Create app.py

app.py

@@ -0,0 +1,118 @@
+import re
+import pandas as pd
+import pytesseract
+from PIL import Image
+import gradio as gr
+import io
+
+# IBS Café schema columns
+COLUMNS = [
+    "Parent Category", "Category", "Store Item Name", "Item Code", "Master Item Name", "EAN Code",
+    "price", "Active", "Priortiy", "Image", "food type", "NoOfMains", "onlineName",
+    "Menu/MRP", "itemTaxInclusive", "taxPct", "brandName", "classificationCode", "HSN CODE"
+]
+
+DEFAULTS = {
+    "Item Code": "",
+    "Master Item Name": "",
+    "EAN Code": "",
+    "Active": "Y",
+    "Priortiy": "",
+    "Image": "",
+    "food type": "",
+    "NoOfMains": "",
+    "itemTaxInclusive": "Y",
+    "taxPct": "5",
+    "brandName": "Nescafe",
+    "classificationCode": "",
+    "HSN CODE": "",
+}
+
+price_pattern = re.compile(r'(?<!\d)(?:₹|Rs\.?\s*)?\d{2,4}(?:\.\d{1,2})?(?!\d)')
+
+def parse_menu_text(text):
+    lines = [line.strip() for line in text.splitlines() if line.strip()]
+    rows = []
+    parent_category, category = "", ""
+    
+    for line in lines:
+        if line.isupper() and len(line.split()) <= 4:
+            # treat as category or parent
+            if not parent_category:
+                parent_category = line.title()
+            else:
+                category = line.title()
+            continue
+        
+        prices = price_pattern.findall(line)
+        if prices:
+            name = re.sub(price_pattern, '', line).strip(" -:–")
+            if '/' in line and len(prices) > 1:
+                # multi-size e.g., 149/199
+                for i, price in enumerate(prices):
+                    size_label = f" ({['Regular', 'Large', 'XL'][i]})" if i < 3 else f" (Option {i+1})"
+                    store_name = f"{name}{size_label}"
+                    row = build_row(parent_category, category, store_name, price)
+                    rows.append(row)
+            else:
+                price = prices[0]
+                store_name = name
+                row = build_row(parent_category, category, store_name, price)
+                rows.append(row)
+    
+    return rows
+
+def build_row(parent, category, name, price):
+    base = {**DEFAULTS}
+    base.update({
+        "Parent Category": parent,
+        "Category": category,
+        "Store Item Name": name,
+        "price": price,
+        "onlineName": name,
+        "Menu/MRP": price
+    })
+    return base
+
+def ocr_and_extract(image):
+    if image is None:
+        return "Please upload an image.", None, None
+    
+    img = Image.open(image)
+    text = pytesseract.image_to_string(img)
+    rows = parse_menu_text(text)
+    
+    df = pd.DataFrame(rows, columns=COLUMNS)
+    csv_buffer = io.StringIO()
+    df.to_csv(csv_buffer, index=False)
+    
+    json_output = {"rows": rows, "needs_review": []}
+    return text, csv_buffer.getvalue(), json_output
+
+# Gradio UI
+with gr.Blocks(title="Menu → IBS Schema Extractor") as demo:
+    gr.Markdown("## 🧾 Menu OCR to IBS Café Schema")
+    gr.Markdown("Upload a menu image and extract structured data in CSV + JSON formats.")
+    
+    with gr.Row():
+        image_input = gr.Image(type="filepath", label="Upload Menu Image")
+    
+    extract_btn = gr.Button("Extract")
+    
+    with gr.Tab("Extracted Text"):
+        text_output = gr.Textbox(label="OCR Text", lines=10)
+    
+    with gr.Tab("CSV Output"):
+        csv_output = gr.Textbox(label="CSV Data", lines=10)
+        csv_file = gr.File(label="Download CSV")
+    
+    with gr.Tab("JSON Output"):
+        json_output = gr.JSON(label="Structured JSON")
+    
+    extract_btn.click(
+        ocr_and_extract,
+        inputs=[image_input],
+        outputs=[text_output, csv_output, json_output]
+    )
+
+demo.launch()