import json from typing import Dict from datetime import datetime from core.doc_processor import ( extract_text_from_bytes, extract_text_from_docx, parse_pdf_date ) from core.text_processor import chunk_text from core.vector_store import create_or_load_vectorstore, similarity_search from services.llm_service import get_llm, get_unified_resume_prompt from services.db_service import store_candidate_data async def analyze_resume_document( file_bytes: bytes, job_role: str, file_extension: str ) -> Dict: if file_extension.lower() == ".pdf": resume_text, mod_date = extract_text_from_bytes(file_bytes) file_modified_date = parse_pdf_date(mod_date) if mod_date else datetime.now() elif file_extension.lower() == ".docx": resume_text, mod_date = extract_text_from_docx(file_bytes) file_modified_date = mod_date if mod_date else datetime.now() else: raise ValueError(f"Unsupported file format: {file_extension}") if resume_text.startswith("Error"): raise ValueError(resume_text) chunks = chunk_text(resume_text) vector_store = create_or_load_vectorstore(chunks) analysis_result = analyze_resume_with_llm(vector_store, resume_text, job_role, file_modified_date) try: db_record = await store_candidate_data(analysis_result) analysis_result["id"] = db_record.get("id") analysis_result["stored_in_db"] = True except Exception as e: print(f"Error storing candidate data: {str(e)}") analysis_result["stored_in_db"] = False analysis_result["db_error"] = str(e) return analysis_result def analyze_resume_with_llm(vectorstore, resume_text: str, job_role: str, file_modified_date: datetime) -> Dict: try: retrieved_docs = similarity_search(vectorstore, job_role) relevant_content = "\n".join([doc.page_content for doc in retrieved_docs]) unified_prompt = get_unified_resume_prompt() llm = get_llm() chain = unified_prompt | llm response = chain.invoke( { "resume_text": resume_text, "relevant_content": relevant_content, "job_role": job_role, } ) response_text = response.content.strip() if response_text.startswith("`"): response_text = response_text.strip("`") if response_text.startswith("json"): response_text = response_text[4:] parsed_response = json.loads(response_text.strip()) default_response = { "candidate_name": "", "gender": None, "date_of_birth": None, "email": None, "mobile_number": None, "address": None, "pan_number": None, "aadhar_number": None, "passport_number": None, "highest_education": None, "university_name": None, "designation": None, "skills": "", "skill_score": [], "total_experience": None, "region": None, "relevant_experience": None, "experience_summary": "No relevant experience analysis available", "hr_questions": [] } result = {**default_response, **parsed_response} result["file_modified_date"] = file_modified_date return result except json.JSONDecodeError as e: print(f"JSON Parse Error: {str(e)}") return { "candidate_name": "", "gender": None, "date_of_birth": None, "email": None, "mobile_number": None, "address": None, "pan_number": None, "aadhar_number": None, "passport_number": None, "highest_education": None, "university_name": None, "designation": None, "skills": "", "skill_score": [], "total_experience": None, "region": None, "relevant_experience": None, "experience_summary": "Could not parse resume details from response", "hr_questions": [], "file_modified_date": file_modified_date } except Exception as e: print(f"Unexpected Error: {str(e)}") return { "candidate_name": "", "gender": None, "date_of_birth": None, "email": None, "mobile_number": None, "address": None, "pan_number": None, "aadhar_number": None, "passport_number": None, "highest_education": None, "university_name": None, "designation": None, "skills": "", "skill_score": [], "total_experience": None, "region": None, "relevant_experience": None, "experience_summary": f"Error analyzing resume: {str(e)}", "hr_questions": [], "file_modified_date": file_modified_date }