Create app.py

app.py

@@ -0,0 +1,234 @@
+import gradio as gr
+import requests
+import base64
+import re
+import json
+from PIL import Image
+import io
+import tempfile
+import os
+
+def get_molecular_structure(compound_name):
+    """
+    Get molecular structure image for a chemical compound
+    Returns: dict with image URL, metadata, and status
+    """
+    try:
+        # Clean compound name
+        clean_compound = re.sub(r'[^a-zA-Z0-9\s-]', '', compound_name.strip())
+        if not clean_compound:
+            return {
+                "success": False,
+                "error": "Invalid compound name",
+                "image_url": None
+            }
+        
+        print(f"Processing compound: {clean_compound}")
+        
+        # Step 1: Get InChIKey from NIH Resolver
+        inchikey_url = f"https://cactus.nci.nih.gov/chemical/structure/{clean_compound}/stdinchikey"
+        
+        try:
+            response = requests.get(inchikey_url, timeout=10)
+            if response.status_code == 200 and response.text.strip():
+                inchikey = response.text.strip()
+                # Remove InChIKey= prefix if present
+                inchikey = inchikey.replace('InChIKey=', '')
+                print(f"Found InChIKey: {inchikey}")
+            else:
+                raise Exception("InChIKey not found")
+        except Exception as e:
+            print(f"InChIKey lookup failed: {e}")
+            # Try direct NIH image as fallback
+            return get_direct_nih_image(clean_compound)
+        
+        # Step 2: Get CID from PubChem
+        cid_url = f"https://pubchem.ncbi.nlm.nih.gov/rest/pug/compound/inchikey/{inchikey}/cids/JSON"
+        
+        try:
+            response = requests.get(cid_url, timeout=10)
+            if response.status_code == 200:
+                cid_data = response.json()
+                if 'IdentifierList' in cid_data and 'CID' in cid_data['IdentifierList']:
+                    cid = cid_data['IdentifierList']['CID'][0]
+                    print(f"Found CID: {cid}")
+                else:
+                    raise Exception("CID not found in response")
+            else:
+                raise Exception(f"CID request failed: {response.status_code}")
+        except Exception as e:
+            print(f"CID lookup failed: {e}")
+            return get_direct_nih_image(clean_compound)
+        
+        # Step 3: Get image from PubChem
+        image_url = f"https://pubchem.ncbi.nlm.nih.gov/rest/pug/compound/cid/{cid}/PNG"
+        
+        try:
+            response = requests.get(image_url, timeout=15)
+            if response.status_code == 200 and len(response.content) > 1000:
+                # Save image temporarily and return public URL
+                return save_and_return_image(response.content, {
+                    "compound": compound_name,
+                    "cid": cid,
+                    "inchikey": inchikey,
+                    "source": "PubChem"
+                })
+            else:
+                raise Exception("Invalid image data")
+        except Exception as e:
+            print(f"PubChem image failed: {e}")
+            return get_direct_nih_image(clean_compound)
+            
+    except Exception as e:
+        return {
+            "success": False,
+            "error": str(e),
+            "image_url": None,
+            "compound": compound_name
+        }
+
+def get_direct_nih_image(compound_name):
+    """Fallback to direct NIH image"""
+    try:
+        nih_url = f"https://cactus.nci.nih.gov/chemical/structure/{compound_name}/image"
+        response = requests.get(nih_url, timeout=15)
+        
+        if response.status_code == 200 and len(response.content) > 1000:
+            return save_and_return_image(response.content, {
+                "compound": compound_name,
+                "source": "NIH Direct"
+            })
+        else:
+            return {
+                "success": False,
+                "error": "No molecular structure found",
+                "image_url": None,
+                "compound": compound_name
+            }
+    except Exception as e:
+        return {
+            "success": False,
+            "error": f"NIH fallback failed: {str(e)}",
+            "image_url": None,
+            "compound": compound_name
+        }
+
+def save_and_return_image(image_data, metadata):
+    """Save image and return file path for Gradio"""
+    try:
+        # Create PIL Image
+        image = Image.open(io.BytesIO(image_data))
+        
+        # Save to temporary file
+        temp_file = tempfile.NamedTemporaryFile(delete=False, suffix='.png')
+        image.save(temp_file.name, 'PNG')
+        
+        return {
+            "success": True,
+            "image_path": temp_file.name,  # For Gradio display
+            "image_url": f"data:image/png;base64,{base64.b64encode(image_data).decode()}",  # For API
+            "metadata": metadata,
+            "compound": metadata.get("compound", ""),
+            "size": len(image_data)
+        }
+    except Exception as e:
+        return {
+            "success": False,
+            "error": f"Image processing failed: {str(e)}",
+            "image_url": None
+        }
+
+def molecular_structure_interface(compound_name):
+    """Gradio interface function"""
+    if not compound_name or not compound_name.strip():
+        return None, "Please enter a compound name"
+    
+    result = get_molecular_structure(compound_name.strip())
+    
+    if result["success"]:
+        info = f"""
+**Compound:** {result['compound']}
+**Source:** {result['metadata']['source']}
+**Image Size:** {result['size']} bytes
+"""
+        if 'cid' in result['metadata']:
+            info += f"**PubChem CID:** {result['metadata']['cid']}\n"
+        if 'inchikey' in result['metadata']:
+            info += f"**InChIKey:** {result['metadata']['inchikey']}\n"
+        
+        return result["image_path"], info
+    else:
+        return None, f"❌ Error: {result['error']}"
+
+def api_endpoint(compound_name):
+    """API endpoint that returns JSON"""
+    result = get_molecular_structure(compound_name)
+    return result
+
+# Create Gradio interface
+with gr.Blocks(title="Molecular Structure Generator") as demo:
+    gr.Markdown("# 🧪 Molecular Structure Generator")
+    gr.Markdown("Generate molecular structure images for chemical compounds using PubChem and NIH databases.")
+    
+    with gr.Row():
+        with gr.Column():
+            compound_input = gr.Textbox(
+                label="Chemical Compound Name",
+                placeholder="Enter compound name (e.g., 'aspirin', 'caffeine', 'Rochelle salt')",
+                value="Rochelle salt"
+            )
+            generate_btn = gr.Button("Generate Structure", variant="primary")
+        
+        with gr.Column():
+            image_output = gr.Image(label="Molecular Structure", type="filepath")
+            info_output = gr.Markdown(label="Information")
+    
+    generate_btn.click(
+        molecular_structure_interface,
+        inputs=[compound_input],
+        outputs=[image_output, info_output]
+    )
+    
+    # Examples
+    gr.Examples(
+        examples=[
+            ["Rochelle salt"],
+            ["aspirin"],
+            ["caffeine"],
+            ["benzene"],
+            ["glucose"],
+            ["ethanol"]
+        ],
+        inputs=[compound_input]
+    )
+    
+    # API section
+    gr.Markdown("## 🔗 API Usage")
+    gr.Markdown("""
+    **Endpoint:** `POST /api/molecular-structure`
+    
+    **Example Request:**
+    ```bash
+    curl -X POST "https://your-space.hf.space/api/molecular-structure" \
+         -H "Content-Type: application/json" \
+         -d '{"compound_name": "aspirin"}'
+    ```
+    
+    **Example Response:**
+    ```json
+    {
+      "success": true,
+      "image_url": "data:image/png;base64,iVBORw0KGgoAAAA...",
+      "metadata": {
+        "compound": "aspirin",
+        "cid": 2244,
+        "inchikey": "BSYNRYMUTXBXSQ-UHFFFAOYSA-N",
+        "source": "PubChem"
+      },
+      "size": 3456
+    }
+    ```
+    """)
+
+if __name__ == "__main__":
+    demo.launch(server_name="0.0.0.0", server_port=7860)