Files
SnapOtter/apps/api/src/routes/tools/background-replace.ts
T

238 lines
8.5 KiB
TypeScript

import { randomUUID } from "node:crypto";
import { removeBackground } from "@snapotter/ai";
import { getBundleForTool, TOOL_BUNDLE_MAP } from "@snapotter/shared";
import type { FastifyInstance, FastifyReply, FastifyRequest } from "fastify";
import sharp from "sharp";
import { z } from "zod";
import { registerAiJobHandler } from "../../jobs/ai-handlers.js";
import { enqueueToolJob } from "../../jobs/enqueue.js";
import { autoOrient } from "../../lib/auto-orient.js";
import { compositeOnColor, createGradientBackground } from "../../lib/bg-effects.js";
import { formatZodErrors, stripInternalPaths } from "../../lib/errors.js";
import { isToolInstalled } from "../../lib/feature-status.js";
import { validateImageBuffer } from "../../lib/file-validation.js";
import { decodeToSharpCompat, needsCliDecode } from "../../lib/format-decoders.js";
import { decodeHeic } from "../../lib/heic-converter.js";
import { receiveUpload } from "../../lib/upload-stream.js";
import { getAuthUser } from "../../plugins/auth.js";
const HEX_RE = /^#[0-9a-fA-F]{6}$/;
const settingsSchema = z.object({
backgroundType: z.enum(["color", "gradient"]).default("color"),
color: z.string().regex(HEX_RE).default("#ffffff"),
gradientColor1: z.string().regex(HEX_RE).optional(),
gradientColor2: z.string().regex(HEX_RE).optional(),
gradientAngle: z.number().int().min(0).max(360).default(180),
feather: z.number().int().min(0).max(20).default(0),
format: z.enum(["png", "webp"]).default("png"),
});
/**
* Soften the alpha edges of a subject PNG by blurring its alpha channel.
* Keeps RGB intact; only the transparency boundary gets smoothed.
*/
async function featherEdges(subjectBuffer: Buffer, radius: number): Promise<Buffer> {
// Read the subject as raw RGBA plus a separately-blurred copy of its alpha,
// then overwrite the alpha channel in place. joinChannel does not reliably
// re-tag the merged channel as alpha, so we splice the raw bytes directly.
const { data: rgba, info } = await sharp(subjectBuffer)
.ensureAlpha()
.raw()
.toBuffer({ resolveWithObject: true });
const { data: blurredAlpha } = await sharp(subjectBuffer)
.extractChannel(3)
.blur(radius)
.raw()
.toBuffer({ resolveWithObject: true });
for (let p = 3, a = 0; p < rgba.length; p += 4, a++) {
rgba[p] = blurredAlpha[a];
}
return sharp(rgba, { raw: { width: info.width, height: info.height, channels: 4 } })
.png()
.toBuffer();
}
// -- AI job handler (runs inside the BullMQ worker) --
registerAiJobHandler("background-replace", async (input, data, ctx) => {
const settings = settingsSchema.parse(data.settings);
ctx.report(5, "Removing background");
const subjectPng = await removeBackground(input, ctx.scratchDir, {}, (percent, stage) => {
// Scale rembg progress into 5..80 range
const scaled = 5 + Math.round(percent * 0.75);
ctx.report(Math.min(scaled, 80), stage);
});
// Feather alpha edges before compositing
let subject = subjectPng;
if (settings.feather > 0) {
ctx.report(82, "Feathering edges");
subject = await featherEdges(subjectPng, settings.feather);
}
ctx.report(85, "Compositing background");
let composited: Buffer;
if (settings.backgroundType === "gradient") {
const meta = await sharp(subject).metadata();
if (!meta.width || !meta.height) throw new Error("Cannot read subject dimensions");
const gradBg = await createGradientBackground(
meta.width,
meta.height,
settings.gradientColor1 ?? "#ffffff",
settings.gradientColor2 ?? "#000000",
settings.gradientAngle,
);
composited = await sharp(gradBg)
.composite([{ input: subject, blend: "over" }])
.png()
.toBuffer();
} else {
composited = await compositeOnColor(subject, settings.color);
}
// Encode to requested output format
const fmt = settings.format;
const result =
fmt === "webp" ? await sharp(composited).webp({ lossless: true }).toBuffer() : composited;
const base = data.filename.replace(/\.[^.]+$/, "");
const outName = `${base}_bg.${fmt}`;
return {
buffer: result,
filename: outName,
contentType: fmt === "webp" ? "image/webp" : "image/png",
};
});
export function registerBackgroundReplace(app: FastifyInstance) {
app.post(
"/api/v1/tools/image/background-replace",
async (request: FastifyRequest, reply: FastifyReply) => {
const toolId = "background-replace";
if (!isToolInstalled(toolId)) {
const bundle = getBundleForTool(toolId);
return reply.status(501).send({
error: "Feature not installed",
code: "FEATURE_NOT_INSTALLED",
feature: TOOL_BUNDLE_MAP[toolId],
featureName: bundle?.name ?? toolId,
estimatedSize: bundle?.estimatedSize ?? "unknown",
});
}
const userId = getAuthUser(request)?.id ?? null;
const jobId = randomUUID();
let fileBuffer: Buffer | null = null;
let filename = "image";
let settingsRaw: string | null = null;
let clientJobId: string | null = null;
let fileId: string | null = null;
let inputKey: string | null = null;
try {
const parts = request.parts();
for await (const part of parts) {
if (part.type === "file") {
const upload = await receiveUpload(part, jobId);
inputKey = upload.key;
filename = upload.filename;
} else if (part.fieldname === "settings") {
settingsRaw = part.value as string;
} else if (part.fieldname === "clientJobId") {
const raw = part.value as string;
if (/^[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}$/i.test(raw)) {
clientJobId = raw;
}
} else if (part.fieldname === "fileId") {
fileId = part.value as string;
}
}
} catch (err) {
return reply.status(400).send({
error: "Failed to parse multipart request",
details: stripInternalPaths(err instanceof Error ? err.message : String(err)),
});
}
if (!inputKey) {
return reply.status(400).send({ error: "No image file provided" });
}
const { getObjectBuffer, putObject } = await import("../../lib/object-storage.js");
fileBuffer = await getObjectBuffer(inputKey);
if (!fileBuffer || fileBuffer.length === 0) {
return reply.status(400).send({ error: "No image file provided" });
}
const validation = await validateImageBuffer(fileBuffer, filename);
if (!validation.valid) {
return reply.status(400).send({ error: `Invalid image: ${validation.reason}` });
}
let settings: z.infer<typeof settingsSchema>;
try {
const parsed = settingsRaw ? JSON.parse(settingsRaw) : {};
const result = settingsSchema.safeParse(parsed);
if (!result.success) {
return reply
.status(400)
.send({ error: "Invalid settings", details: formatZodErrors(result.error.issues) });
}
settings = result.data;
} catch {
return reply.status(400).send({ error: "Settings must be valid JSON" });
}
try {
if (validation.format === "heif") {
fileBuffer = await decodeHeic(fileBuffer);
const ext = filename.match(/\.[^.]+$/)?.[0];
if (ext) filename = `${filename.slice(0, -ext.length)}.png`;
}
if (needsCliDecode(validation.format)) {
fileBuffer = await decodeToSharpCompat(fileBuffer, validation.format);
const ext = filename.match(/\.[^.]+$/)?.[0];
if (ext) filename = `${filename.slice(0, -ext.length)}.png`;
}
fileBuffer = await autoOrient(fileBuffer);
} catch (err) {
request.log.error({ err, toolId }, "Input decoding failed");
return reply.status(422).send({
error: "Processing failed",
details: stripInternalPaths(err instanceof Error ? err.message : "Unknown error"),
});
}
const decodedKey = `uploads/${jobId}/${filename}`;
if (decodedKey !== inputKey) {
await putObject(decodedKey, fileBuffer);
inputKey = decodedKey;
} else {
await putObject(inputKey, fileBuffer);
}
const progressJobId = clientJobId || jobId;
await enqueueToolJob({
jobId,
toolId,
userId,
pool: "ai",
inputRefs: [inputKey],
filename,
settings,
clientJobId: clientJobId ?? undefined,
fileId: fileId ?? undefined,
kind: "ai-tool",
});
return reply.status(202).send({ jobId: progressJobId, async: true });
},
);
}