From 70f09fa9da083bf863f476833ec34114425c53ea Mon Sep 17 00:00:00 2001
From: Andreas Knuth <andreas.knuth@gmail.com>
Date: Sun, 12 Jul 2026 22:59:21 -0500
Subject: [PATCH] sdfsdf

---
 vision_runner.ts | 27 +++++++++++++++++++++++----
 1 file changed, 23 insertions(+), 4 deletions(-)

diff --git a/vision_runner.ts b/vision_runner.ts
index d1773c4..fd06178 100644
--- a/vision_runner.ts
+++ b/vision_runner.ts
@@ -47,7 +47,7 @@ interface Args {
   api: string;
   limit: number;
   only: string | null;
-  dpi: number;
+  dpi: number | "auto";
   maxPages: number;
   force: boolean;
   outDir: string;
@@ -73,7 +73,10 @@ function parseArgs(): Args {
     api,
     limit: parseInt(get("--limit", "0")!, 10) || 0,
     only: get("--only"),
-    dpi: parseInt(get("--dpi", "150")!, 10) || 150,
+    dpi: ((): number | "auto" => {
+      const v = get("--dpi", "auto")!;
+      return v === "auto" ? "auto" : parseInt(v, 10) || 150;
+    })(),
     maxPages: parseInt(get("--max-pages", "8")!, 10) || 8,
     force: a.includes("--force"),
     outDir: get("--out-dir", path.dirname(input))!,
@@ -234,6 +237,19 @@ async function renderPdf(pdfPath: string, dpi: number, maxPages: number, tmpDir:
   return files;
 }
 
+/** Hat das PDF eine nennenswerte Textebene? (Hybrid: Werte getippt → 150 DPI
+ *  reicht. Reiner Scan: keine Textebene, Handschrift möglich → 200 DPI.) */
+async function hasTextLayer(pdfPath: string, maxPages: number): Promise<boolean> {
+  try {
+    const { stdout } = await execFileP("pdftotext", ["-l", String(maxPages), pdfPath, "-"], {
+      timeout: 30_000, maxBuffer: 10 * 1024 * 1024,
+    });
+    return stdout.replace(/\s+/g, "").length > 100;
+  } catch {
+    return false; // im Zweifel als Scan behandeln → hohe Auflösung
+  }
+}
+
 // ---------------------------------------------------------------------------
 // VLM-Aufruf (llama-server, OpenAI-kompatibel, JSON-Schema)
 // ---------------------------------------------------------------------------
@@ -481,8 +497,11 @@ async function main(): Promise<void> {
 
     const tmpDir = await fsp.mkdtemp(path.join(os.tmpdir(), "bvs-"));
     try {
-      progress(`${tag} … rendere`);
-      const images = await renderPdf(pdfPath, args.dpi, args.maxPages, tmpDir);
+      const dpi = args.dpi === "auto"
+        ? ((await hasTextLayer(pdfPath, args.maxPages)) ? 150 : 200)
+        : args.dpi;
+      progress(`${tag} … rendere (${dpi} dpi)`);
+      const images = await renderPdf(pdfPath, dpi, args.maxPages, tmpDir);
 
       let vis: VisionRaw | null = null;
       let lastErr = "";