gfdfg

debug_text.ts

@@ -0,0 +1,48 @@
+import * as pdfjsLib from 'pdfjs-dist/legacy/build/pdf.js';
+import * as fs from 'fs';
+
+async function debugText(filePath: string) {
+    console.log(`\nLese PDF für Text-Debugging: ${filePath}`);
+    
+    const dataBuffer = fs.readFileSync(filePath);
+    const loadingTask = pdfjsLib.getDocument({ data: new Uint8Array(dataBuffer) });
+    const pdfDocument = await loadingTask.promise;
+    
+    console.log(`Anzahl Seiten: ${pdfDocument.numPages}`);
+    
+    // Wir nehmen uns Seite 1 vor
+    const page = await pdfDocument.getPage(1);
+    const textContent = await page.getTextContent();
+    
+    console.log(`\n--- RAW TEXT ITEMS (Die ersten 40 Fragmente) ---`);
+    for (let i = 0; i < Math.min(40, textContent.items.length); i++) {
+        const item = textContent.items[i] as any;
+        console.log(`Y: ${item.transform[5].toFixed(1).padStart(6)} | X: ${item.transform[4].toFixed(1).padStart(6)} | Text: "${item.str}"`);
+    }
+
+    // Unsere Sortier- und Bereinigungslogik anwenden
+    const sortedItems = textContent.items
+        .map((item: any) => ({
+            text: item.str,
+            x: item.transform[4],
+            y: item.transform[5]
+        }))
+        .sort((a: any, b: any) => {
+            if (Math.abs(b.y - a.y) > 5) return b.y - a.y; 
+            return a.x - b.x;
+        });
+
+    const textRaw = sortedItems.map((i: any) => i.text).join('').toUpperCase().replace(/[^A-Z]/g, '');
+    
+    console.log(`\n--- BEREINIGTER SUCH-STRING (Die ersten 200 Zeichen) ---`);
+    console.log(textRaw.substring(0, 200));
+    console.log(`\nEnthält 'BUYERINFORMATIONSHEET'? -> ${textRaw.includes('BUYERINFORMATIONSHEET')}`);
+}
+
+const args = process.argv.slice(2);
+const pdfArgIndex = args.indexOf('--pdf');
+if (pdfArgIndex !== -1 && args[pdfArgIndex + 1]) {
+    debugText(args[pdfArgIndex + 1]).catch(console.error);
+} else {
+    console.error("Bitte --pdf Parameter angeben.");
+}