Spaces:
Sleeping
Sleeping
Ayush soni
commited on
Commit
·
e98e646
1
Parent(s):
6034171
app.py fixwd
Browse files
app.py
CHANGED
|
@@ -5,28 +5,28 @@ from llm_processor import load_llm_model, generate_json_from_text
|
|
| 5 |
# Load LLM model on startup
|
| 6 |
load_llm_model()
|
| 7 |
|
| 8 |
-
def process_invoice(
|
| 9 |
-
|
| 10 |
-
|
| 11 |
|
| 12 |
-
# Step 1: Extract
|
| 13 |
-
raw_text = extract_text_from_image(
|
| 14 |
if not raw_text or "No text detected" in raw_text:
|
| 15 |
return raw_text, {"error": "No text could be extracted from the image."}
|
| 16 |
|
| 17 |
-
# Step 2:
|
| 18 |
json_data = generate_json_from_text(raw_text)
|
| 19 |
-
|
| 20 |
return raw_text, json_data
|
| 21 |
|
| 22 |
|
|
|
|
| 23 |
### Gradio UI
|
| 24 |
with gr.Blocks() as demo:
|
| 25 |
gr.Markdown("# 🧾 Invoice Processing App")
|
| 26 |
gr.Markdown("Upload an invoice image. The app extracts **OCR text** and generates **structured JSON**.")
|
| 27 |
|
| 28 |
with gr.Row():
|
| 29 |
-
input_file = gr.File(label="Upload Invoice Image", type="
|
| 30 |
|
| 31 |
with gr.Row():
|
| 32 |
raw_text_output = gr.Textbox(label="Extracted OCR Text", lines=10)
|
|
|
|
| 5 |
# Load LLM model on startup
|
| 6 |
load_llm_model()
|
| 7 |
|
| 8 |
+
def process_invoice(file_bytes):
|
| 9 |
+
if file_bytes is None:
|
| 10 |
+
return "No file uploaded", {"error": "No file uploaded"}
|
| 11 |
|
| 12 |
+
# Step 1: Extract text
|
| 13 |
+
raw_text = extract_text_from_image(file_bytes)
|
| 14 |
if not raw_text or "No text detected" in raw_text:
|
| 15 |
return raw_text, {"error": "No text could be extracted from the image."}
|
| 16 |
|
| 17 |
+
# Step 2: JSON generation
|
| 18 |
json_data = generate_json_from_text(raw_text)
|
|
|
|
| 19 |
return raw_text, json_data
|
| 20 |
|
| 21 |
|
| 22 |
+
|
| 23 |
### Gradio UI
|
| 24 |
with gr.Blocks() as demo:
|
| 25 |
gr.Markdown("# 🧾 Invoice Processing App")
|
| 26 |
gr.Markdown("Upload an invoice image. The app extracts **OCR text** and generates **structured JSON**.")
|
| 27 |
|
| 28 |
with gr.Row():
|
| 29 |
+
input_file = gr.File(label="Upload Invoice Image", type="binary", file_types=[".png", ".jpg", ".jpeg"])
|
| 30 |
|
| 31 |
with gr.Row():
|
| 32 |
raw_text_output = gr.Textbox(label="Extracted OCR Text", lines=10)
|