Update main.py

main.py

@@ -59,15 +59,21 @@ def extract_json_from_response(response_text):
         logging.error(f"Unexpected error extracting JSON: {str(e)}")
         raise
 
-def process_with_gemini(model, prompt, text):
+def process_with_gemini(model, prompt, text, answers_text=None):
     """Send combined text (job description and applicant documents) to Gemini and return structured JSON response."""
+    
+    # Build the applicant section with answers if provided
+    applicant_section = f"Applicant Documents:\n{text}"
+    if answers_text:
+        applicant_section += f"\n\nApplicant's Answers to Screening Questions:\n{answers_text}"
+    
     full_prompt = f"""
 You are an expert HR recruiter. Based on the job description and the applicant's documents provided below, assess the applicant's suitability for the job.
+
 Job Description:
 {prompt}
 
-Applicant Documents:
-{text}
+{applicant_section}
 
 Provide a detailed analysis that includes:
 - A score out of 100 evaluating the applicant's fit for the role.
@@ -96,6 +102,7 @@ def process_application():
       - job_id: form data string
       - job_description: form data string
       - applicant_id: form data string
+      - questions_answers: (optional) form data string - JSON string containing Q&A pairs
       - files: one or more PDF files (e.g., CV, cover letter)
     Returns a JSON response with score, recommendation, and justification.
     """
@@ -104,6 +111,7 @@ def process_application():
         job_id = request.form.get('job_id')
         job_description = request.form.get('job_description')
         applicant_id = request.form.get('applicant_id')
+        questions_answers = request.form.get('questions_answers')  # New field for Q&A
 
         if not job_id or not job_description or not applicant_id:
             return jsonify({'error': 'job_id, job_description, and applicant_id are required'}), 400
@@ -132,15 +140,34 @@ def process_application():
             # Extract text and combine
             combined_text += read_pdf(temp_file.name) + "\n\n"
 
+        # Parse and format questions_answers if provided
+        answers_text = None
+        if questions_answers:
+            try:
+                qa_data = json.loads(questions_answers)
+                # Format the Q&A data for better readability
+                formatted_answers = []
+                if isinstance(qa_data, list):
+                    for idx, qa in enumerate(qa_data, 1):
+                        question = qa.get('question', 'N/A')
+                        response = qa.get('response', 'N/A')
+                        formatted_answers.append(f"Q{idx}: {question}\nA{idx}: {response}")
+                else:
+                    formatted_answers.append(str(qa_data))
+                answers_text = "\n\n".join(formatted_answers)
+            except json.JSONDecodeError:
+                logging.warning("questions_answers is not valid JSON, treating as plain text")
+                answers_text = questions_answers
+
         # Configure Gemini and process combined text
         model = configure_gemini(api_key)
-        extracted_data = process_with_gemini(model, job_description, combined_text)
+        extracted_data = process_with_gemini(model, job_description, combined_text, answers_text)
 
         # Clean up temporary files
         for temp_file in temp_files:
             os.remove(temp_file)
 
-        # Optionally, you could log or store the job_id and applicant_id alongside the result
+        # Log the processing
         logging.info(f"Processed application for applicant_id: {applicant_id} for job_id: {job_id}")
 
         return jsonify(extracted_data)