launcher fetches.
- .github/workflows/release.yml — auto-publishes both Release assets
on v1-december-* tag push.
Scrubbed via maintainer-side scrub-for-public.ps1: same drop rules as
the prior import, plus build-patcher-zip.ps1 added to the tools/
allow-list so it survives the scrub.
66 lines
2.9 KiB
C#
66 lines
2.9 KiB
C#
using SkiaSharp;
|
|
|
|
namespace DorkNet.Server.Services;
|
|
|
|
/// <summary>
|
|
/// img.* CDN transforms. The 2020 watch posts a full-frame capture
|
|
/// (e.g. the avatar standing in their dorm) and then re-requests the
|
|
/// same blob with <c>?cropSquare=1&width=256</c> on the profile
|
|
/// thumbnail surface. We honour <c>cropSquare</c> server-side so the
|
|
/// profile pic is face-zoomed, but we deliberately ignore the watch's
|
|
/// <c>width=256</c> hint and return the cropped image at the source's
|
|
/// native resolution: the watch's UI renders the same blob across
|
|
/// multiple sizes (small thumb chip and the big profile frame), and a
|
|
/// pre-shrunk 256² source gets stretched across a ~500px UI element and
|
|
/// looks blurry. Keeping the source resolution lets the watch's image
|
|
/// pipeline downsample for display without losing detail.
|
|
/// </summary>
|
|
public static class ImageTransformService
|
|
{
|
|
public readonly record struct Result(byte[] Bytes, string ContentType);
|
|
|
|
public static Result? TryTransform(
|
|
byte[] sourceBytes,
|
|
string sourceContentType,
|
|
bool cropSquare,
|
|
int? widthHint)
|
|
{
|
|
// widthHint is intentionally unused — see class comment. Kept on the
|
|
// signature so the controller can pass the parsed value without the
|
|
// call site having to know which params we currently respect.
|
|
_ = widthHint;
|
|
if (!cropSquare) return null;
|
|
if (sourceBytes.Length == 0) return null;
|
|
|
|
using var input = SKBitmap.Decode(sourceBytes);
|
|
if (input is null) return null;
|
|
if (input.Width == input.Height) return null; // already square, pass through
|
|
|
|
using var cropped = CenterCropSquare(input);
|
|
using var image = SKImage.FromBitmap(cropped);
|
|
|
|
// Re-encode in the same family the source advertised, so an uploaded
|
|
// JPEG stays JPEG (size budget for profile pic round-trips) and
|
|
// anything PNG-shaped stays PNG (alpha preserved for UI overlays).
|
|
var (format, contentType, quality) = sourceContentType switch
|
|
{
|
|
"image/jpeg" or "image/jpg" => (SKEncodedImageFormat.Jpeg, "image/jpeg", 92),
|
|
"image/webp" => (SKEncodedImageFormat.Webp, "image/webp", 92),
|
|
_ => (SKEncodedImageFormat.Png, "image/png", 100),
|
|
};
|
|
using var encoded = image.Encode(format, quality);
|
|
return new Result(encoded.ToArray(), contentType);
|
|
}
|
|
|
|
private static SKBitmap CenterCropSquare(SKBitmap input)
|
|
{
|
|
var side = Math.Min(input.Width, input.Height);
|
|
var x = (input.Width - side) / 2;
|
|
var y = (input.Height - side) / 2;
|
|
var dst = new SKBitmap(new SKImageInfo(side, side, input.ColorType, input.AlphaType));
|
|
using var canvas = new SKCanvas(dst);
|
|
canvas.DrawBitmap(input, new SKRect(x, y, x + side, y + side), new SKRect(0, 0, side, side));
|
|
return dst;
|
|
}
|
|
}
|