Yapamıyor Türkçe ocr yapacak aynı abbyfine redaer format çeviri gibi Adobe redaer DC ocr format çevirisi gibi çalışmalı

index.html

@@ -42,7 +42,8 @@
                         <option value="json">JSON</option>
                         <option value="markdown">Markdown</option>
                         <option value="text">Plain Text</option>
-                    </select>
+                        <option value="formatted">Formatted Text</option>
+</select>
                 </div>
                 <button id="processBtn" class="bg-indigo-600 hover:bg-indigo-700 text-white font-medium py-3 px-6 rounded-lg transition duration-200 flex items-center justify-center mt-6 sm:mt-auto">
                     <i data-feather="cpu" class="mr-2"></i> Process Files

script.js

@@ -240,18 +240,67 @@ document.addEventListener('DOMContentLoaded', function() {
             reader.readAsArrayBuffer(file);
         });
     }
-    
-    async function extractTextFromImage(file) {
+async function extractTextFromImage(file) {
         return new Promise((resolve, reject) => {
             Tesseract.recognize(
                 file,
-                'eng',
-                { logger: m => console.log(m) }
-            ).then(({ data: { text } }) => {
-                resolve(text);
+                'tur+eng', // Turkish + English languages
+                { 
+                    logger: m => console.log(m),
+                    preserve_interword_spaces: true,
+                    tessedit_pageseg_mode: 6, // Assume a single uniform block of text 
+                    tessedit_char_whitelist: 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789.,!?-(){}[]/\\\'" ',
+                    tessedit_create_hocr: 1 // Include formatting info
+                }
+            ).then(({ data: { text, hocr } }) => {
+                if (outputFormat.value === 'formatted') {
+                    // Process formatted output similar to Adobe/Abbyy
+                    const formatted = processFormattedOCR(hocr);
+                    resolve(formatted);
+                } else {
+                    resolve(text);
+                }
             }).catch(reject);
         });
-    }
+
+        function processFormattedOCR(hocr) {
+            // Parse hOCR output to preserve formatting and layout
+            const parser = new DOMParser();
+            const doc = parser.parseFromString(hocr, 'text/html');
+            const paragraphs = doc.querySelectorAll('.ocr_par');
+            
+            let formattedText = '';
+            
+            paragraphs.forEach(par => {
+                const lines = par.querySelectorAll('.ocr_line');
+                lines.forEach(line => {
+                    const words = line.querySelectorAll('.ocrx_word');
+                    let lineText = '';
+                    
+                    words.forEach((word, index) => {
+                        const wordText = word.textContent || '';
+                        const wordConfidence = parseFloat(word.getAttribute('title')
+                            .match(/x_wconf (\d+)/)[1]);
+                        
+                        // Apply formatting based on confidence and context
+                        if (wordConfidence < 60) {
+                            lineText += `[${wordText}] `;
+                        } else if (wordConfidence < 80 && index > 0) {
+                            lineText += `${wordText}`;
+                        } else {
+                            lineText += `${wordText} `;
+                        }
+                    });
+                    
+                    formattedText += lineText.trim() + '\n';
+                });
+                
+                formattedText += '\n';
+            });
+            
+            return formattedText;
+        }
+}
     
     function displayResult(result) {
         const resultCard = document.createElement('div');

style.css

@@ -20,7 +20,6 @@
     max-height: 150px;
     object-fit: contain;
 }
-
 #resultsContainer pre {
     background-color: #f8f9fa;
     padding: 16px;
@@ -30,4 +29,19 @@
     word-wrap: break-word;
     max-height: 500px;
     overflow-y: auto;
-}
+    font-family: monospace;
+    line-height: 1.5;
+}
+
+/* OCR confidence styling */
+pre [confidence-low] {
+    background-color: #ffebee;
+    color: #c62828;
+    padding: 0 2px;
+}
+
+pre [confidence-medium] {
+    background-color: #fff8e1;
+    color: #f57f17;
+    padding: 0 2px;
+}