Update app.py

app.py

@@ -4,7 +4,7 @@ import numpy as np
 
 model = ocr_predictor(det_arch='db_resnet50', reco_arch='crnn_vgg16_bn', pretrained=True)
 
-def extract_perfect_table(image):
+def extract_6_column_table(image):
     if image is None:
         return "Please upload an image."
     
@@ -14,9 +14,9 @@ def extract_perfect_table(image):
     
     markdown_rows = []
     
-    # Precise Column Boundaries (Adjusted for 4-column accountancy papers)
-    # Bucket 1: < 22% | Bucket 2: 22%-62% | Bucket 3: 62%-82% | Bucket 4: > 82%
-    col_bounds = [0.22, 0.62, 0.82] 
+    # NEW: Boundaries for 6 columns (Account Name, Code, Statement, Group, Sub-Group, Normally)
+    # Adjusted based on your specific screenshot layout
+    col_bounds = [0.30, 0.38, 0.52, 0.68, 0.90] 
 
     for page in json_export['pages']:
         words_list = []
@@ -33,7 +33,7 @@ def extract_perfect_table(image):
         rows = []
         current_row = [words_list[0]]
         for i in range(1, len(words_list)):
-            if abs(words_list[i]['y'] - current_row[-1]['y']) < 0.015:
+            if abs(words_list[i]['y'] - current_row[-1]['y']) < 0.012: # Tighter row grouping
                 current_row.append(words_list[i])
             else:
                 rows.append(current_row)
@@ -41,52 +41,55 @@ def extract_perfect_table(image):
         rows.append(current_row)
         
         for row in rows:
-            slots = ["", "", "", ""]
-            row.sort(key=lambda w: w['x']) # Ensure words in row are left-to-right
+            # 6 slots for your specific table structure
+            slots = ["", "", "", "", "", ""]
+            row.sort(key=lambda w: w['x'])
             
             for w in row:
                 x = w['x']
-                text = w['text']
+                t = w['text']
                 
-                # Logic to handle the "Debit Credit" header merging
-                if "Debit" in text and x > 0.55 and x < 0.75:
-                    slots[2] += text + " "
-                elif "Credit" in text and x > 0.75:
-                    slots[3] += text + " "
-                # General bucket logic for the rest of the rows
+                # Forced Column Logic for "Debit/Credit"
+                if t in ["Debit", "Credit"] and x > 0.85:
+                    slots[5] = t
                 elif x < col_bounds[0]:
-                    slots[0] += text + " "
+                    slots[0] += t + " "
                 elif x < col_bounds[1]:
-                    slots[1] += text + " "
+                    slots[1] += t + " "
                 elif x < col_bounds[2]:
-                    slots[2] += text + " "
+                    slots[2] += t + " "
+                elif x < col_bounds[3]:
+                    slots[3] += t + " "
+                elif x < col_bounds[4]:
+                    slots[4] += t + " "
                 else:
-                    slots[3] += text + " "
+                    slots[5] += t + " "
             
             clean_slots = [s.strip() for s in slots]
             markdown_rows.append("| " + " | ".join(clean_slots) + " |")
 
     return "\n".join(markdown_rows)
 
+# UI
 with gr.Blocks() as demo:
-    gr.Markdown("## 📑 Final Accountancy Table Extractor")
-    gr.Markdown("Fixed header alignment for Debit/Credit columns.")
+    gr.Markdown("## 📑 6-Column Accounting Extractor")
+    gr.Markdown("Separates Account Name, Code, Group, Sub-Group, and Normal Balance.")
     
     with gr.Row():
         with gr.Column():
             img_in = gr.Image(type="pil")
-            btn = gr.Button("Extract Table", variant="primary")
+            btn = gr.Button("Extract Complex Table", variant="primary")
         with gr.Column():
-            out = gr.Textbox(label="Copy-Paste Ready", lines=20, elem_id="out_box")
+            out = gr.Textbox(label="Result", lines=25, elem_id="out_box")
             copy_btn = gr.Button("📋 Copy Table")
             copy_btn.click(None, None, None, js="""
                 () => {
                     const text = document.querySelector('#out_box textarea').value;
                     navigator.clipboard.writeText(text);
-                    alert('Table copied! Header columns are now separated.');
+                    alert('Copied! Column merges have been fixed.');
                 }
             """)
 
-    btn.click(extract_perfect_table, inputs=img_in, outputs=out)
+    btn.click(extract_6_column_table, inputs=img_in, outputs=out)
 
 demo.launch()