Update app.py

app.py

@@ -544,87 +544,81 @@ async def voice_chat_endpoint(
 
 @app.post("/analyze-report")
 async def analyze_clinical_report(
-    file: Optional[UploadFile] = File(None),
-    text: Optional[str] = Form(None),
+    file: UploadFile = File(...),
+    patient_id: Optional[str] = Form(None),
     temperature: float = Form(0.5),
     max_new_tokens: int = Form(1024)
 ):
     """
-    Analyze a clinical patient report either from uploaded file or direct text input.
+    Analyze a clinical patient report from an uploaded file.
     
     Parameters:
     - file: Uploaded clinical report file (PDF, TXT, DOCX)
-    - text: Direct text input of the clinical report
+    - patient_id: Optional patient ID to associate with this report
     - temperature: Controls randomness of response (0.1-1.0)
     - max_new_tokens: Maximum length of response
     
     Returns structured analysis of the patient report.
     """
     try:
-        # Validate input
-        if not (file or text):
-            raise HTTPException(status_code=400, detail="Either file or text input is required")
+        # Validate file type
+        content_type = file.content_type
+        allowed_types = [
+            'application/pdf',
+            'text/plain',
+            'application/vnd.openxmlformats-officedocument.wordprocessingml.document'
+        ]
         
-        # Extract text from file if provided
-        report_text = text
-        if file:
-            if file.filename.lower().endswith('.pdf'):
-                # PDF processing
-                import PyPDF2
-                pdf_reader = PyPDF2.PdfReader(file.file)
-                report_text = "\n".join([page.extract_text() for page in pdf_reader.pages])
-            elif file.filename.lower().endswith(('.txt', '.md')):
-                # Plain text
-                report_text = (await file.read()).decode('utf-8')
-            elif file.filename.lower().endswith(('.docx', '.doc')):
-                # Word document
-                from docx import Document
-                doc = Document(io.BytesIO(await file.read()))
-                report_text = "\n".join([para.text for para in doc.paragraphs])
-            else:
-                raise HTTPException(status_code=400, detail="Unsupported file format")
+        if content_type not in allowed_types:
+            raise HTTPException(
+                status_code=400,
+                detail=f"Unsupported file type: {content_type}. Supported types: PDF, TXT, DOCX"
+            )
+
+        # Read file content
+        file_content = await file.read()
         
-        # Clean and validate the extracted text
-        if not report_text or len(report_text.strip()) < 50:
-            raise HTTPException(status_code=400, detail="Report text is too short or empty")
+        # Extract text based on file type
+        if content_type == 'application/pdf':
+            text = extract_text_from_pdf(file_content)
+        elif content_type == 'text/plain':
+            text = file_content.decode('utf-8')
+        elif content_type == 'application/vnd.openxmlformats-officedocument.wordprocessingml.document':
+            doc = Document(io.BytesIO(file_content))
+            text = "\n".join([para.text for para in doc.paragraphs])
         
-        # Create analysis prompt
-        prompt = (
-            "You are a senior clinical analyst. Analyze this patient report and provide:\n"
-            "1. SUMMARY: Concise summary of key findings\n"
-            "2. DIAGNOSES: List of confirmed or suspected diagnoses\n"
-            "3. RISK FACTORS: Important risk factors identified\n"
-            "4. RED FLAGS: Any urgent concerns that need attention\n"
-            "5. RECOMMENDATIONS: Suggested next steps for care\n\n"
-            f"PATIENT REPORT:\n{report_text[:15000]}"  # Limit input size
-        )
-        
-        # Get AI analysis
-        raw_response = agent.chat(
-            message=prompt,
-            history=[],
-            temperature=temperature,
-            max_new_tokens=max_new_tokens
+        # Clean and validate text
+        text = clean_text_response(text)
+        if len(text.strip()) < 50:
+            raise HTTPException(
+                status_code=400,
+                detail="Extracted text is too short (minimum 50 characters required)"
+            )
+
+        # Analyze the report
+        analysis = await analyze_patient_report(
+            patient_id=patient_id,
+            report_content=text,
+            file_type=content_type,
+            file_content=file_content
         )
-        
-        # Structure the response
-        structured_response = structure_medical_response(raw_response)
-        
-        # Add suicide risk assessment
-        risk_level, risk_score, risk_factors = detect_suicide_risk(raw_response)
-        structured_response["suicide_risk"] = {
-            "level": risk_level.value,
-            "score": risk_score,
-            "factors": risk_factors
-        }
-        
-        return JSONResponse(content=structured_response)
-    
+
+        return JSONResponse(content={
+            "status": "success",
+            "analysis": analysis,
+            "patient_id": patient_id,
+            "file_type": content_type,
+            "file_size": len(file_content)
+        })
+
     except HTTPException:
         raise
     except Exception as e:
-        logger.error(f"Error analyzing clinical report: {e}")
-        raise HTTPException(status_code=500, detail="Error processing clinical report")
+        logger.error(f"Error in report analysis: {str(e)}")
+        raise HTTPException(
+            status_code=500,
+            detail=f"Failed to analyze report: {str(e)}"
+        )
 if __name__ == "__main__":
     import uvicorn
     uvicorn.run(app, host="0.0.0.0", port=8000)