Update app.py
Browse files
app.py
CHANGED
|
@@ -43,4 +43,57 @@ def extract_data(raw_text, fields_to_extract):
|
|
| 43 |
# Fallback: Safely strip markdown code blocks without using complex regex
|
| 44 |
# that might break code editors during copy-pasting
|
| 45 |
if output_text.startswith("```"):
|
| 46 |
-
# Remove the starting ```json or
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 43 |
# Fallback: Safely strip markdown code blocks without using complex regex
|
| 44 |
# that might break code editors during copy-pasting
|
| 45 |
if output_text.startswith("```"):
|
| 46 |
+
# Remove the starting ```json or
|
| 47 |
+
# Remove the ending ```
|
| 48 |
+
output_text = re.sub(r"\n?```$", "", output_text)
|
| 49 |
+
|
| 50 |
+
# Parse the text into an actual JSON dictionary for the Gradio UI
|
| 51 |
+
structured_data = json.loads(output_text.strip())
|
| 52 |
+
return structured_data
|
| 53 |
+
|
| 54 |
+
except json.JSONDecodeError:
|
| 55 |
+
return {
|
| 56 |
+
"error": "The model failed to return valid JSON. It returned this instead:",
|
| 57 |
+
"raw_output": output_text
|
| 58 |
+
}
|
| 59 |
+
except Exception as e:
|
| 60 |
+
return {"error": str(e)}
|
| 61 |
+
|
| 62 |
+
# -------------------------
|
| 63 |
+
# Build the Gradio UI
|
| 64 |
+
# -------------------------
|
| 65 |
+
with gr.Blocks(theme=gr.themes.Soft()) as demo:
|
| 66 |
+
gr.Markdown("# 🛟 The Data Rescuer")
|
| 67 |
+
gr.Markdown(f"**Powered by `{MODEL_ID}`** | Turn messy transcripts, notes, and OCR text into clean JSON data.")
|
| 68 |
+
|
| 69 |
+
with gr.Row():
|
| 70 |
+
# Left Column: Inputs
|
| 71 |
+
with gr.Column():
|
| 72 |
+
raw_input = gr.Textbox(
|
| 73 |
+
label="1. Paste Unstructured Text",
|
| 74 |
+
placeholder="Paste your messy meeting notes, emails, or raw text here...",
|
| 75 |
+
lines=12
|
| 76 |
+
)
|
| 77 |
+
|
| 78 |
+
schema_input = gr.Textbox(
|
| 79 |
+
label="2. What fields do you want to extract?",
|
| 80 |
+
placeholder="e.g., Company Name, Contact Person, Deadline, Action Items (list)",
|
| 81 |
+
lines=3
|
| 82 |
+
)
|
| 83 |
+
|
| 84 |
+
extract_btn = gr.Button("Extract Structured Data", variant="primary")
|
| 85 |
+
|
| 86 |
+
# Right Column: Output
|
| 87 |
+
with gr.Column():
|
| 88 |
+
json_output = gr.JSON(label="Structured Output")
|
| 89 |
+
|
| 90 |
+
# Connect the button to the function
|
| 91 |
+
extract_btn.click(
|
| 92 |
+
fn=extract_data,
|
| 93 |
+
inputs=[raw_input, schema_input],
|
| 94 |
+
outputs=json_output
|
| 95 |
+
)
|
| 96 |
+
|
| 97 |
+
# Launch the app
|
| 98 |
+
if __name__ == "__main__":
|
| 99 |
+
demo.launch()
|