feat(scan): support iX1600 duplex scanning on Linux

agents/fedeo-device-agent/src/api.ts

@@ -1,4 +1,4 @@
-import { readFile } from "node:fs/promises"
+import { readFile, stat } from "node:fs/promises"
 import { basename } from "node:path"
 import { AgentConfig, AgentHeartbeat, NextScanJobResponse, ScanResult } from "./types.js"
 
@@ -16,10 +16,11 @@ export class FedeoApi {
         }
     }
 
-    private async request<T>(path: string, init: RequestInit = {}): Promise<T> {
+    private async request<T>(path: string, init: RequestInit = {}, timeoutMs = 30_000): Promise<T> {
         const response = await fetch(this.url(path), {
             ...init,
             headers: this.headers(init.headers),
+            signal: AbortSignal.timeout(timeoutMs),
         })
 
         if (!response.ok) {
@@ -51,6 +52,9 @@ export class FedeoApi {
     }
 
     async uploadScan(jobId: string, result: ScanResult) {
+        if ((await stat(result.path)).size > 100 * 1024 * 1024) {
+            throw new Error("Die Scan-Datei überschreitet das Upload-Limit von 100 MiB. Bitte den Stapel aufteilen oder die Auflösung reduzieren.")
+        }
         const form = new FormData()
         const fileBuffer = await readFile(result.path)
         const file = new File([fileBuffer], result.filename || basename(result.path), {
@@ -62,6 +66,6 @@ export class FedeoApi {
         return this.request(`/instance-agent/scan-jobs/${jobId}/upload`, {
             method: "POST",
             body: form,
-        })
+        }, 120_000)
     }
 }

agents/fedeo-device-agent/src/commands.ts

@@ -8,7 +8,7 @@ export type CommandResult = {
 
 export const commandExists = (command: string) =>
     new Promise<boolean>((resolve) => {
-        const child = spawn("sh", ["-lc", `command -v ${command}`])
+        const child = spawn("sh", ["-c", 'command -v "$1" >/dev/null 2>&1', "--", command])
         child.on("error", () => resolve(false))
         child.on("close", (code) => resolve(code === 0))
     })
@@ -16,33 +16,50 @@ export const commandExists = (command: string) =>
 export const runCommand = (
     command: string,
     args: string[],
-    options: { timeoutMs?: number } = {}
+    options: { timeoutMs?: number; cwd?: string } = {}
 ) =>
     new Promise<CommandResult>((resolve, reject) => {
         const child = spawn(command, args, {
             stdio: ["ignore", "pipe", "pipe"],
+            cwd: options.cwd,
         })
 
         const stdout: Buffer[] = []
         const stderr: Buffer[] = []
 
+        let timedOut = false
+        let killTimeout: ReturnType<typeof setTimeout> | undefined
         const timeout = options.timeoutMs
             ? setTimeout(() => {
+                timedOut = true
                 child.kill("SIGTERM")
-                reject(new Error(`${command} wurde nach ${options.timeoutMs} ms beendet`))
+                killTimeout = setTimeout(() => child.kill("SIGKILL"), 2000)
             }, options.timeoutMs)
             : null
 
         child.stdout.on("data", (chunk) => stdout.push(Buffer.from(chunk)))
         child.stderr.on("data", (chunk) => stderr.push(Buffer.from(chunk)))
-        child.on("error", reject)
-        child.on("close", (code) => {
+        child.on("error", (error) => {
             if (timeout) clearTimeout(timeout)
+            if (killTimeout) clearTimeout(killTimeout)
+            reject(error)
+        })
+        child.on("close", (code, signal) => {
+            if (timeout) clearTimeout(timeout)
+            if (killTimeout) clearTimeout(killTimeout)
+            if (timedOut) {
+                reject(new Error(`${command} wurde nach ${options.timeoutMs} ms beendet`))
+                return
+            }
+            if (signal) {
+                reject(new Error(`${command} wurde durch ${signal} beendet`))
+                return
+            }
 
             resolve({
                 stdout: Buffer.concat(stdout).toString("utf8"),
                 stderr: Buffer.concat(stderr).toString("utf8"),
-                code: code ?? 0,
+                code: code ?? 1,
             })
         })
     })

agents/fedeo-device-agent/src/config.ts

@@ -60,6 +60,7 @@ export const loadConfig = (): AgentConfig => {
         scanResolution: numberFromEnv(process.env.FEDEO_SCAN_RESOLUTION, 300),
         scanMode: optional(process.env.FEDEO_SCAN_MODE) || "Color",
         scanSource: optional(process.env.FEDEO_SCAN_SOURCE),
+        scanTimeoutSeconds: numberFromEnv(process.env.FEDEO_SCAN_TIMEOUT_SECONDS, 900),
         scanPostprocess: booleanFromEnv(process.env.FEDEO_SCAN_POSTPROCESS, false),
         postprocessProfile: postprocessProfileFromEnv(process.env.FEDEO_SCAN_POSTPROCESS_PROFILE),
         postprocessPython: optional(process.env.FEDEO_SCAN_POSTPROCESS_PYTHON) || defaultPostprocessPython(),

agents/fedeo-device-agent/src/main.ts

@@ -1,10 +1,13 @@
 #!/usr/bin/env node
 import os from "node:os"
+import path from "node:path"
+import { rm } from "node:fs/promises"
 import { FedeoApi } from "./api.js"
 import { loadConfig } from "./config.js"
 import { log } from "./logger.js"
 import { listPrinters } from "./print/cups.js"
 import { hasSane, listScanners, runScan } from "./scan/sane.js"
+import { AgentHeartbeat } from "./types.js"
 
 const sleep = (ms: number) => new Promise((resolve) => setTimeout(resolve, ms))
 
@@ -25,11 +28,14 @@ const main = async () => {
 
     while (true) {
         try {
-            const scannerNames = await listScanners()
-            const printerNames = await listPrinters()
+            const scannerNames = await listScanners().catch((error) => {
+                log.warn("Scanner-Erkennung fehlgeschlagen", stringifyError(error))
+                return []
+            })
+            const printerNames = await listPrinters().catch(() => [])
             const scanAvailable = await hasSane()
 
-            const heartbeat = await api.heartbeat({
+            const heartbeatPayload: AgentHeartbeat = {
                 capabilities: {
                     scan: scanAvailable,
                     print: printerNames.length > 0,
@@ -44,7 +50,8 @@ const main = async () => {
                     node: process.version,
                     uptimeSeconds: Math.round(os.uptime()),
                 },
-            })
+            }
+            const heartbeat = await api.heartbeat(heartbeatPayload)
 
             if (heartbeat.pendingScanJobs > 0) {
                 log.info(`${heartbeat.pendingScanJobs} Scan-Auftrag/Aufträge warten`)
@@ -62,14 +69,44 @@ const main = async () => {
                 scannerName: next.job.scannerName || config.scannerName || "default",
             })
 
+            // Keep online status fresh during long scans without probing an occupied USB scanner.
+            let heartbeatRunning = false
+            const heartbeatTimer = setInterval(async () => {
+                if (heartbeatRunning) return
+                heartbeatRunning = true
+                try {
+                    await api.heartbeat(heartbeatPayload)
+                } catch (error) {
+                    log.warn("Heartbeat während Scan fehlgeschlagen", stringifyError(error))
+                } finally {
+                    heartbeatRunning = false
+                }
+            }, 30_000)
+
             try {
                 await api.updateScanJobStatus(next.job.id, "running")
-                const scanResult = await runScan(config, next.job)
-                await api.uploadScan(next.job.id, scanResult)
+                const jobId = next.job.id
+                const reportProgress = async (message: string) => {
+                    await api.updateScanJobStatus(jobId, "running", message).catch((error) => {
+                        log.warn("Scan-Status konnte nicht übertragen werden", stringifyError(error))
+                    })
+                }
+                const scanResult = await runScan(config, next.job, reportProgress)
+                await reportProgress(`${scanResult.pageCount} Seiten gescannt. ${scanResult.warnings.join(" ")}`.trim())
+                try {
+                    await api.uploadScan(jobId, scanResult)
+                } catch (error) {
+                    throw new Error(`${stringifyError(error)} Scan-Datei bleibt zur Wiederherstellung unter ${scanResult.path}.`)
+                }
+
+                await rm(path.dirname(scanResult.path), { recursive: true, force: true }).catch((error) => {
+                    log.warn("Scan gespeichert; temporäre Dateien konnten nicht entfernt werden", stringifyError(error))
+                })
 
                 log.info("Scan-Auftrag abgeschlossen", {
                     jobId: next.job.id,
                     file: scanResult.filename,
+                    pages: scanResult.pageCount,
                 })
             } catch (error) {
                 const message = stringifyError(error)
@@ -79,6 +116,8 @@ const main = async () => {
                 })
 
                 await api.updateScanJobStatus(next.job.id, "failed", message)
+            } finally {
+                clearInterval(heartbeatTimer)
             }
         } catch (error) {
             log.error("Agent-Schleife fehlgeschlagen", stringifyError(error))

agents/fedeo-device-agent/src/scan/postprocess.ts

@@ -1,66 +1,45 @@
 import path from "node:path"
 import { fileURLToPath } from "node:url"
 import { AgentConfig, ScanResult } from "../types.js"
-import { commandExists, runCommand } from "../commands.js"
+import { runCommand } from "../commands.js"
 
-const currentFile = fileURLToPath(import.meta.url)
-const agentRoot = path.resolve(path.dirname(currentFile), "../..")
-const postprocessScript = path.join(agentRoot, "scripts/opencv_postprocess.py")
+const agentRoot = path.resolve(path.dirname(fileURLToPath(import.meta.url)), "../..")
+const documentScript = path.join(agentRoot, "scripts/scan_document.py")
 
-const extensionMimeTypes: Record<string, string> = {
-    ".pdf": "application/pdf",
-    ".png": "image/png",
-    ".tif": "image/tiff",
-    ".tiff": "image/tiff",
-    ".jpg": "image/jpeg",
-    ".jpeg": "image/jpeg",
+export const checkScanRuntime = async (config: AgentConfig, postprocess: boolean) => {
+    const imports = postprocess && config.postprocessStrict ? "import PIL, cv2, numpy" : "import PIL"
+    try {
+        const result = await runCommand(config.postprocessPython, ["-c", imports], { timeoutMs: 10_000 })
+        if (result.code === 0) return
+    } catch {
+        // Report missing runtimes before the scanner consumes any paper.
+    }
+    throw new Error("Python/Pillow (bei strikter Nachbearbeitung zusätzlich OpenCV) fehlt. Bitte npm run setup:opencv ausführen oder python3-pil installieren.")
 }
 
-const ensureOutputExtension = (filename: string, format: AgentConfig["scanFormat"]) => {
-    const ext = path.extname(filename)
-    if (ext) return filename
-    return `${filename}.${format}`
-}
-
-export const hasOpenCvPostprocessRuntime = async (config: AgentConfig) => {
-    if (!await commandExists(config.postprocessPython)) return false
-
-    const result = await runCommand(config.postprocessPython, [
-        "-c",
-        "import cv2, PIL, numpy",
-    ], { timeoutMs: 10_000 })
-
-    return result.code === 0
-}
-
-export const postprocessScan = async (
+export const createScanDocument = async (
     config: AgentConfig,
-    inputPath: string,
-    outputFilename: string,
-    outputFormat: AgentConfig["scanFormat"],
+    pages: string[],
+    outputPath: string,
+    resolution: number,
+    postprocess: boolean,
     profile: AgentConfig["postprocessProfile"]
 ): Promise<ScanResult> => {
-    const filename = ensureOutputExtension(outputFilename, outputFormat)
-    const outputPath = path.join(config.workDir, filename)
+    const args = [documentScript, "--output", outputPath, "--resolution", String(resolution), "--profile", profile]
+    if (postprocess) args.push("--postprocess")
+    if (config.postprocessStrict) args.push("--strict")
+    args.push("--inputs", ...pages)
 
-    const result = await runCommand(config.postprocessPython, [
-        postprocessScript,
-        "--input",
-        inputPath,
-        "--output",
-        outputPath,
-        "--profile",
-        profile,
-    ], { timeoutMs: 5 * 60 * 1000 })
+    const result = await runCommand(config.postprocessPython, args, {
+        timeoutMs: config.scanTimeoutSeconds * 1000,
+    })
+    if (result.code !== 0) throw new Error(result.stderr || "Scan-Datei konnte nicht erstellt werden")
 
-    if (result.code !== 0) {
-        throw new Error(result.stderr || `OpenCV-Nachbearbeitung wurde mit Code ${result.code} beendet`)
-    }
-
-    const extension = path.extname(outputPath).toLowerCase()
+    const details = JSON.parse(result.stdout) as { pageCount: number; warnings: string[] }
     return {
         path: outputPath,
-        filename,
-        mimeType: extensionMimeTypes[extension] || "application/octet-stream",
+        filename: path.basename(outputPath),
+        mimeType: outputPath.endsWith(".pdf") ? "application/pdf" : outputPath.endsWith(".png") ? "image/png" : "image/tiff",
+        ...details,
     }
 }

agents/fedeo-device-agent/src/scan/sane.ts

@@ -1,149 +1,118 @@
-import { mkdirSync } from "node:fs"
+import { mkdir, mkdtemp, readdir, stat } from "node:fs/promises"
 import path from "node:path"
 import { AgentConfig, ScanJob, ScanResult } from "../types.js"
 import { commandExists, runCommand } from "../commands.js"
-import { hasOpenCvPostprocessRuntime, postprocessScan } from "./postprocess.js"
-import { log } from "../logger.js"
+import { checkScanRuntime, createScanDocument } from "./postprocess.js"
 
-const mimeTypes = {
-    pdf: "application/pdf",
-    png: "image/png",
-    tiff: "image/tiff",
-}
-
-const stringSetting = (settings: Record<string, unknown> | undefined, key: string) => {
-    const value = settings?.[key]
+const stringSetting = (settings: Record<string, unknown>, key: string) => {
+    const value = settings[key]
     return typeof value === "string" && value.trim() ? value.trim() : undefined
 }
 
-const numberSetting = (settings: Record<string, unknown> | undefined, key: string) => {
-    const value = settings?.[key]
-    if (typeof value === "number" && Number.isFinite(value)) return value
-    if (typeof value === "string" && value.trim()) {
-        const parsed = Number(value)
-        if (Number.isFinite(parsed)) return parsed
-    }
-
-    return undefined
-}
-
-const booleanSetting = (settings: Record<string, unknown> | undefined, key: string, fallback: boolean) => {
-    const value = settings?.[key]
+const booleanSetting = (settings: Record<string, unknown>, key: string, fallback: boolean) => {
+    const value = settings[key]
     if (typeof value === "boolean") return value
     if (typeof value === "string") return ["1", "true", "yes", "ja", "on"].includes(value.trim().toLowerCase())
     return fallback
 }
 
-const profileSetting = (
-    settings: Record<string, unknown> | undefined,
-    fallback: AgentConfig["postprocessProfile"]
-): AgentConfig["postprocessProfile"] => {
-    const value = settings?.postprocessProfile
-    if (value === "document" || value === "receipt" || value === "raw") return value
-    return fallback
-}
-
-const ensureFilenameExtension = (filename: string, format: AgentConfig["scanFormat"]) => {
-    const ext = path.extname(filename)
-    if (!ext) return `${filename}.${format}`
-
-    const expectedExt = `.${format}`
-    if (ext.toLowerCase() === expectedExt) return filename
-    return `${filename.slice(0, -ext.length)}${expectedExt}`
-}
-
-const fallbackRawResult = (scanOutputPath: string, jobId: string): ScanResult => ({
-    path: scanOutputPath,
-    filename: `${jobId}.raw.png`,
-    mimeType: "image/png",
-})
-
 export const hasSane = () => commandExists("scanimage")
 
 export const listScanners = async () => {
     if (!await hasSane()) return []
-
     const result = await runCommand("scanimage", ["-L"], { timeoutMs: 10_000 })
     if (result.code !== 0) return []
-
-    return result.stdout
-        .split(/\r?\n/)
-        .map((line) => line.trim())
-        .filter((line) => line.startsWith("device `"))
-        .map((line) => line.match(/device `([^']+)'/)?.[1])
+    return result.stdout.split(/\r?\n/)
+        .map((line) => line.trim().match(/^device `([^']+)'/)?.[1])
         .filter((device): device is string => Boolean(device))
 }
 
-export const runScan = async (config: AgentConfig, job: ScanJob): Promise<ScanResult> => {
-    if (!await hasSane()) {
-        throw new Error("scanimage ist nicht installiert oder nicht im PATH")
-    }
-
-    mkdirSync(config.workDir, { recursive: true })
+export const runScan = async (
+    config: AgentConfig,
+    job: ScanJob,
+    reportProgress: (message: string) => Promise<void> = async () => {}
+): Promise<ScanResult> => {
+    if (!await hasSane()) throw new Error("scanimage ist nicht installiert oder nicht im PATH")
 
     const settings = job.settings || {}
-    const format = stringSetting(settings, "format") as AgentConfig["scanFormat"] | undefined || config.scanFormat
-    const resolution = numberSetting(settings, "resolution") || config.scanResolution
+    const format = stringSetting(settings, "format") || config.scanFormat
+    if (!["pdf", "png", "tiff"].includes(format)) throw new Error("Scan-Format muss PDF, PNG oder TIFF sein")
+    const resolution = Number(settings.resolution ?? config.scanResolution)
+    if (!Number.isFinite(resolution) || resolution < 75 || resolution > 1200) {
+        throw new Error("Die Scan-Auflösung muss zwischen 75 und 1200 dpi liegen")
+    }
     const mode = stringSetting(settings, "mode") || config.scanMode
     const source = stringSetting(settings, "source") || config.scanSource
     const scannerName = job.scannerName || config.scannerName
-    const filename = ensureFilenameExtension(job.requestedFilename || `${job.id}.${format}`, format)
-    const outputPath = path.join(config.workDir, filename)
-    const shouldPostprocess = booleanSetting(settings, "postprocess", config.scanPostprocess)
-    const postprocessProfile = profileSetting(settings, config.postprocessProfile)
-    const scanFormat = shouldPostprocess ? "png" : format
-    const scanOutputPath = shouldPostprocess
-        ? path.join(config.workDir, `${job.id}.raw.png`)
-        : outputPath
+    const paperSize = stringSetting(settings, "paperSize") || "auto"
+    if (!["auto", "A4", "Letter"].includes(paperSize)) throw new Error("Unbekanntes Papierformat")
+    const batch = Boolean(source && /adf|document feeder/i.test(source))
+    if (batch && format !== "pdf") throw new Error("Für Einzug- und Duplex-Scans bitte PDF als Format wählen")
+    const postprocess = booleanSetting(settings, "postprocess", config.scanPostprocess)
+    const profile = stringSetting(settings, "postprocessProfile") || config.postprocessProfile
+    if (!["document", "receipt", "raw"].includes(profile)) throw new Error("Unbekanntes Nachbearbeitungsprofil")
+    await checkScanRuntime(config, postprocess)
 
-    const args = [
-        "--format",
-        scanFormat,
-        "--resolution",
-        String(resolution),
-        "--mode",
-        mode,
-        "--output-file",
-        scanOutputPath,
-    ]
+    // Each physical scan has its own directory. A filename cannot escape it or overwrite another job.
+    await mkdir(config.workDir, { recursive: true, mode: 0o700 })
+    const jobDir = await mkdtemp(path.join(path.resolve(config.workDir), "scan-"))
+    const requested = path.basename((job.requestedFilename || "scan").replace(/\\/g, "/"))
+    const stem = requested.replace(/\.[^.]*$/, "").replace(/[\x00-\x1f]/g, "").trim() || "scan"
+    const outputPath = path.join(jobDir, `${stem}.${format}`)
 
-    if (source) args.push("--source", source)
-    if (scannerName) args.push("--device-name", scannerName)
+    try {
+        // Device/source come first: changing source can change the remaining SANE options.
+        const args: string[] = []
+        if (scannerName) args.push("--device-name", scannerName)
+        if (source) args.push("--source", source)
+        args.push("--mode", mode, "--resolution", String(resolution), "--format=png")
+        if (paperSize !== "auto") {
+            const [width, height] = paperSize === "A4" ? [210, 297] : [215.9, 279.4]
+            if (scannerName?.startsWith("fujitsu:")) {
+                args.push("--page-width", String(width), "--page-height", String(height))
+            }
+            args.push("-l", "0", "-t", "0", "-x", String(width), "-y", String(height))
+        }
+        if (batch) {
+            // Relative pattern avoids interpreting '%' in a user-configured work directory.
+            args.push("--batch=page-%06d.png", "--batch-start=1", "--batch-increment=1")
+        } else {
+            args.push("--output-file", "page-000001.png")
+        }
+        await reportProgress(batch ? "Dokumente werden eingezogen und gescannt..." : "Seite wird gescannt...")
+        const result = await runCommand("scanimage", args, {
+            cwd: jobDir,
+            timeoutMs: config.scanTimeoutSeconds * 1000,
+        })
 
-    const result = await runCommand("scanimage", args, { timeoutMs: 5 * 60 * 1000 })
-
-    if (result.code !== 0) {
-        throw new Error(result.stderr || `scanimage wurde mit Code ${result.code} beendet`)
-    }
-
-    if (shouldPostprocess) {
-        if (!await hasOpenCvPostprocessRuntime(config)) {
-            const message = "OpenCV-Nachbearbeitung ist aktiviert, aber python3 mit cv2, Pillow und numpy ist nicht verfügbar"
-            if (config.postprocessStrict) throw new Error(message)
-
-            log.warn(`${message}. Rohscan wird ohne Korrektur hochgeladen.`, {
-                jobId: job.id,
-                python: config.postprocessPython,
-            })
-            return fallbackRawResult(scanOutputPath, job.id)
+        // SANE_STATUS_NO_DOCS (7) is the expected end of an unlimited ADF batch.
+        // Never turn paper jams, I/O errors or canceled scans into a successful partial upload.
+        if (result.code !== 0 && !(batch && result.code === 7)) {
+            throw new Error(result.stderr.trim() || `scanimage wurde mit Code ${result.code} beendet`)
+        }
+        const names = (await readdir(jobDir))
+            .filter((name) => /^page-\d+\.png$/.test(name))
+            .sort((a, b) => Number(a.match(/\d+/)![0]) - Number(b.match(/\d+/)![0]))
+        const pages: string[] = []
+        for (const [index, name] of names.entries()) {
+            const page = path.join(jobDir, name)
+            // Some scanimage versions leave a zero-byte file for the final NO_DOCS attempt.
+            if ((await stat(page)).size === 0) {
+                if (batch && result.code === 7 && index === names.length - 1) continue
+                throw new Error(`Unvollständige Scan-Seite: ${name}`)
+            }
+            if (Number(name.match(/\d+/)![0]) !== index + 1) throw new Error("Im Scan fehlen Seiten")
+            pages.push(page)
+        }
+        if (!pages.length) throw new Error("Keine Seiten gescannt. Bitte Papier in den Einzug legen.")
+        if (source?.toLowerCase().includes("duplex") && pages.length % 2 !== 0) {
+            throw new Error("Der Duplex-Scan ist unvollständig: Zu einer Vorderseite fehlt die Rückseite.")
         }
 
-        try {
-            return await postprocessScan(config, scanOutputPath, filename, format, postprocessProfile)
-        } catch (error) {
-            if (config.postprocessStrict) throw error
-
-            log.warn("OpenCV-Nachbearbeitung fehlgeschlagen. Rohscan wird ohne Korrektur hochgeladen.", {
-                jobId: job.id,
-                error: error instanceof Error ? error.message : String(error),
-            })
-            return fallbackRawResult(scanOutputPath, job.id)
-        }
-    }
-
-    return {
-        path: outputPath,
-        filename,
-        mimeType: mimeTypes[format] || "application/octet-stream",
+        await reportProgress(`${pages.length} Seiten gescannt. Datei wird erstellt...`)
+        return await createScanDocument(config, pages, outputPath, resolution, postprocess, profile as AgentConfig["postprocessProfile"])
+    } catch (error) {
+        const message = error instanceof Error ? error.message : String(error)
+        throw new Error(`${message} Rohscans bleiben zur Wiederherstellung in ${jobDir}.`)
     }
 }

agents/fedeo-device-agent/src/types.ts

@@ -9,6 +9,7 @@ export type AgentConfig = {
     scanResolution: number
     scanMode: string
     scanSource?: string
+    scanTimeoutSeconds: number
     scanPostprocess: boolean
     postprocessProfile: "document" | "receipt" | "raw"
     postprocessPython: string
@@ -45,4 +46,6 @@ export type ScanResult = {
     path: string
     filename: string
     mimeType: string
+    pageCount: number
+    warnings: string[]
 }