Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -5,6 +5,9 @@ from PIL import Image
|
|
| 5 |
import gradio as gr
|
| 6 |
import io
|
| 7 |
|
|
|
|
|
|
|
|
|
|
| 8 |
# IBS Café schema columns
|
| 9 |
COLUMNS = [
|
| 10 |
"Parent Category", "Category", "Store Item Name", "Item Code", "Master Item Name", "EAN Code",
|
|
@@ -48,7 +51,6 @@ def parse_menu_text(text):
|
|
| 48 |
if prices:
|
| 49 |
name = re.sub(price_pattern, '', line).strip(" -:–")
|
| 50 |
if '/' in line and len(prices) > 1:
|
| 51 |
-
# multi-size e.g., 149/199
|
| 52 |
for i, price in enumerate(prices):
|
| 53 |
size_label = f" ({['Regular', 'Large', 'XL'][i]})" if i < 3 else f" (Option {i+1})"
|
| 54 |
store_name = f"{name}{size_label}"
|
|
@@ -83,28 +85,31 @@ def ocr_and_extract(image):
|
|
| 83 |
rows = parse_menu_text(text)
|
| 84 |
|
| 85 |
df = pd.DataFrame(rows, columns=COLUMNS)
|
| 86 |
-
csv_buffer = io.StringIO()
|
| 87 |
-
df.to_csv(csv_buffer, index=False)
|
| 88 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 89 |
json_output = {"rows": rows, "needs_review": []}
|
| 90 |
-
return text,
|
| 91 |
|
| 92 |
# Gradio UI
|
| 93 |
with gr.Blocks(title="Menu → IBS Schema Extractor") as demo:
|
| 94 |
-
gr.Markdown("## 🧾 Menu OCR
|
| 95 |
-
gr.Markdown("Upload a menu image and extract structured data in
|
| 96 |
|
| 97 |
with gr.Row():
|
| 98 |
image_input = gr.Image(type="filepath", label="Upload Menu Image")
|
| 99 |
|
| 100 |
extract_btn = gr.Button("Extract")
|
| 101 |
-
|
| 102 |
with gr.Tab("Extracted Text"):
|
| 103 |
text_output = gr.Textbox(label="OCR Text", lines=10)
|
| 104 |
|
| 105 |
-
with gr.Tab("
|
| 106 |
-
|
| 107 |
-
csv_file = gr.File(label="Download CSV")
|
| 108 |
|
| 109 |
with gr.Tab("JSON Output"):
|
| 110 |
json_output = gr.JSON(label="Structured JSON")
|
|
@@ -112,7 +117,7 @@ with gr.Blocks(title="Menu → IBS Schema Extractor") as demo:
|
|
| 112 |
extract_btn.click(
|
| 113 |
ocr_and_extract,
|
| 114 |
inputs=[image_input],
|
| 115 |
-
outputs=[text_output,
|
| 116 |
)
|
| 117 |
|
| 118 |
demo.launch()
|
|
|
|
| 5 |
import gradio as gr
|
| 6 |
import io
|
| 7 |
|
| 8 |
+
# Uncomment and edit this path if you’re on Windows
|
| 9 |
+
# pytesseract.pytesseract.tesseract_cmd = r"C:\Program Files\Tesseract-OCR\tesseract.exe"
|
| 10 |
+
|
| 11 |
# IBS Café schema columns
|
| 12 |
COLUMNS = [
|
| 13 |
"Parent Category", "Category", "Store Item Name", "Item Code", "Master Item Name", "EAN Code",
|
|
|
|
| 51 |
if prices:
|
| 52 |
name = re.sub(price_pattern, '', line).strip(" -:–")
|
| 53 |
if '/' in line and len(prices) > 1:
|
|
|
|
| 54 |
for i, price in enumerate(prices):
|
| 55 |
size_label = f" ({['Regular', 'Large', 'XL'][i]})" if i < 3 else f" (Option {i+1})"
|
| 56 |
store_name = f"{name}{size_label}"
|
|
|
|
| 85 |
rows = parse_menu_text(text)
|
| 86 |
|
| 87 |
df = pd.DataFrame(rows, columns=COLUMNS)
|
|
|
|
|
|
|
| 88 |
|
| 89 |
+
# Save Excel file in memory
|
| 90 |
+
excel_buffer = io.BytesIO()
|
| 91 |
+
with pd.ExcelWriter(excel_buffer, engine='openpyxl') as writer:
|
| 92 |
+
df.to_excel(writer, index=False, sheet_name="Menu")
|
| 93 |
+
excel_buffer.seek(0)
|
| 94 |
+
|
| 95 |
json_output = {"rows": rows, "needs_review": []}
|
| 96 |
+
return text, excel_buffer, json_output
|
| 97 |
|
| 98 |
# Gradio UI
|
| 99 |
with gr.Blocks(title="Menu → IBS Schema Extractor") as demo:
|
| 100 |
+
gr.Markdown("## 🧾 Menu OCR → IBS Café Excel Generator")
|
| 101 |
+
gr.Markdown("Upload a menu image and extract structured data in Excel (.xlsx) + JSON formats.")
|
| 102 |
|
| 103 |
with gr.Row():
|
| 104 |
image_input = gr.Image(type="filepath", label="Upload Menu Image")
|
| 105 |
|
| 106 |
extract_btn = gr.Button("Extract")
|
| 107 |
+
|
| 108 |
with gr.Tab("Extracted Text"):
|
| 109 |
text_output = gr.Textbox(label="OCR Text", lines=10)
|
| 110 |
|
| 111 |
+
with gr.Tab("Excel Output"):
|
| 112 |
+
excel_file = gr.File(label="Download Excel (.xlsx)")
|
|
|
|
| 113 |
|
| 114 |
with gr.Tab("JSON Output"):
|
| 115 |
json_output = gr.JSON(label="Structured JSON")
|
|
|
|
| 117 |
extract_btn.click(
|
| 118 |
ocr_and_extract,
|
| 119 |
inputs=[image_input],
|
| 120 |
+
outputs=[text_output, excel_file, json_output]
|
| 121 |
)
|
| 122 |
|
| 123 |
demo.launch()
|